Preparation and characterization of nano-structured silica from rice husk

Preparation and characterization of nano-structured silica from rice husk

Materials Science and Engineering A364 (2004) 313–323 Preparation and characterization of nano-structured silica from rice husk Tzong-Horng Liou∗ Dep...

482KB Sizes 0 Downloads 15 Views

Materials Science and Engineering A364 (2004) 313–323

Preparation and characterization of nano-structured silica from rice husk Tzong-Horng Liou∗ Department of Chemical Engineering, Hsiuping Institute of Technology, 11 Gungye Rd., Dali City, Taichung 412, Taiwan, ROC Received 13 February 2003; received in revised form 14 August 2003

Abstract Uniformly sized ultrafine silica powder can be obtained by nonisothermal decomposition of rice husk in an oxidizing atmosphere. The properties of reactant and product including morphology, particle size, surface area, pore volume and pore distribution, have been investigated by TEM, SEM, XRD, FTIR, ICP-MS, and EA. At a heating rate of 5 K/min, the specific surface area of the silica powder was 235 m2 /g, the average pore diameter was 5.4 nm, and the average particle size was 60 nm. The products obtained from various heating rates were all amorphous. By using a thermogravimetric analysis technique, a mechanism including two reaction stages was observed for the thermal decomposition of rice husk in air. The activation energy was found to be 166 ± 10 kJ/mol. This method can conveniently provide preparation of silica of high surface area and nanometer grade. © 2003 Elsevier B.V. All rights reserved. Keywords: Rice husk; Silica; Nano-structure; Preparation; Characterization

1. Introduction Silica (SiO2 ) is a basic raw material that is widely used in electronics, ceramic, and polymer material industries. Because of their small-diameter particles, ultrafine silica powders have many technological applications, such as thixotropic agents, thermal insulators, composite fillers, etc. [1]. At present, nanoscale silica materials are prepared using several methods, including vapor-phase reaction, sol–gel and thermal decomposition technique [2–6]. However, their high cost of preparation has limited their wide application. In contrast, rice husk is an agricultural byproduct whose major constituents are organic materials and hydrated silicon. Because the silicon atoms in the rice husk have been naturally and uniformly dispersed by molecular units, very fine particle size, with very high purity and surface area silica powder can be prepared under controlled conditions. In addition, the reaction can occur more easily than the conventional experimental technique. This process has the benefit not only of producing valuable silica powder but also of reducing disposal and pollution problems. ∗

Tel.: +886-4-24961198; fax: +886-4-24961187. E-mail address: [email protected] (T.-H. Liou).

0921-5093/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2003.08.045

The organic materials in rice husk consist of celluloses (55–60 wt.%, including cellulose and hemicellulose) and lignin (22 wt.%) [7]. Approximately one-fifth of the ash is obtained on burning rice husk in air. The ash contains >90% silica by mass with minor amounts of metallic elements. Because the ash is obtained as a fine powder, it does not require further grinding [8], making it the most economical source of nanoscale silica. There have been several reports on the formation of silica from rice husk [9–11]. Real et al. [12] found that a homogeneous size distribution of nanometric silica particles could be obtained by burning rice husk at 873–1073 K in a pure oxygen atmosphere. Della et al. [13] observed that active silica with a high specific area could be produced from rice husk ash after heat-treating at 973 K in air. Kalapathy et al. [14] used an improved method to produce silica with lower sodium content by adding silicate solution to pH 1.5 hydrochloric, citric, or oxalic acid solutions until pH 4.0 was reached. Kamath and Proctor [15] reported that the rice husk could be dissolved in sodium hydroxide solution, and then titrated with acid to obtain silica gel. These previous studies focused on the process of manufacturing silica from rice husk. However, the mechanism of thermal decomposition as well as the surface characteristic of product under nonisothermal condition have received


T.-H. Liou / Materials Science and Engineering A364 (2004) 313–323

much less attention. In this paper, characterization of silica prepared by heating the rice husk at various heating rates (5, 10, 15, and 20 K/min) in air is investigated. In addition, the rice husk is treated with acid leaching to remove impurities from the samples. The physical and chemical examinations selected in the study include transmission electron microscope, scanning electron microscope, X-ray diffractometer, Fourier infrared spectrometer, N2 -adsorption apparatus, inductivity coupled plasma-mass spectrometer and element analyzer. A thermogravimetric analysis (TGA) is carried out to investigate the decomposition process of rice husk. A theoretical mechanism is also developed to account for the experimental results. These results of this study may be a potentially attractive method to produce valuable silicon nitride and silicon carbide powder [16–19], as well as the manufacture of supported-metal catalyst and zeolite with high value of specific surface area [20,21].

2. Experimental 2.1. Material used and sample preparation The samples of rice husk used in the study were obtained from a rice mill, and the rice husk was washed thoroughly with distilled water to remove adhering soil and dust. An adequate process of acid leaching was carried out as described in a previous report [22]. Analytical grade HCl solution (Merck & Co.) was used as the acid treatment. The water-rinsed rice husk and acidic solution were mixed under stirring in a glass round-bottomed flask at 373 K with a thermostat for 1 h. The solution was filtered and the husk was washed with distilled water several times until the filtrate was free from acid. The acid-leached husk was then dried at 373 K for 24 h. The samples were pulverized, and then ground and screened through an ASTM standard sieve to obtain the desired grain sizes (325 mesh size). 2.2. Experimental apparatus and procedures The preparation of silica was carried out using a combustion reaction in a tubular reactor made of quartz. First, a weighed amount of rice husk was packed into the reactor, and highly purified air (99.999%, San-Fu Chem. Co.) was used as reaction gas. The reactor was inserted into a furnace and heated at the desired rates. At the end of the heating period, the reactor was withdrawn from the furnace and cooled to room temperature. The experiments on the reaction characteristic of rice husk were carried out using a TGA technique, as shown in Fig. 1. A Perkin-Elmer TGA7 thermogravimetric analyzer was used for the thermal decomposition measurement. The main components of the system were a sensitive ultra-microbalance (as small as 0.1 ␮g), an alumina reactor, a movable furnace, and a computer. A platinum sample pan was used as

a sample container. The effect of resistance to thermal and mass transfer on oxidation reaction was eliminated by placing small amounts of specimen (6 ± 0.2 mg) into the sample pan. The reaction temperature was controlled by a standard platinum–rhodium thermocouple close to the sample. The flow rate of air was fixed at 60 ml/min. All silica samples were prepared by heating the rice husk at temperatures between 300 and 1000 K, using heating rates of 5, 10, 15, and 20 K/min. Several tests were performed to ensure that the sampling technique employed was valid. The remaining amount of rice husk decomposed in air is defined here as W/W0 , where W0 and W represent initial and instantaneous mass of sample, respectively. 2.3. Analysis of metallic impurities and organic elements The amounts of metallic impurities in the samples were estimated by an inductively coupled plasma-mass spectrometer (Kontron Plasmakon, model S-35). The reactant and product were dissolved and placed into a solution of HNO3 and HF. The amounts of fundamental organic element in raw material sample were determined using a Heraeus elemental analyzer. The dried sample was powdered to 325 mesh size (ASTM), and this powder was employed in the analysis. 2.4. Analysis of physical properties The crystalline structures of reactant and product were examined by X-ray diffractometer (Siemens, model D-500) using Cu K␣ and a scan speed of 2.5◦ /min. Transmission electron microscope (TEM) studies were performed on a Jeol JEM-1200CX II instrument operated at 120 kV. The sample was ultrasonically dispersed in alcohol for 20 min, and the suspension was then pipetted onto a copper microgrid with carbon film. Reactant and product morphology were inspected by scanning-electron microscope (Topcon, model ABT-150S). The specific surface area, pore volume and pore diameter of samples were measured by a N2 adsorption analyzer (Micrometric, model ASAP 2000). Prior to the measurement, samples were purged in a nitrogen flow at 423 K for 1 h. Further investigation of the samples decomposed at various heating rates was carried out by Fourier infrared spectrometer (Shimadzu, model FTIR-8300). The sample for IR analysis was prepared by mixing KBr with 10 wt.% silica powder and then pressing into a thin tablet.

3. Results and discussion 3.1. Analysis of metallic impurities and organic elements Metallic ingredients will have a substantial effect on the quality of silica from rice husk. Therefore, it is preferable to treat rice husk with an optimally acidic solution, so as to effectively diminish impurities and to obtain highly purified silica powder. The amounts of metallic ingredients present

T.-H. Liou / Materials Science and Engineering A364 (2004) 313–323


Fig. 1. Schematic diagram of the experimental apparatus for the preparation of silica from rice husk.

in the water-rinsed and acid-leached at various heating rates samples are given in Table 1. All the ash samples contain magnesium, calcium, iron, sodium, potassium, phosphorus, manganese, and zinc; of which the concentration of calcium is greatest. The water-rinsed sample has an impurity content of 4.1 wt.%. However, the acid-leached and heated samples have impurity content of 0.20–0.27 wt.%, indicating that the impurity content is significantly reduced by the acid-leached and thermal decomposition process. The decreased impurities mainly result from chemical reaction between acid and metals, after which the metals are removed by filtration. Additionally, the metals are also probably carried out from the volatiles during thermal decomposition. About 95% of impurities are extracted after treatment of the sample with

thermal decomposition. For various heating rates, the total residual amount of metallic ingredients from a high heating rate is larger than that from a low one. The resulting purity of silica is better than 99.7% after burning the rice husk. In addition, Conradt et al. [23] reported that omission of acid pre-treatment would yield a considerably reduced surface area, which was attributed to the presence of alkali promoting structural changes of the silica during heat treatment. Fundamental organic element analysis reveals that the average organic composition of rice husk was 38.01 wt.% carbon, 5.28 wt.% hydrogen, 36.10 wt.% oxygen, and 1.94 wt.% nitrogen. However, the reported compositions of rice husk differ widely, as affected by the type of paddy and climate [24].

Table 1 Amount of metallic ingredients in rice husk for both before and after burning samples β (K/min)

0a 5b 10b 15b 20b a b

Metallic ingredients as oxides (ppm) Mg


4680 84.7 222 85.8 65.4

15460 1810 1480 2000 1630

Sample was water-rinsed and unburnt. Sample was acid-leached and burnt in air.

Fe 750 <10 <10 45.4 84.1

Na 3550 <10 <10 <10 345


P 8970 <10 122 <10 37.3

5430 30 256 151 514

Mn 1800 <10 <10 <10 <10

Zn 200 <10 <10 <10 <10

Total 40840 1974.7 2120 2322.2 2695.8


T.-H. Liou / Materials Science and Engineering A364 (2004) 313–323

Fig. 2. Adsorption–desorption isotherms of nitrogen at 77 K for both before and after reaction samples: (a) specimen burnt at 5 K/min in air; (b) unburnt specimen.

3.2. Analysis of pore structures The nitrogen adsorption–desorption isotherms for reactant and silica sample prepared by heating the rice husk at 5 K/min, are shown in Fig. 2. The lower portion of the loop is traced out on adsorption, and the upper portion on desorption. Fig. 2a shows a hysteresis loop which resembles type IV of Brunaur’s classification. At low values of P/P0 the isotherm is similar to type II, but then adsorption increases markedly at P/P0 above 0.4, where pore condensation takes place. By changing the heating rate of 10–20 K/min, similar hysteresis loops were also observed. However, the range of relative pressures over which hysteresis was observed increased with increasing heating rate. These results indicate that the silica product is a porous material. The capillary condensation curve may be used to evaluate the pore size and pore volume distribution. Fig. 2b shows no obvious hysteresis loop for the raw material sample. In addition, the isotherm is convex over the entire range, indicating that the forces of adsorption between adsorbate and adsorbent are relatively weak. The difference in Fig. 2a and b indicates that pore opening takes place when the rice husk is burned in air, which is probably relative to the conversation of silanol groups to siloxane bridges [8]. The volume adsorbed plotted versus the adsorbed layer of thickness (t) for the rice husk ash samples, fired at various

heating rates, is shown in Fig. 3. There is a linear region extending through the origin, the slope of which is proportional to surface area [25]. From the t-plots, it can be seen that the volume adsorbed of samples increased with decreasing heating rate. The results indicate that the accessibility of nitrogen molecules into the pore system of these adsorbents can be increased by decreasing the heating rate. Ibrahim et al. [26] reported that the appearance of the upward deviation in the t-plot is a good criterion for the increased extent of pore vacation. Fig. 4 shows a plot of the increment of pore volume per increment in pore size, versus pore size. This result is determined by the Barrett–Joyner–Halenda (BJH) method (based on the desorption branch). The raw material sample has only a small pore size distribution, with maximum at about 3.5 nm in diameter. After burning the rice husk at 5–20 K/min, a double peak is observed in the silica samples having maxima at about 2.5 and 4.0 nm. The larger pores probably result from the residual spaces between fine particles existing or formed in the silica preparation, and the smaller pores are developed within the particles by the oxidation procedure. A same trend is also observed by Real et al. [12], for acid-leached rice husk annealed at 873 K in O2 . Table 2 summarizes the results of specific surface area and pore structure of samples reacted at various heating rates. The Brunauer–Emmett–Teller (BET) method is used for the

T.-H. Liou / Materials Science and Engineering A364 (2004) 313–323

Fig. 3. Nitrogen adsorption as a function of adsorbed layer of thickness (t) for both before and after reaction samples.

Fig. 4. Differential pore size distribution of the rice husk calculated from the desorption isotherm by the BJH procedure.



T.-H. Liou / Materials Science and Engineering A364 (2004) 313–323

Table 2 Surface area and pore characteristics of the rice husk for both before and after burning samples β (K/min)

BET surface area (m2 /g)

Total pore volume (ml/g)

Maximum pore volume (ml/g)

Average pore diameter (nm, 4V/A)

Median pore diameter (nm, 4V/A)

0a 5b 10b 15b 20b

10.4 235.0 232.9 228.8 185.2

0.025 0.318 0.313 0.307 0.248

0.002 0.055 0.053 0.052 0.041

9.56 5.42 5.47 5.27 5.36

0.962 0.710 0.717 0.713 0.706

a b

Sample was acid-leached and unburnt. Sample was acid-leached and burnt in air.

analysis of specific surface area. Comparing before and after reaction, the specific surface area is clearly increased. The silica samples are highly porous materials which have a larger internal surface area. A major contribution for the increase of specific surface area is that the organic matter has been broken up during the thermal decomposition of rice husk, thus leaving a highly porous structure. The table also shows that the specific surface area of silica powder decreased with increasing heating rate. This decrease of specific surface area can be attributed to portion of the pores having been destroyed and to pore structure variation, with increasing heating rate. At a heating rate of 5 K/min, the specific surface area of silica powder reaches value of 235 m2 /g. Compared to that, acid-leached rice husk pyrolyzed in N2 atmosphere had a surface area of 261 m2 /g [17], which is much higher than this present observation. As shown in Table 2, the total cumulative pore volume of the unreacted sample is 0.025 ml/g, much smaller than that of reacted samples, which have an average value of 0.30 ml/g. The total cumulative pore volume decreased from 0.32 to 0.25 ml/g as the heating rate increase from 5 to 20 K/min. A similar tread is also observed in the distribution of maximum pore volume. The average pore diameter (by BET method) of unreacted sample is 9.6 nm, which is much larger than that of reacted samples, which have an average value of 5.4 nm. The heating rate has a smaller effect on the pore size, and the sizes of average pore diameter and median pore diameter have not obviously changed with changes in the heating rate. The predominant type of pores was found to be mesopores. 3.3. Analysis of physical properties Fig. 5 shows X-ray diffraction analysis of rice husk before and after reaction at various heating rates. All samples are amorphous. However, in the ash samples (Fig. 5b–e), the typical silica characteristic is observed at a broad peak centered at 2θ = 22.5◦ , which can be attributed to the presence of disordered cristobalite. Chakraverty et al. [11] reported that the ash obtained from combustion of untreated and various HCl-treated husk samples of 973 K all showed amorphous form, which is consistent with the present study. They also reported that acid treatment of husk did not affect the silica structure.

Fig. 6 shows the resulting IR spectrum when the samples are heated at various heating rates. As can be seen, there are strong absorption peaks at ∼475, 805, and 1115 cm−1 , just as those from commercial grade silica, indicating the presence of silica. For various heating rates, the IR spectrum shows no significant changes in the peak position. However, the intensity of the peaks increases with increasing heating rate, indicating that the silica yield is increased with increasing heating rate, which is consistent with the results of reaction characteristic analysis, as indicated in Table 3. Fig. 7 shows SEM micrographs of rice husk sample and silica product prepared using 5 K/min in air. Fig. 7a shows the outer epidermis of rice husk, which is well organized and has a corrugated structure. Fig. 7b shows the inner epidermis of rice husk, which has a lamella structure. The morphology is different for the outer and inner surfaces of rice husk. The silica is mainly localized in the tough interlayer (epidermis) of the rice husk and also filling in the spaces between the epidermal cells [27,28]. Fig. 7c shows that the surface of ground rice husk powder was covered with flakes. When rice husk was burnt in air, Fig. 7d shows that many residual pores are distributed within the ash sample, indicating that the silica is a highly porous material with a large internal surface area. The rice husk might have become broken up during thermal decomposition of organic matter, thus leaving a highly porous structure. The silica obtained after burning of rice husk in air has a pure white colour. By comparison of these micrographs, it is observed that the surface of unreacted samples is relatively nonporous, whereas a burnt sample exhibits a porous surface, as indicated by the pore structure analysis (Fig. 2). After combustion of rice husk in air, the morphology tended to maintain its original shape, although the product

Table 3 Reaction characteristics of the rice husk β (K/min)

Ti (K)

Tm1 (K)

Tm2 (K)

Tf (K)

W∞ /W0 (%)

5 10 15 20

425.6 436.1 446.6 453.5

569.8 580.9 587.6 591.6

577.8 596.3 607.7 615.0

801.1 819.9 837.8 849.2

23.0 23.4 23.8 23.9

Samples were acid-leached and burnt in air.

T.-H. Liou / Materials Science and Engineering A364 (2004) 313–323

Fig. 5. X-ray diffractogram of rice husk: (a) unburnt specimen; (b)–(e) specimens burnt at 5, 10, 15, and 20 K/min in air, respectively.

Fig. 6. FTIR spectrogram of rice husk: (a)–(d) specimens burnt at 5, 10, 15, and 20 K/min in air, respectively.



T.-H. Liou / Materials Science and Engineering A364 (2004) 313–323

Fig. 7. Scanning electron micrographs of rice husk: (a) outer epidermis of rice husk (300×); (b) inner epidermis of rice husk (40×); (c) powder of unburnt specimen (1000×); (d) inner surface of burnt specimen at 5 K/min in air (1000×).

is highly brittle and loose, so when carefully pinched with the fingers, it could be divided into individual entities. On the other hand, the burnt-husk is easily converted into a ultrafine powder with very soft grinding in an agate mortar. The same observation was also reported by other workers [8,12,23]. The TEM of the silica sample which is ground into powder is presented in Fig. 8. Fig. 8a shows that uniform and nanosized silica powders were obtained when rice husk was burnt in air. The shape of the silica grains is spherical with an average particle size of 60 nm. Because the silicon is homogeneously distributed over the rice husk with molecular grade, therefore very small sized silica particles can be obtained. Fig. 8b shows that the other morphology of silica from burned of rice husk consisted of whisker-like structures ranging in width from 10 to 20 nm.

3.4. Analysis of reaction characteristics The broken lines (TG curves) in Fig. 9 represent the variation of weight remaining (W/W0 ) with reaction temperature, for the heating rates of 5, 10, 15, and 20 K/min. The TGA plots of rice husk combustion obtained from various heating rate are similar, and the thermal decomposition effect appears in the samples heated at 400–900 K. This weight decrease is attributed to the removal of combustible volatiles. It is also observed that there is one inflection point shown on the TG curves, indicating that the thermal decomposition of rice husk takes place in two main stages. Furthermore, with decreasing heating rate, the decomposition starts at lower temperatures because as the heating rate decreases, the time needed to achieve a certain temperature

T.-H. Liou / Materials Science and Engineering A364 (2004) 313–323


Fig. 8. Transmission electron micrographs of rice husk: (a) reacted specimen which is ground into powder (50,000×); (b) surface of reacted specimen which is ground into powder (100,000×).

increases [29]. From Fig. 9, about 53% of the total mass is decomposed and volatilized at the end of the first stage, with the remaining mass removed on completion of the second stage. Further observation shows that rapid decomposition of the rice husk powder takes place in the first stage, after which the decomposition is slower, perhaps because part of the reactant is not as easily decomposed in the second stage. The solid lines (DTG curves) in Fig. 9 represent the variation of derivative of weight (dW/dt) with temperature at each heating rate. The peak height and position are very sensitive to the heating rate. Note that in the first stage, there is an overlap of two peaks present in each heating rate. The peaks shift to the high-temperature region when the heating rate increases. Furthermore, the height of the peaks, which is associated with the instantaneous rate of thermal decomposition, also increases with the heating rate. Liou et al. [22] previously reported the pyrolysis kinetics of husk in nitrogen atmosphere, and also obtained the same conclusion. The reaction characteristics of rice husk are summarized in Table 3. In the reaction temperature range, a continuous increase in both initial (Ti ) and final (Tf ) reaction temperatures, with increasing heating rate, is obtained. According to the peak temperature (Tm1 and Tm2 ) at which instantaneous rate of thermal decomposition is maximum, the main reaction temperature of the rice husk is clearly in the range of 570–615 K. However, a temperature beyond 800 K is necessary for complete decomposition of the rice husk. Table 3 also shows that the average yield (W∞ /W0 ) of silica is about 23.5 wt.%. The yield is slightly increased by increasing the heating rate.

Liou et al. [22] employed the Friedman method to evaluate the apparent activation energy from the TG data for the pyrolysis of rice husk, with no oxygen present in the carrier gas. With oxygen present in the carrier gas in this work, the derivation of activation energy follows that of the previous investigation. The decomposition rate of rice husk, dα/dt, can be represented by an Arrhenius-type equation:   −Ea dα = k0 exp f(α) (1) dt RT where α = (initial mass−instantaneous mass)/(initial mass− final mass), t is the reaction time, k0 is the frequency factor, R is the gas constant, Ea is the activation energy, T is the reaction temperature in degrees Kelvin, and f(α) is a function of conversion independent of temperature. Taking a natural logarithm of Eq. (1), the plot of ln(dα/dt) versus 1/T from Fig. 9 at various heating rates yields a set of straight lines. The average activation energy value obtained from the slope of straight lines is 166 ± 10 kJ/mol. Results from Liou et al. [22] showed an activation energy of 116–205 kJ/mole in the pure nitrogen atmosphere. Thus, there is about 20% reduction of activation energy when oxygen is present in the carrier gas. The observed reaction characteristics of rice husk can be explained on the basis of the thermal decomposition behavior of its major constituents: celluloses, lignin, and ash (mostly silica). There are two stages in the generation of silica during the thermal decomposition process, and it can be clearly observed that the volatilization of gas products is accompanied by the formation of tar during the thermal decomposition reaction. From the DTG curves, there are two peaks observed in the first stage reaction. This indicates


T.-H. Liou / Materials Science and Engineering A364 (2004) 313–323

Fig. 9. Effect of heating rate on the thermal decomposition of the specimen: air flow rate, 60 ml/min; sample loading, 6 mg; initial grain size, 38–45 ␮m.

that there are two components of organic matter (mainly celluloses and lignin) participating in the first stage reaction. The first stage of the thermal decomposition is the loss of volatile compounds, followed by the decomposition of celluloses and lignin to the intermediate, which may be organic material of smaller molecular weight. In the second stage, there is no peak observed. Hence the thermal decomposition mechanism of rice husk in that stage is attributed to the further oxidation of carbon in the residual intermediate to form other volatile species, tar and char. The final char is composed of high-purity silica. From the above observation, there are two competitive reactions contained in the mechanism, so the thermal decomposition of celluloses and lignin in the rice husk results in the formation of silica.

Patel et al. [7] reported that the possible bonding of organic molecules and silicon in rice husk is mainly hemicellulose (a mixture of d-xylose, l-arabinose, methylglucoronic acid, and d-galactose). The reason for the formation was that the four organic materials in rice husk are aldehydes (monosaccharides) which become polar due to an electromeric effect. According to the above hypothesis, it can be concluded that the silicon atoms in rice husk attach to oxygen atoms, forming a silanol group (Si–OH). When the rice husk is burned in air, the dehydroxylation or conversion of silanol groups to siloxane bridges (Si–O) leads to pore opening of organic matters, and then transforms into amorphous silica [8,30]. Because the major constituents of rice husk include more than one organic component, the thermal decomposition reaction is complicated, with more than one

T.-H. Liou / Materials Science and Engineering A364 (2004) 313–323

pathway of chemical decomposition observed in the present work.

4. Conclusions Nano-structured silica powders with high specific surface area were obtained from nonisothermal decomposition of rice husk in an air atmosphere. The product was characterized in terms of silica content, specific surface area, pore volume, particle size and pore size distribution. The properties of silica in rice husk ash are strongly influenced by heating rate. About 95% of the impurities are extracted after thermal decomposition of the samples. Amorphous and ultrafine silica particles, average 60 nm in size, could be obtained. The results of TGA reveal that two reaction stages are included in the thermal decomposition system. The initial and final reaction temperatures, as well as the reaction range, all increase when the heating rate is increased. The observed thermal behavior is explained on the basic of two competitive reactions, degradation of celluloses and lignin, which are the major organic constituents of rice husk. The results of this study are useful for the preparation of valuable and widely applicable nanoscale silica from rice husk, also helping to solve disposal and pollution problems. Acknowledgements The author expresses thanks to the National Science Council of Taiwan for its financial support under Project No. NSC 89-2214—E164-001. References [1] L. Sun, K. Gong, Ind. Eng. Chem. Res. 40 (2001) 5861. [2] M. Tomozawa, D.L. Kim, V. Lou, J. Non-Cryst. Solids 296 (2001) 102.


[3] P.A. Tanner, B. Yan, H. Zhang, J. Mater. Sci. 35 (2000) 4325. [4] G. Wu, J. Wang, J. Shen, T. Yang, Q. Zhang, B. Zhou, Z. Deng, F. Bin, D. Zhou, F. Zhang, J. Non-Cryst. Solids 275 (2000) 169. [5] S. Sadasivan, D.H. Rasmussen, F.P. Chen, R.K. Kannabiran, Colloids Surf. A: Physicochem. Eng. Aspects 132 (1998) 45. [6] G.H. Bogush, M.A. Tracy, C.F. Zukoski IV, J. Non-Cryst. Solids 104 (1988) 95. [7] M. Patel, A. Karere, P. Prasanna, J. Mater. Sci. 22 (1987) 2457. [8] J. James, M.S. Rao, Am. Ceram. Soc. Bull. 65 (1986) 1177. [9] K. Kamiya, A. Oka, H. Nasu, J. Sol-Gel Sci. Technol. 19 (2000) 495. [10] K.G. Mansaray, A.E. Ghaly, Biomass Bioenerg. 17 (1999) 19. [11] A. Chakraverty, P. Mishra, H.D. Banerjee, J. Mater. Sci. 23 (1988) 21. [12] C. Real, M.D. Alcala, J.M. Criado, J. Am. Ceram. Soc. 79 (1996) 2012. [13] V.P. Della, I. Kuhn, D. Hotza, Mater. Lett. 57 (2002) 818. [14] U. Kalapathy, A. Proctor, J. Shultz, Bioresour. Technol. 85 (2002) 285. [15] S.R. Kamath, A. Proctor, Cereal Chem. 75 (1998) 484. [16] P. Grothy, G.M. Pudukottah, J. Am. Ceram. Soc. 82 (1999) 1393. [17] T.H. Liou, F.W. Chang, Ind. Eng. Chem. Res. 35 (1996) 3375. [18] A.W. Weimer, J.R. Cassiday, D.W. Susnitzky, C.K. Black, D.R. Beaman, J. Mater. Sci. 31 (1996) 6005. [19] B.B. Nayak, B.C. Mohanty, S.K. Singh, J. Am. Ceram. Soc. 79 (1996) 1197. [20] F.W. Chang, M.T. Tsay, M.S. Kuo, C.M. Yang, Appl. Catal. A 226 (2002) 213. [21] H. Hamdan, M.N.M. Muhid, S. Endud, E. Listiorini, Z. Ramli, J. Non-Cryst. Solids 211 (1997) 126. [22] T.H. Liou, F.W. Chang, J.J. Lo, Ind. Eng. Chem. Res. 36 (1997) 568. [23] R. Conradt, P. Pimkhaokham, U. Leela-Adison, J. Non-Cryst. Solids 145 (1992) 75. [24] V.M.H. Govindarao, J. Sci. Ind. Res. 39 (1980) 495. [25] C.N. Satterfield, Heterogeneous Catalysis in Industrial Practice, 2nd ed., McGraw-Hill, New York, 1991, p. 146. [26] D.M. Ibrahim, S.A. El-Hemaly, S.A. Abo-El-Enein, S. Hanafi, M. Helmy, Thermochim. Acta 37 (1980) 347. [27] N.K. Sharma, W.S. Williams, A. Zangvil, J. Am. Ceram. Soc. 67 (1984) 715. [28] R.V. Krishnarao, M.M. Godkhindi, Ceram. Int. 18 (1992) 243. [29] B. Andricic, T. Kovacic, I. Klaric, Polym. Degrad. Stab. 79 (2003) 265. [30] S.B. Hanna, L.M. Farag, N.A.L. Mansour, Thermochim. Acta 81 (1984) 77.