Microporous and Mesoporous Materials

Transcription

1 Microporous and Mesoporous Materials 139 (2011) Contents lists available at ScienceDirect Microporous and Mesoporous Materials journal homepage: Synthesis of hydrophilic and hydrophobic xerogels with superior properties using sodium silicate Pradip B. Sarawade a, Jong-Kil Kim b, Askwar Hilonga a, Dang Viet Quang a, Hee Taik Kim a, a Department of Chemical Engineering, Hanyang University, 1271 Sa 3-dong, Sangnok-gu, Ansan-si, Gyeonggi-do , Republic of Korea b E&B Nanotech. Co., Ltd., Ansan-si, Gyeonggi-do, Republic of Korea article info abstract Article history: Received 2 August 2010 Received in revised form 20 September 2010 Accepted 19 October 2010 Available online 23 October 2010 Keywords: Silica xerogels Ambient pressure drying Inexpensive Hydrophobic Highly porous hydrophilic and hydrophobic silica xerogels were synthesized by surface modification of silica hydrogels at ambient pressure drying. The silica hydrogels were prepared by a sol gel polymerization of an inexpensive silica precursor (sodium silicate) under atmospheric conditions. In order to minimize shrinkage due to drying, the hydrogel surface was modified using trimethylchlorosilane (TMCS) in the presence of ethanol/n-hexane solution before ambient pressure drying (APD). Properties of the final product were investigated using Field-Emission Scanning Electron Microscopy (FE-SEM), Transmission Electron Microscopy (TEM), Fourier Transform Infrared Spectroscopy (FTIR), Thermogravimetric and Differential Analysis (TG DTA), and nitrogen physisorption studies. The final product was observed to have an extremely high specific surface area (783 m 2 /g) and a large cumulative pore volume (2.74 cm 3 /g). Highly porous hydrophilic xerogels were obtained after heat-treating the modified xerogels. At temperatures above 450 C the surface alkyl groups (ACH 3 ) were significantly oxidized and, consequently, the properties of the resulting xerogels were altered. Products obtained via the proposed inexpensive approach have superior properties and the method exploits an inexpensive silica source (sodium silicate). Thus it is feasible for large-scale economic industrial production. Ó 2010 Elsevier Inc. All rights reserved. 1. Introduction Corresponding author. Tel.: ; fax: address: khtaik@yahoo.com (H.T. Kim). Aerogels are the most highly porous nanostructured materials. They exhibit large surface area (1200 m 2 /g), high porosity (80 98%), low bulk density (0.03 g/cm 3 ), extremely low thermal conductivities (0.005 W/mk), and unique acoustic properties (sound velocities as low as 100 m/s) [1,2]. Because of these properties, aerogels are utilized as thermal super-insulators in solar energy systems, refrigerators, and thermal flasks [3]. Despite these applications, the high production costs have thus far prevented their commercial use. Meanwhile, applications for porous silica xerogels continuously expand as their production costs decrease and their properties improve. Hydrophobic and hydrophilic silica xerogels with superior physical properties such as high surface area and large pore volume have potential applications in fields such as adsorbents, separations, biomedicine, sensors, drug delivery systems, catalyst carriers, thermal insulation, glazing, paints, and oil spill clean-up [4 8]. Conventional silica xerogels have relatively high density, low surface area, and small pore volume, restricting their applications. Recent observations suggest that the properties of porous materials improve following modification with silica gels (alcogel or hydrogel) during synthesis before the ambient pressure drying (APD) [9 13]. Moreover, silylating hydrogels and drying at ambient pressure can give less-dense silica xerogels. During the drying process, non-polar alky groups (which repel each other) replace surface OH groups, resulting in the spring back-effect, which preserves the silica gel network and, hence, the porosity [14]. Surface modification of silica hydrogels by alkyl groups has been reported to preserve the porous network even after drying at ambient pressure [15]. Prakash et al. [16] have synthesized silica aerogel films at ambient pressure via solvent exchange and surface modification processes. Solvent exchange is a lengthy and tedious process because it simply depends on diffusion of the solution within the gel. Hence, its take several days to produce silica aerogels at ambient pressure. Schwertfeger et al. [17] developed a new synthesis for sodium silicate-based silica aerogel at ambient pressure. Since then, many researchers have focused on synthesizing sodium silicate-based silica aerogels at ambient pressure. Nevertheless, the solvent exchange process, which is required for silica aerogel synthesis at ambient pressure, makes it a tedious process. Recently, Shi et al. [18] reported a new method, called one-step solvent exchange and surface modification process. This method is based on combining different solvents (trimethylchlorosilane (TMCS), n-hexane, and ethanol) for surface modification. Though Shi et al. used ethanol to reduce the reaction between the silylating /$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi: /j.micromeso

2 P.B. Sarawade et al. / Microporous and Mesoporous Materials 139 (2011) agent and water, the ethanol consumes large amount of the silylating agent in the surface modification process. We found that this process may be more suitable for surface modification of high content silica hydrogels, as there is less water present in the pore compared to low content silica hydrogels, leaving less silylating agent consumed. Here we observed that very little silylating reagent is required for surface modification of water silica gels with low pore content than those with higher pore content. The amount of silylating agent consumed for surface modification of silica hydrogels depends upon the amount of pore water, since the silylating reagent can directly react with pore water. Also, modifying silica hydrogels (high pore water content) can take a long time to replace the pore fluids. Therefore, a one-step solvent exchange and surface modification process is more suitable to produce hydrophobic silica xerogel, reducing both the time and the amount of silylating agent required. This study is not intended to compare silica aerogel and xerogel properties. Our intention was to produce high porous silica xerogels (hydrophobic and hydrophilic) with extreme physical properties, such as high surface area and large pore volume, using one-step solvent exchange and surface modification of silica hydrogels and drying at ambient pressure. In the present study, we synthesized low density, relatively transparent, very high specific surface area, and high pore volume hydrophobic silica xerogels by surface modification of hydrogel followed by APD. The silica precursor utilized in this study is relatively inexpensive (sodium silicate) and the reaction procedures employed are considerably versatile. We employed a simultaneous solvent exchange and surface modification process (one-step solvent exchange and surface modification) to reduce the synthesis duration and the drying shrinkage of silica xerogels (at an ambient pressure) from 7 to 2 days. The results for the modified silica xerogel were compared with those for the unmodified silica xerogel. Moreover, hydrophilic silica xerogels resulted from heattreating modified silica xerogel. The specific surface area, pore volume and pore diameter of TMCS modified hydrophobic silica gel increase slightly with increase in heating temperature from 150 to 500 C. This paper reports, in detail, the results obtained. 2. Materials and methods 2.1. Preparation of silica hydrogel using sodium silicate (water-glass) Silica hydrogel was prepared by sol gel polymerization using sodium silicate as a silica precursor (molar ratio SiO 2 :Na 2 O = 3.4), purchased from Shinwoo Materials Co. Ltd., South Korea. Trimethylchlorosilane (TMCS) (silylating agent) and sulfuric acid (acid catalyst) were purchased from Duksan chemical. Scheme 1 shows the method used to prepare the mesoporous hydrophobic and hydrophilic silica xerogels at an ambient pressure via simultaneous solvent exchange and surface modification. In order to prepare the silica sol, the aqueous sodium silicate solution (molar ratio SiO 2 :Na 2 O = 3.4) was mixed with purified water and sulfuric acid (40%) through a line mixer and finally through a nozzle as shown in Scheme 1. The flow rate of sodium silicate solution (200 g/min), purified water (150 g/min), and sulfuric acid (200 g/min) was controlled by using chemical feed pump and air chamber. Then the acidic (ph 1) silica sol was collected in a gelation unit prior to gelation. The gelation time was about 4 min, since the silica concentration in the silica sol solution was considerably high and the ph was strongly acidic. After gelation, the resultant silica hydrogels were aged for two different aging conditions i.e. (i) acidic low temperature (ALT) and (ii) basic high temperature (BHT) in order to see the effect of aging time and temperature on the textural properties of the silica xerogels. For this, 100 g of silica xerogel was aged in acidic condition (ph 4) and the temperature of 40 C for various durations (0 70 h). For the second aging condition, the same amount of silica xerogel was aged in basic condition (ph 9) and the temperature of 70 C for various durations (0 70 h) Washing (removal of byproducts), solvent exchange/surface modification and drying silylated wet gels The prepared silica gel was thoroughly washed with a continuous flow of water, for 12 h to remove trapped sodium ions (Na + ) from the wet silica gel. The removal of Na + from the washed gel was confirmed by a sodium ion detector (NeoMet, ISTEK, ph/ise meter). Furthermore, the washed gel was chemically modified with trimethylchlorosilane in the presence of n-hexane/ethanol solution. In this case, the gel (100 g) was mixed with n-hexane (100 ml) solution, TMCS, and ethanol for 12 h at room temperature (25 C). In all cases, the molar ratio of ethanol/tmcs (M 1 ) was fixed at 1. Ethanol was used to slow down the reaction between pore water and TMCS. The molar ratio of n-hexane/tmcs (M 2 ) was fixed at to minimize shrinkage, as reported elsewhere [19]. After surface modification, the gel was removed and washed with fresh n-hexane to remove unreacted TMCS. The resultant material was dried at an ambient pressure (room temperature, 25 C) for 12 h, followed by further drying at ambient air (80 C for 2 h). These drying procedures reduced shrinkage. To completely evaporate the pore liquid, the product was finally dried at 150 C for an hour at ambient pressure. The hydrophobic silica gels were cooled to room temperature and characterized using various techniques. Furthermore, to obtain hydrophilic silica xerogel, the hydrophobic xerogel samples were heated in air at various temperatures: 150, 200, 250, 300, 350, 400, 450, and 500 C for an hour Characterization methods Silica xerogel hydrophobicity and hydrophilicity were examined by measuring the percentage of absorbed water after putting the samples directly on a water surface. The weight change over time was examined using an electronic microbalance (Model OHAUS EPG214C, USA) at 10 5 g accuracy. Oil absorption studies were carried out on silica xerogel powder. Hydrophobic and hydrophilic silica xerogels were crushed into refined powder, and oil absorption was measured as follows. Ten grams of the sample was placed on a polyethylene plate (which does not absorb oil). Di-n-butyl phthalate (DOP) was dropped little by little on the center of each sample from a burette and was thoroughly kneaded with a spatula after each drop. The dropping and kneading were repeated until the entire mixture no longer became a solid putty lump. The amount of DOP used was determined, and oil absorption was expressed as ml of oil absorbed/g of the sample. The tapping densities of the silica xerogel were calculated from the mass to volume ratios. The volume was calculated by placing the silica xerogel in a graduated cylinder. An electronic microbalance (Model OHAUS EPG214C USA) measured the mass. The percentage porosity of the prepared silica xerogel was determined as follows: % Porosity ¼ð1 q b =q s Þ100 ð1þ where, q s and q b are the skeletal and bulk densities (of silica xerogel), respectively. The specific surface area and pore size distributions (PSDs) of xerogel were analyzed using Brunauer Emmet and Teller (BET) and BJH nitrogen gas adsorption and desorption methods (ASAP 2020, Micromeritics, USA). The BJH nitrogen gas absorption method was used to obtain the average pore diameter for the silica gel. BET analysis from the N 2 gas adsorbed at various partial pressures (10 points 0.05 < p/p o < 0.3, nitrogen molecular cross sectional area = nm 2 ) was employed to determine the surface area,

3 140 P.B. Sarawade et al. / Microporous and Mesoporous Materials 139 (2011) Chemical feed pump Air Chamber Flow line SiO 2 :Na 2 O =3.4 Line Mixer Nozzle Purified Water Chemical feed pump Air Chamber Flow line Hydrosol (Silica sol) Chemical feed pump Air Chamber Flow line H 2 SO 4 (40%) (Gelation Unit) Valve Drying & Heat-treatment Surface Modification Purified Hydrogel Aging & Washing Scheme 1. A flow chart showing experimental procedures for the synthesis of hydrophobic and hydrophilic mesoporous silica gels at an ambient pressure. and a single condensation point (p/p o = 0.99) was used to find the cumulative pore volume. Before N 2 adsorption, the sample was degassed at 150 C. Pore size distributions were calculated from the desorption isotherms [20 22]. To study the thermal stability of the xerogel, in terms of hydrophobicity retention, samples were examined by Thermo Gravimetric and Differential Thermal Analysis (TG DTA). Ten milligrams of hydrophobic xerogel was heattreated in air from 25 to 1000 C at a rate of 1.5 C min 1 using a microprocessor based Parr temperature controller (Model 4846) connected to a muffle furnace (A.H. JEON Industrial Co. Ltd., Korea). The thermal stability refers to the temperature at which the silica xerogel retains its hydrophobicity [23]. Surface modification was confirmed using Infrared (IR) spectroscopy, Perkin-Elmer (Model No. 783). For this purpose, the silica xerogel was ground into a refined powder, mixed with KBr, and pressed to form a sample pellet for FTIR measurements. Microstructure studies of xerogel samples were carried out by Field-Emission Scanning Electron Microscopy (FE-SEM) and transmission electron microscopy (TEM, JSM 6700 F, JEOL). 3. Results and discussion calculated by the concentration of excess sulfuric acid (H 2 SO 4 ) in the silica sol. These data can be explained using the general theory of silica polymerization. The condensation polymerization shows that silica gelation in aqueous solution involves an ionic mechanism. As the amount of sulfuric acid in the sol increases from 0.1 to 1 N, the gelation time decreases from 60 to 4 min (Fig. 1). The decrease arises from the increase in acid in the sol, increasing the condensation reaction rate. Therefore, silica clusters aggregate faster, forming a three-dimensional, porous silica network in a short time [28]. At relatively lower sulfuric acid concentrations silica particles have very little ionic charge. Consequently, they first aggregate into chains and finally into a three-dimensional gel network. Below 0.5 N sulfuric acid the silica particles are positively charged; hence, they repel each other. Sols with the longest gelation time have a maximum temporary stability at about 1 N because they are positively charged between 0.5 N and 1 N sulfuric acid; hence the gelation time of silica sols increases significantly because they attract each other. At around 1 N the silica monomers rapidly convert into particles that simultaneously aggregate into a gel network Effect of excess H 2 SO 4 in silica sol on gelation time Sodium silicate (Na 2 SiO 3 ) has been, and probably always been, the cheapest source of relatively pure silicic acid to make silica gel. Sodium silicate reacts with water and acid (sulfuric acid) to give silicic acid as shown in the following chemical equation: Na 2 SiO 3 þ H 2 O þ H 2 SO 4! SiðOHÞ 4 þ Na 2 SO 4 The silicic acid condenses to form small silica particles, chains, and consequently a silica network (silica gel/slurry) as shown below: SiðOHÞ 4 þðohþ 4 Si!ðOHÞ 3 BSiAOASiBðOHÞ 3 þ H 2 O These reaction mechanisms have been suggested to take place when using other sources of silica [24 26]. The silica gel preparation involves a sol-to-gel transition. This transition is referred as a gelation, where the sol becomes highly viscous and ceases to move [27]. The gelation time depends strongly on the amount of catalyst added to the sol (Fig. 1). The degree of gelation was ð2þ ð3þ Gelation time (minute) Normality (N) of excess H 2SO 4 in silica sol Fig. 1. Effect of normality (N) of excess H 2 SO 4 in silica sol on gelation time.

4 P.B. Sarawade et al. / Microporous and Mesoporous Materials 139 (2011) Variation of textural properties with aging conditions (ph, temperature and time) Aging conditions (temperature, time, and ph) have a strong impact on the textural properties of dried silica xerogels, including BET surface area, pore volume, and pore size. In order to observe the effects of ph and aging temperature with respect to aging time on textural properties of dried silica xerogel, the gels were treated with two different conditions: (i) acidic low temperature and (ii) basic high temperature for various durations (time periods), as shown in Figs It is evident that, at basic high temperature, the surface area gradually decreases from 588 to 190 m 2 /g over time. On the other hand, the BET surface area for acidic low temperature samples decreased slightly from 588 to 445 m 2 /g. A decrease in BET surface area at higher temperatures is likely due to hydrothermal effects on the silica gel [29 31]. In Figs. 3 and 4 the pore volume and pore size gradually increase over time for basic high temperature samples. In addition, pore size and pore volume at above 20 h increased suddenly (uptake) at basic high temperature than at acidic low temperature aging conditions. This can be attributed to the precipitation rate of silica monomers into the pores as well as into the surface of the gel network. The precipitation rate is higher at above 20 h at basic high temperature than at acidic low temperature. If the precipitation occurs at the pore walls, there is a chance to decrease the pore size which will contribute to the microporosity. This increase in microporosity contributes to the increase in internal surface area and thereby the total pore volume [32,33]. Moreover, this difference is due to the fact that the condensation in silica gels continues long after BET specific surface area (m 2 /g) : ph 4.0, Temp. 40 o C (ALT) : ph 9.0, Temp. 70 o C (BHT) Time (hrs) Fig. 2. Effect of two aging conditions on the BET surface area of the silica gels: (i) acidic low temperature (ALT) and (ii) basic high temperature (BHT). Pore volume (cm 3 /g) : ph 4.0, Temp. 40 o C (ALT) : ph 9.0, Temp. 70 o C (BHT) Time (hrs) Fig. 3. Effect of two aging conditions on the pore volume of the silica gels: (i) acidic low temperature (ALT) and (ii) basic high temperature (BHT). BET average pore diameter (nm) : ph 4.0, Temp. 40 o C (ALT) : ph 9.0, Temp. 70 o C (BHT) Time (hrs) Fig. 4. Effect of two aging conditions on the pore size of the silica gels: (i) acidic low temperature (ALT) and (ii) basic high temperature (BHT). gelation [33,34]. Moreover, aging silica hydrogels may modify their texture by continued condensation reactions leading to a more branched gel network. A more branched gel network leads to larger interconnected particles resulting in a larger pore size [35,36]. In addition, aging at higher temperature allowed reorganization of the pore structure by several processes, such as dissolution, condensation, de- and re-polymerization, and syneresis, resulting in increased pore size and pore volume [37]. The data obtained for the samples aged at acidic low temperature and basic high temperature are complied in Table Simultaneous solvent exchange and surface modification of silica wet gel Generally, colloidal silica polymerization forms a weak network. During the drying process, liquid and vapor coexist within the gel pores. As the liquid begins to evaporate, a meniscus forms at the liquid vapor interface. The formation of a liquid vapor interface within the gel results in surface tension and creates concave menisci in the gel pores. With the progressive evaporation, the menisci recede into the gel body and build a compressive force that acts on the pore walls. This force causes considerable shrinkage due to partial collapse of the gel network [38]. In addition, the Table 1 Effects of washing ph and aging temperature on the physical and textural properties of the unmodified water-glass-based silica xerogels (hydrophilic) dried at ambient pressure (APD). No. Aging time (h) Bulk density (g/cm 3 ) Porosity (%) Oil absorption (ml/g) Water absorption (ml/g) Acidic low temperature (ALT) (ph 4; temp, 40 C) (a 1 ) (b 1 ) (c 1 ) (d 1 ) (e 1 ) (f 1 ) (g 1 ) Basic high temperature (BHT) (ph 9; temp, 70 C) (a 2 ) (b 2 ) (c 2 ) (d 2 ) (e 2 ) (f 2 ) (g 2 )

5 142 P.B. Sarawade et al. / Microporous and Mesoporous Materials 139 (2011) terminal silanol groups (SiAOH) present on the silica surface undergo a condensation reaction forming new siloxane bonds, which ultimately result in irreversible shrinkage [39]. Attaching the nonpolar alkyl group to the silica surface replaces OH with H, reducing irreversible shrinkage. Also, gel collapse ceases when the gel structure is strong enough to withstand the tensile strength of the liquid [40]. Therefore, to reduce irreversible shrinkage during drying, the gel surface was organically modified with tri-methyl groups (found in TMCS). The modification was carried out via a simultaneous solvent exchange and surface modification process using ethanol/ TMCS/n-hexane solution. The expected chemical reactions for the simultaneous solvent exchange/surface modification of the silica hydrogel are presented as follows: C 2 H 5 OH þðch 3 Þ 3 ASiACl!ðCH 3 Þ 3 ASiAOACH 2 CH 3 ½ETMSŠþHCl ð4þ 2ðCH 3 Þ 3 SiACl½TMCSŠþH 2 O!ðCH 3 Þ 3 ASiAOASiAðCH 3 Þ 3 ½HMDSOŠ þ 2HCl ð5þ 2ðCH 3 Þ 3 ASiAOACH 2 CH 3 þ H 2 O!ðCH 3 Þ 3 ASiAOASiAðCH 3 Þ 3 þ 2C 2 H 5 OH ð6þ ðch 3 Þ 3 ASiACl þ BSiAOH! BSiAOASiAðCH 3 Þ 3 þ HCl ð7þ Thus, TMCS reacts with ethanol (Eq. (4)), pore water (Eq. (5)), and the OH group on the silica surface (Eq. (7)), to form ethyltrimethoxysilane [ETMS] and hexamethyldisiloxane (HMDSO); consequently, the hydrophilic surface of the silica network becomes hydrophobic. As the reaction proceeds, transferable yellowish liquids (HCl/residual ethanol) spontaneously come out of the wet gel [41]. During the simultaneous solvent exchange/surface modification process, the reaction between ethanol and TMCS can slow down the reaction rate of TMCS with the OH group on the silica surface. The latter process spontaneously replaces pore water with n-hexane. Chemical surface modification of the hydrogel by nonpolar alkyl/aryl groups is an indispensable step before APD, as it prohibits the formation of new siloxane bonds between the adjacent silica cluster and thereby irreversible gel shrinkage [42,43]. In the present work, tri-methyl groups present in the trimethylchlorosilane organically modified the surface of the water-glass based hydrogel by simultaneous solvent exchange and surface modification, as explained above Hydrophobic properties of sodium silicate-based xerogel Surface modification of all samples was carried out for 12 h at room temperature (25 C). As shown in Fig. 5, the hydrophobicity of silica xerogel granules was measured by putting samples directly on a water surface and measuring the increase in weight after 3 months. The results are tabulated in Table 2. The data for the surface modified silica gel (xerogel granules) have been compared with those for the unmodified silica gel (xerogel). Although the wettability of materials depends upon their surface chemistry is well known, however, water absorption studies show that the unmodified silica xerogel (hydrophilic xerogel) absorbed more water than the TMCS surface modified silica xerogel (hydrophobic xerogel). This discrepancy arises from the non-polar methyl groups (ACH 3 ) attached to the silica gel surface, which repels water molecules, consequently reducing water absorption. The mesoporous silica xerogel granules exhibited hydrophobic behavior as a result of the surface modification. Surface modification of silica xerogel granules was also confirmed by using Fourier Transform Infrared Spectroscopy (FTIR). The FTIR spectra of modified (hydrophobic) xerogel with TMCS, unmodified xerogel (hydrophilic), and heattreated xerogels are provided in Fig. 6. The strong absorption peak at 1260 cm 1 corresponds to the terminal ACH 3 group, and the peak at 840 cm 1 is due to SiAC bonding. These absorption peaks are attributed to the surface modification by TMCS [36,44]. The absorption peak centered at 976 cm 1 corresponds to the OH groups absorbed on the silica surface, the bending of HAOAH bonds, and the stretching of the SiAOH bonds [45]. The peaks centered at 1070 and 495 cm 1 correspond to the SiAOASi bonds and are the most informative of the silica network structure [46,47]. The two weak peaks at 976 and 960 cm 1 for the unmodified silica gel represent the SiAOH stretch vibrations, indicating that there is no modification of the silica gel surface. Moreover, there are no CAH peaks in the FTIR spectra of the unmodified silica gel. SiAOH and SiAH 2 O peaks are present in the FTIR spectra and show the hydrophilic nature of the unmodified gel [48]. A strong peak at 1256 cm 1 for the modified samples indicates significant surface modification by TMCS. The OH and SiAOH peaks are conspicuous for the unmodified xerogel. Fig. 6 also shows the FTIR spectra for heat-treated, TMCS modified xerogel. The peak at 1256 cm 1, which was present in modified xerogel, clearly disappears after heat treatment at 450 C. This change indicates that the attached group (ACH 3 ) gets oxidized at that temperature, converting the hydrophobic gel into its hydrophilic counterpart. The transition temperature of the xerogel (hydrophobic to hydrophilic) was determined using TG/DTA and water absorption measurements Oil adsorption studies Oil absorption studies were carried out on unmodified, modified, and heat-treated silica xerogel powders. The mesoporous hydrophobic (TMCS modified) and hydrophilic (heat-treated at 450 C) silica xerogels had maximum oil absorptions of more than 3.96 and 3.88 ml/g, respectively, measured using dioctyl phthalate. The unmodified xerogel powder has a very low oil absorption compared to the others. The latter retains its maximum porosity because it does not shrink. The method used to measure oil absorption was explained in detail in the characterization section, and the results are summarized in Tables TG DTA studies and the effect of heat-treatment on xerogel hydrophobicity Fig. 5. Photograph showing (a) hydrophobicity of the modified mesoporous silica gel on water surface and (b) its hydrophilicity when heat-treated at 450 C. In order to investigate the thermal stability/transition temperature (hydrophobic to hydrophilic transformation), 25 mg of xerogel was heated at various temperatures ( C). Also, water absorption in heat-treated xerogel was tested as shown in Fig. 5(a and b). Xerogels heat-treated at temperatures below 250 C do not absorb water and float on the surface (Fig. 5(a)). Xerogels heat-treated at 350 C absorb some water, partially sinking below the surface but still floating. Xerogels heat-treated at

6 P.B. Sarawade et al. / Microporous and Mesoporous Materials 139 (2011) Table 2 Surface modification effects (TMCS) with respect to aging conditions on the physical and textural properties of the water-glass-based silica xerogels (hydrophobic) dried at ambient pressure (APD). No. Aging time (h) Bulk density (g/cm 3 ) BET surface area (m 2 /g) Pore diameter (nm) Pore volume (cm 3 /g) Porosity (%) Oil absorption (ml/g) Acidic low temperature (ALT) (ph 4; temp, 40 C) (a 11 ) (b 11 ) (c 11 ) (d 11 ) (e 11 ) (f 11 ) (g 11 ) Basic high temperature (BHT) (ph 9; temp, 70 C) (a 22 ) (b 22 ) (c 22 ) (d 22 ) (e 22 ) (f 22 ) (g 22 ) Water absorption (after 3 month) (ml/g) Relative intensity (a. u.) OH Si-CH 3 OH 3400 Series4 Series7 Series10 (a) Series13 (b) (c) (d) (e) 2900 Unmodified 150 o C (modified) 250 o C 350 o C 450 o C 2400 Si-CH Si-O-Si 900 Si-C Si-CH Wave number (cm -1 ) Fig. 6. FTIR spectra of (a) unmodified silica gel together with those modified with (b) TMCS and heat-treated at (c) 250 C, (d) 350 C and (e) 450 C. 450 C completely sink, settling at the bottom (Fig. 5(b)). Further increases in heating temperature make the xerogel more hydrophilic. The maximum temperature at which the xerogel retains its hydrophobic character (before becomes hydrophilic) was 440 C. At this temperature the surface methyl groups (ACH 3 ) become oxidized, forming a hydrophilic xerogel. The thermal stability of the xerogel and the oxidation temperature for the ACH 3 groups were estimated by thermogravimetric (TG) and differential thermal (DT) analyses, respectively. The TG DTA results for hydrophobic xerogels were compared with those for unmodified xerogel. Fig. 7(a and b) shows the TG DT analyses of the mesoporous TMCS modified and unmodified xerogels heat-treated in an oxygen atmosphere up to 1000 C. The curves clearly show that the modified xerogel exhibited a negligible weight loss up to 440 C, beyond which the xerogel underwent a significant weight loss. This decrease in weight is due to oxidation of surface methyl groups, which can be clearly seen by two sharp exothermic peaks in the DTA curve when the temperature is raised above 440 C [49]. On the other hand, the unmodified silica gel showed a continuous weight loss with increased temperature (Fig. 7(b)) Nitrogen physisorption studies The impact of surface modification by TMCS on textural properties of water-glass based xerogels was investigated by BET analysis. Tables 1 and 2 show the physical and textural properties of xerogels synthesized without (unmodified) and with (modified) TMCS. Table 3 shows the effect of heat-treatment on the physical and textural properties of the hydrophobic xerogel with respect to the two aging conditions (acidic low temperature and basic high temperature) and heating temperature (450 C). TMCS has a significant effect on the physical and textural properties of xerogel samples. TMCS altered specific surface area, average pore size, cumulative pore volume, tapping density, and percentage porosity. The tapping density of the unmodified xerogel is much higher (0.57 g/cm 3 ) than that of its modified counterpart (0.098 g/cm 3 ) and nearly equal to that of the ambient pressure-dried silica aerogel (0.05 g/cm 3 ) [50]. As illustrated in Table 2, the modified xerogel possesses high surface area (783 m 2 /g), large pore size (9.21 nm), and high pore volume (2.74 cm 3 /g). On the other hand, the unmodified xerogel shows low surface area (588 m 2 /g), low

7 144 P.B. Sarawade et al. / Microporous and Mesoporous Materials 139 (2011) Table 3 Effect of heat treatment (450 C) on the physical and textural properties of the TMCS surface modified water-glass-based silica xerogel. No. Heating temp. ( C) Bulk density (g/cm 3 ) Pore diameter (nm) Pore volume (cm 3 /g) Porosity (%) Oil absorption (ml/g) Acidic low temperature (ALT) (ph 4; temp, 40 C) (a 111 ) (a 111 ) (a 111 ) (a 111 ) (a 111 ) (a 111 ) (a 111 ) (a 111 ) Basic high temperature (BHT) (ph 9; temp, 70 C) (a 222 ) (a 222 ) (a 222 ) (a 222 ) (a 222 ) (a 222 ) (a 222 ) (a 222 ) Water absorption (after 3 month) (ml/g) pore volume (0.78 cm 3 /g), and small pore size (5.10 nm). These data indicate that the hydrophobic surface of the pore network reduces the capillary pressure by lowering the surface tension, consequently reducing shrinkage during the aging and drying stages. Previous studies [13] also reported the same phenomenon, wherein surface modification of the silica gel increases its BET surface area. Furthermore, the repulsion between ACH 3 groups on the surface during the final drying stages also tends to expand the surface and affects partial recovery of the wet gel structure [51]. The percentage porosity depends on gel shrinkage during drying. As reported in Tables 1 and 2, the percentage porosity for the xerogels prepared by surface modification was considerably higher (95%) than that for the unmodified xerogels (69.04%). This difference in percentage porosity can be attributed to the nonpolar methyl groups attached to the silica surface. The methyl groups are responsible for the spring back effect [52,53], which preserves the highly porous silica network and reduces collapse. Table 3 illustrates the changes in physical properties of TMCS surface-modified xerogel after heating at 450 C for 1 h. The specific surface area, cumulative pore volume, and average pore diameter increase slightly with increase in heating temperature from 150 to 500 C (Fig. 8, Table 3). Generally, when the xerogel is heated BET specific surface area (m 2 /g) Heating temperature ( o C) Fig. 8. Variation of BET surface area with the increase in heating temperature from 150 to 500 C. Weight (%) 120 (a) Modified (b) Unmodified 100 TGA 440 o C TGA DTA DTA Temperature ( o C) DTA (uv/mg) Exo up Volume adsorbed (cm 3 /g) STP (a) Unmodified (b) Modified (c) Heat-treated Desorption Adsorption Relative Pressure (p/p o) Fig. 7. TG/DTA curves of (a) hydrophilic (unmodified), and (b) hydrophobic (modified) mesoporous silica gel. Fig. 9. N 2 adsorption/desorption isotherms of (a) unmodified silica gel together with (b) those modified with TMCS and (c) heat-treated at 450 C.

8 P.B. Sarawade et al. / Microporous and Mesoporous Materials 139 (2011) dv/dlog(d) (cm 3 /g.nm) Pore diameter (nm) (a) (b) (c) Unmodified Modified Heat-treated Fig. 10. Pore size distributions of (a) unmodified silica gel together with (b) those modified with TMCS and (c) heat-treated at 450 C. at high temperature its particle size (particle diameter) decreases while the pore size increases. Hence increase in pore size, internal pore volume, and pore surface may increase. Consequently, the total pore volume and the total specific surface area increased. This observation is consistent with another report in which it was concluded that the specific surface area increases with the decrease in particle diameter and increase in pore diameter, after heating the gel above 150 C [54 56]. In addition, the pore size and the pore volume increased with increase in heating temperature (Table 3). The increase in total pore volume with increase in heating temperature can be attributed to the extraction of residual solvent (water) which may create more micropore surface area and hence the total pore volume increased. Hence, there was a significant increase in total pore volume, total specific surface area, and total micropore surface area. Also, after heating the modified silica gel at high temperature (above 450 C), the alkyl groups (ACH 3 ) that were attached on the silica surface were oxidized as shown in TG DTA graph (Fig. 7). This oxidation of surface methyl groups may create more micro and mesopores; hence, the total pore volume and, consequently, the total specific surface area may increase with increase in heating temperature (Fig. 8). The xerogel porosities (unmodified, modified, and heat-treated) were further studied by examining nitrogen adsorption desorption isotherms and pore size distributions. Fig. 9 shows the N 2 adsorption/desorption isotherms of unmodified, modified (TMCS), and heat-treated xerogels. The unmodified samples show an adsorption/desorption curve with hysteresis behavior typical of xerogels, which indicates inkbottle shaped pores (H4 type) [48]. On the contrary, the isotherm curves of the modified and heat-treated xerogels are very similar to those of the supercritical/ambient pressure-dried silica aerogels (H1 type) [37], although with a relatively low surface area and pore size. H1 type hysteresis loops indicate a cylindrical-like pore [11]. The limited N 2 uptake at low relative pressures (<0.1) indicates the existence of mesoporous structures [40,57]. The isotherm of the heat-treated xerogel decreased slightly compared to that of its heat-treated counterpart. The increase in pore volume is associated with the removal of residual organic components and ACH 3 groups from the gel. The physisorption isotherms obtained for all xerogels exhibit hysteresis loops which correspond to the characteristic features of mesoporous materials (Type IV isotherms) [58,59]. Fig. 10 shows the pore size distribution (PSDs) curves of unmodified, modified, and heat-treated xerogels. The unmodified xerogel shows a narrow pore size distribution with an average diameter of 5.10 nm, while the modified xerogel shows a broad distribution with an average diameter of 9.21 nm. The heat-treated xerogel shows a broader pore size distribution with an average diameter of nm. After heating the modified samples at 450 C the peak pore diameter shifted to the higher value. This shift arises from the removal of residual organic compounds and ACH 3 Fig. 11. FE-SEM micrograph of (a) unmodified silica gel together with those modified with (b) TMCS and (c) heat-treated at 450 C.

9 146 P.B. Sarawade et al. / Microporous and Mesoporous Materials 139 (2011) Fig. 12. TEM micrograph (low and high magnification) of (a) unmodified silica gel together with (b) those modified with TMCS and (c) heat-treated at 450 C. groups from the gel. As per the IUPAC classification of pores [60], the modified and heat-treated samples show a pronounced peak in the mesopore region (2 50 nm), indicating that the xerogels maintain mesoporosity even after high temperature heattreatment FE-SEM and TEM studies The xerogel mesoporosity was confirmed using Field-Emission Scanning Electron Microscopy (FE-SEM) and Transmission Electron Microscopy (TEM). Fig. 11 shows FE-SEM micrographs of (a) unmodified, (b) TMCS modified, and (c) heat-treated TMCS modified xerogels. Generally, unmodified xerogels have a less porous structure with dense aggregated spheres. The unmodified wet silica gel shrinks more during APD, leading to a dense microstructure of dried xerogel and a loss of mesopores [61]. The modified xerogel shows highly porous structures (Fig. 11(b)) due to reversible shrinkage (spring back effect) that takes place during drying, a consequence of organic modification of tri-methyl groups on the silica surface. The modified xerogel has a mesopore in the structure with an average diameter of 9.21 nm. Fig. 11(c) shows an FE-SEM micrograph of heat-treated modified xerogel. The TMCS modified xerogel that was heat-treated at 450 C appears to have more regular spheres that are small and uniform in size. The pore size for the heat-treated xerogel is wider than that for the unheated counterpart. Fig. 12 shows representative TEM images of unmodified, modified (TMCS), and heat-treated TMCS modified xerogels. The unmodified xerogel shows more dense aggregated silica particles as expected. Samples modified with TMCS and heated at 450 C exhibited a highly porous sponge-like silica network microstructure with a uniform pore size distribution in the range of 9 15 nm. The surface morphology observed by TEM for heat-treated xerogel is more uniform and has a smaller particle size than that for the unheated xerogel. 4. Conclusions Hydrophobic and hydrophilic mesoporous sodium silicatebased silica xerogels were obtained by simultaneous solvent exchange and surface modification of wet silica gel with trimethylchlorosilane (TMCS) followed by ambient pressure drying. A hydrophilic xerogel with better physicochemical properties was obtained by heating the TMCS modified silica gel at 450 C. The silica wet gel was obtained by a novel fast gelation of colloidal silica sol. The surface modifying agent (TMCS) as well as the heattreatment process has a strong effect on the properties of the final product. The properties examined were surface area, pore volume, pore diameter, hydrophobicity, and morphology. The TMCS surface-modified xerogel has high surface area (783 cm 2 /g), low density (0.098 g/cm 3 ), large pore diameter (9.21 nm), and extremely large pore volume (2.74 cm 3 /g). Heat-treatment had a great impact on the xerogel, mainly on hydrophobicity, surface area, pore size, and pore volume. Hydrophilic mesoporous xerogel with superior surface area (788 cm 2 /g), pore diameter (14.49 nm), and density (0.137 g/cm 3 ) was obtained when modified with TMCS and heattreated at 450 C. This product has properties desirable for various applications and sanctions the proposed synthesis, which exploits a low-cost silica source (sodium silicate), for the large-scale commercial production of hydrophobic and hydrophilic silica xerogels at an ambient pressure. Acknowledgement This research is supported by the collaborative research Program among industry, academia, and research institutes through Korea Industrial Technology Association (KOITA) funded by the Ministry of Education Science and Technology (KOITA- 2010).

10 P.B. Sarawade et al. / Microporous and Mesoporous Materials 139 (2011) References [1] L.W. Hrubesh, Chem. Ind. 17 (1990) 824. [2] G.C. Bond, S. Flamerz, Appl. Catal. 33 (1987) 219. [3] A.V. Rao, R.R. Kalesh, Sci. Technol. Adv. Mater. 4 (2003) 509. [4] G.M. Pajnok, Appl. Catal. 72 (1991) 21. [5] A.S. Dias, M. Pillinger, A.A. Valente, Micropor. Mesopor. Mater. 94 (1 3) (2006) 214. [6] M.R. Jamali, Y. Aassadi, F. Shemirani, M.R.M. Hosseini, R.R. Kozani, M.M. Farahani, M.S. Niasani, Anal. Chem. Acta 579 (1) (2006) 68. [7] Q. Tang, Y. Xu, D. Wu, Y. Sun, J. Solid State Chem. 179 (5) (2006) [8] A.V. Rao, N.D. Hegde, H. Hirashima, J. Colloid Interface Sci. 305 (1) (2007) 124. [9] A.Y. Jeong, S.M. Koo, D.P. Kim, J. Sol Gel Sci. Technol. 19 (2000) 483. [10] M.C. Capel-Sanchez, L. Barrio, J.M. Campos-Martin, J.L.g. Fierro, J. Colloid Interface Sci. 277 (2004) 146. [11] G.T. Burns, Q. Deng, R. Field, J.M. Hahn, C.W. Lengtz, Chem. Mater. 11 (1999) [12] V.C. Menon, S. Komarneni, M. Park, M. Schmucker, H. Schnedier, J. Sol Gel Sci. Technol. 11 (1998) 7. [13] P.B. Sarawade, J.K. Kim, H.K. Kim, H.T. Kim, Appl. Surf. Sci. 254 (2007) 574. [14] G.M. Pajonk, S.J. Tichner, in: J. Fricke (Ed.), Proceedings of the First International Symposium on Aerogels, September, Wurzburg, Germany, 1985, p [15] S.T. Reed, C.S. Ashley, C.J. Brinker, R.J. Walko, R. Ellefsoon, J. Gill, SPIE 1328 (1990) 220. [16] S.S. Prakash, C.J. Brinker, A.J. Hurd, S.M. Rao, Nature 374 (1995) 439. [17] F. Schwertfeger, D. Frank, M. Schmidt, J. Non-Cryst. Solids 225 (1998) 24. [18] F. Shi, L. Wang, J. Liu, Mater. Lett. 60 (2006) [19] S.W. Hwang, T.Y. Kim, S.H. Hyun, J. Colloid Interface Sci. 322 (2008) 224. [20] C.J. Brinker, S.W. Sherere, Sol Gel Science. The Physics and Chemistry of Sol Gel Processing, Academic Press, San Diego, 1990, p [21] R. Rouquerol, D. Avnir, C.W. Fairbridge, D.H. Everett, J.H. Haynes, N. Pernicone, J.D.F. Ramsay, K.S.W. Sing, K.K. Unger, Pure Appl. Chem. 66 (1994) [22] F. Rouquerol, K.S.W. Sing, Jean Rouquerol Adsorption by Powders and Porous Solids: Principles, Methodology and Applications, Academic Press, [23] Z. Bi, Z. Zhang, F. Xu, Y. Qian, J. Yu, J. Colloid Interface Sci. 214 (1999) 368. [24] F.D. Ovcharenako, Y.I. Tarasevich, Kolloidn Zh. 35 (3) (1973) 467. [25] E.M. Flanigen, R.W. Grose, U.S. Pat. 3494,874 (Union Carbide Corp., 1970). [26] K.J. Murata, W.G. Schlecht, U.S. Geol. Surv. Bull. 950 (1946) [27] Sao Carlos (SP), Sol Gel Sci. Tech., in: M.A. Aegerter, Jafelicci Jr., D.F. Souza, E.D. Zanotto (Eds.), World Scientific, Brazil, 1989, pp [28] D.J. O Connor, B.A. Sexton, Surface Analysis Method in Materials Science, Springer, [29] V.V. Sidorchuk, R. Leboda, D.G. Maydannik, V.A. Tertykh, J. Colloid Interface Sci. 171 (1995) 168. [30] S.S. Hayrapetyan, H.G. Khachatryan, J. Colloid Interface Sci. 276 (2004) 373. [31] R.A. Strom, Y. Masmoudi, A. Rigacci, G. Petermann, l. Gullberg, B. Chevalier, M.A. Einarsrud, J. Sol Gel Sci. Technol. 41 (2007) 291. [32] M.E. Winyall, E.G. Acker, US. Pat., 3959,174 (1976). [33] D.J. Suh, T.J. Park, J. Mater. Sci. Lett. 16 (1997) 490. [34] S. Smitha, P. Shajesh, P.R. Aravind, S.R. kumar, P.K. Pillai, K.G.K. Warrier, Micropor. Mesopor. Mater. 91 (2006) 286. [35] C.Y. Kim, A.R. Jang, B.I. Kim, D.H. Suh, J. Sol Gel Sci. Technol. 48 (2008) 336. [36] R. Takahashi, K. Nakanishi, N. Soga, J. Non-Cryst. Solids 189 (1995) 66. [37] D.J. Suh, T.J. Park, J.H. Sonn, J.C. Lim, J. Mater. Sci. Lett. 18 (1999) [38] H.D. Gesser, P.C. Goswami, Chem. Rev. 89 (1989) 765. [39] S.D. Bhagat, Y.H. Kim, G.B. Yi, Y.S. Ahn, J.G. Yeo, Micropor. Mesopor. Mater. 253 (2006) [40] C.E. Kim, J.S. Yoon, H.J. Hwang, J. Sol Gel Sci. Technol. 49 (2009) 47. [41] L.J. Wang, S.Y. Zhao, M. Yang, Mater. Chem. Phys. 113 (2009) 485. [42] R. Deshpande, D.M. Smith, C.J. Brinker, J. Non-Cryst. Solids 144 (1992) 32. [43] X.C. Zhou, L.P. Zhong, Y.P. Xu, Inorg. Mater. 44 (2008) 976. [44] C.J. Pouchert, The Aldrich Library of Infrared Spectra, third ed., Aldrich Chemistry Co., Wisconisan, WI, [45] S. Lee, Y.C. Chad, Mater. Lett. 61 (2007) [46] R.A. Oweini, H.E. Rassy, J. Mol. Struct. 919 (2009) 140. [47] C.J. Brinker, S.W. Scherere, Sol Gel Science. The Physics and Chemistry of Sol Gel Processing, Academic Press, San Diego, 1990, p [48] H. Liu, W. Sha, A.T. Cooper, M. Fan, Colloid Surf. A: Physicochem. Eng. Aspects (2008). [49] A.P. Rao, A.V. Rao, G.M. Pajonk, P.M. Shewale, J. Mater. Sci. 42 (2007) [50] S.D. Bhagat, Y.H. Kim, Y.S. Ahn, J.G. Yeo, Micropor. Mesopor. Mater. 96 (2006) 237. [51] S.D. Bhagat, K.T. Park, Y.H. Kim, J.S. Kim, J.H. Han, Solid State Sci. 10 (9) (2008) [52] G.W. Scherer, S. Haereid, E. Nilsen, M.A. Einarsud, J. Non-Cryst. Solids 202 (1996) 104. [53] A.P. Rao, A.V. Rao, J.L. Gurav, J. Porous Mater. 15 (2008) 507. [54] Z.J. Li, C.R. Liu, Q.S. Zhao, J. Non-Cryst. Solids 265 (2000) 189. [55] S.K. Kang, S.Y. Choi, J. Mater. Sci. 35 (2000) [56] F. Shi, L. Wang, J. Liu, M. Zeng, J. Mater. Sci. Technol. 23 (2007) 402. [57] A. Fidalgo, M.E. Rosa, L.M. Ilharco, Chem. Mater. 15 (2003) [58] A.C. Pierre, E. Elaloui, G.M. Pajonk, Langmuir 14 (1998) 66. [59] W.C. Li, A.H. Lu, S.C. Guo, J. Colloid Interface Sci. 254 (2002) 153. [60] K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Moscou, R.A. Pierotti, J. Rouquerol, T. Siemieniewska, Pure Appl. Chem. 57 (4) (1985) 603. [61] J.L. Gurav, A.V. Rao, A.P. Rao, D.Y. Nadargi, S.D. Bhagat, J. Alloys Compd. 476 (2009) 397.