Powder Technology 196 (2009) 286–291
Contents lists available at ScienceDirect
Powder Technology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p ow t e c
Titania–silica composites with less aggregated particles Askwar Hilonga a, Jong-Kil Kim b, Pradip B. Sarawade a, Hee Taik Kim a,⁎ a b
Department of Chemical Engineering, Hanyang University, 1271 Sa 3-dong, Sangnok-gu, Ansan-si, Gyeonggi-do 426-791, Republic of Korea E&B Nanotech. Co., Ltd, Republic of Korea
a r t i c l e
i n f o
Article history: Received 11 March 2009 Received in revised form 31 May 2009 Accepted 7 August 2009 Available online 15 August 2009 Keywords: TiOCl2 Sodium silicate Titania–silica Less aggregated BET
a b s t r a c t Less aggregated titania–silica composite was developed by a versatile and reproducible method using relatively cheap precursors. The final product has more suitable properties than the conventional materials. The composite was synthesized by using sodium silicate, as a silica precursor, and freshly prepared TiOCl2 as a titania source. The final product was obtained after subsequent calcination for 5 h at 300 to 1000 °C. The primary particles of the composite, as examined by SEM technique, are generally less aggregated. The XRD patterns for the calcined samples indicate the presence of TiO2 and there is a significant increase of peak intensity as the calcination temperature increases. EDS and XPS analyses confirmed the formation of pure composite rich in Ti, Si, and O. Nitrogen physisorption studies reveal that the composite is mesoporous and have large BET surface area (~375 m2/g). A simple experiment of photoreduction of methyl orange under solar radiation was attempted to demonstrate the reliability and improvement of titania–silica composite in practice. It was found out that its efficiency is high as compared to P-25 TiO2 under solar light. The results demonstrate that composite with desirable properties for various applications can be obtained via the present route. © 2009 Elsevier B.V. All rights reserved.
1. Introduction Titania–silica composite has been currently reported to be a good source of paper filler due to its high degree of whiteness and opacity [1]. The composite simultaneously utilizes the best properties of TiO2 and SiO2, namely, enhanced opacity and oil absorption, respectively. Several publications also report the use of titania–silica composite for catalysis and as support for several chemical reactions [2–4]. Also it has got potential applications to new devices such as non-linear optics [5]. Generally, compared with pure TiO2, titania–silica composite has a high thermal stability, improved mechanical strength, and good dispersibility in resin and solvents. Titania–silica composite can express its effectiveness when its particles are less aggregated. For instance, in the latter form the unacceptable whiteness can be prevented in cosmetics due to the increase in transparency [6]. Resin films, coating materials, catalysis and similar applications also require the material with less aggregated particles. When the primary particles become finer, the aggregation degree tends to be higher. The particles having such a high aggregation degree are not only inferior in transparency but also in the handling of the powder since a solvent is readily included in spaces between primary particles or in voids of the steric structure. In this case, the material will encounter inevitable setback in its practical application. Notwithstanding the need observed, there is limited research-based information on the versatile method of synthesizing
⁎ Corresponding author. Tel.: +82 31 400 5493; fax: +82 31 500 3579. E-mail address:
[email protected] (H.T. Kim). 0032-5910/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.powtec.2009.08.004
titania–silica composite with less aggregated particles. Moreover, in order to increase the efficiency and effectiveness in various applications, the ideal composition of a suitable titania–silica composite will be the one in which titania is a main component [6]. It should have a large BET surface area and should not grow to have a steric structure due to the connected primary particles. In the current study an attempt is made to develop such materials. Review of the current literature disclosed an intensive work undertaken to develop titania–silica composite with some suitable properties [7–10]; but most often the issue of particles aggregation was overlooked. Much of the work done was about the internal properties of the composite nanoparticles. In many cases, particles appeared in aggregated form. Moreover, precursors that are traditionally used for the synthesis of titania–silica composites have some limitations. Expensive silica precursor, Tetraethoxysilane (TEOS), has been frequently used [11–13]; while TiCl4 has been avoided as the titania source because of its tendency to polymerize uncontrollably and form large titania agglomerates rapidly [14]. This paper describes the use of TiCl4 via the preparation of TiOCl2 as described in the Experimental section. This versatile method combined with the control of gelation of silicic acid, derived from relatively cheap silica precursor (sodium silicate), has finally resulted into the materials with desired properties, namely, large surface area, high thermal stability, and more titania content with less aggregated particles. To the best of our knowledge, there are rare researches done on the use of sodium silicate as a silica precursor for the synthesis of titania–silica composite with desired properties (particularly less aggregated particles).
A. Hilonga et al. / Powder Technology 196 (2009) 286–291
B. Knoblich and Th. Gerber have made a valuable contribution on understanding the aggregation process in SiO2 sols from sodium silicate solutions [15]. Basically they have examined the fractal dimension of clusters, gelation time and the isoelectric point of SiO2 in aqueous solution. On the other hand, the influence of the aging time on TiO2 nanoparticles formed in silica sol-gel matrix was also reported [16]. The possibility of replacing the relatively expensive silica precursors that are traditionally used for the synthesis of titania–silica composite with relatively cheap ones (sodium silicate) may boost applications of the composite, particularly at the industrial scale. In a continuous work on development of titania–silica composites with various properties, we have synthesized the ones with less aggregated particles. In this work, results on the formation of titania–silica composite by using freshly prepared TiOCl2 solution as a TiO2 precursor and silicic acid from sodium silicate as a silica precursor is presented. The influence of calcination temperature on the properties of the final product was widely examined. Moreover, a simple experiment of photoreduction of methyl orange under solar radiation was attempted to demonstrate the reliability and improvement of titania–silica composite in practice. In this case, the catalytic efficiency of the developed composite was compared with P-25 TiO2 under solar light. Our results will provide a basis for further developments of a cheap route to titania–silica composite with less aggregated particles for various applications. 2. Experimental 2.1. Preparation of TiOCl2 stock solution The preparation of TiOCl2 aqueous solution has been described elsewhere [1] and here we adapt it with some modifications as follows: 16.5 ml of reagent grade TiCl4 (from Duksan Chemical, South Korea) was
287
dropped slowly in the 3-neck flask (inserted in ice bath) containing 33.5 ml of distilled water. Then 3 drops of 1 M HCl (from Duksan Chemical, South Korea) was added to prevent the hydrolysis and precipitation. The obtained product was labeled as the TiOCl2 stock solution. Since TiOCl2 was freshly prepared and used immediately, all Ti atoms were in monomeric form. 2.2. Reaction of TiOCl2 with silicic acid In general, the formation of silicic acid in aqueous solution results from an ion exchange of Na+ for H+ in sodium silicate solution by using an ion exchange resin. The removal of Na+ will result into the formation of silicic acid that has got a low pH value, ca. 2. In order to induce gelation, the pH can be slowly increased by adding few drops of ammonia solution, as a catalyst. It is known that at smaller or larger pH values the gel formation proceeds quickly and the gelation time reaches a minimum at nearly pH = 6 [17]. Thus, the slow increase in pH (up to pH 6) of the solution allows uniform formation of particles and the aggregation can be avoided. In this study, sodium silicate solution (24% SiO2, 7.4% Na2O — Shinwoo Materials Co. Ltd., South Korea) was diluted to 5 wt.% SiO2 solution with distilled water. Then silicic acid solution for gelling was produced by replacing Na+ with H+ by using the ion exchange resin (IR-120, pH 2.4 — Rhom & Hass, Germany). 4 g of silicic acid so obtained was put in a beaker inserted in water bath at 65 °C. Then 10 ml of freshly prepared TiOCl2 was added slowly and stirred for 5 min before adding 50 ml of water. The reaction was continued for 2 h before adding 10 ml of ammonia solution (28% NH3 — Shinwoo Materials Co. Ltd., South Korea) for catalyzing the hydrolysis of silicic acid. It was left to mature for 2 h. The solids were filtered and washed with water and dried at 80 °C for 2 h. The final product was calcined at 300 to 1000 °C for 5 h. A summary of the synthesis procedure is provided in Scheme 1.
Scheme 1. A flow chart of the synthesis procedure for titania–silica composite (TSC). The numbers/description affixed with the TSC acronym simply show the reaction stages.
288
A. Hilonga et al. / Powder Technology 196 (2009) 286–291
It should be noted that, one of the main advantages of the use of TEOS is that all byproducts are volatile, usually they are alcohols. But since the precursor that is employed in the present study may leave residual (such as Na+ and Cl−) behind in the structure and consequently alter the properties of the material, the as-synthesized materials were first thoroughly washed with water and then it was rinsed with ethanol. By following this procedure all byproducts were successfully removed, as examined by using pH/ISEK Meter pH-250L, Korea. The purity of the final product was also confirmed by EDS and XPS analyses, whereby the elements present in the composite were Ti, Si and O only. 2.3. Photocatalytic activities The photoactivity of the titania–silica composite and P-25 TiO2 was measured by degrading methyl orange (C14H14N3NaO3S) in water solution. In this study, we simply conducted the photocatalytic experiments in the presence of natural sunlight (since the main target was just to compare the performance of the developed composite with the standard catalyst — P-25 TiO2). Natural sunlight has been reported (proved) to be a feasible source of illumination to perform different photocatalytic reactions [18 and references therein]. Thus, in our case, the photocatalytic reactions were carried out in a glass reactor of 250 ml under stirring for 2 sunny days. The same weight (1.0 g/l) of the composite and P-25 TiO2 was used for an easy comparison. The initial concentration of methyl orange solution was 10.0 mg/l. After 2 days, 10 ml of dispersion was centrifuged to separate the photocatalyst. The supernatant solution was analyzed by a UV–vis spectrophotometer. The observation related to the photoreduction of methyl orange is briefly presented in Section 3.5 of this report.
Fig. 1. SEM micrograph showing titania–silica composite with less aggregated particles. Insets show EDS spectra and SEM micrograph at high magnification (200 nm).
2.4. Characterization The morphology of the titania–silica composite samples was characterized by scanning electron microscopy (FE-SEM, Hitachi, S-4800) with an accelerating voltage of 15 kV. The FE-SEM was coupled with energy dispersive spectroscopy (EDS) to assess the purity and elemental composition of the final product. X-ray photoelectron spectroscopy (XPS) spectra were recorded by a Quantum 2000 XPS system (Physical Electronics, Inc.). The vacuum in the analytical chamber during measurements was better than 5 × 10− 9 Torr. X-ray diffraction patterns (XRD-6000, Shimadzu) were used to determine the crystallinity of the composite. The accelerating voltage and applied current were 40 kV and 100 mA, respectively. The crystallite size of the powders was determined by Scherrer equation [19]. The Brunauer–Emmett–Teller (BET) surface area and the porosity of the samples were studied by a nitrogen adsorption instrument (Micrometrics ASAP 2020). All the samples measured were degassed at 250 °C for 1 h prior to actual measurements. Thermogravimetric analysis (TGA) and Differential TGA (DTGA) were performed using a microprocessor-based Parr temperature controller (Model 4846) connected to a muffle furnace (A.H.JEON Industrial Co. Ltd. Korea) at a heating rate of 10 °C/min from room temperature to 1000 °C. The catalysis efficiency of the samples was analyzed by a UV–vis spectrometer (Perkin-Elmer Lamda 800).
Fig. 2. SEM micrograph showing titania–silica composite at Ti/Si = 4.9.
regular and smooth with the increase of calcination temperature. In our similar study in which sodium silicate was directly used (without preliminary step of producing silicic acid), the particles of the developed materials tend to aggregate into large spheres as pointed by arrows in Fig. 4. Since the formation of the silica matrix is influenced by the hydrolysis and condensation (to provide charged silica oligomers able to interact with Na+ ions and to avoid large particle growth) processes, there must be a role of
3. Results and discussion 3.1. FE-SEM, EDS, and XPS The FE-SEM micrographs for the representative samples are shown in Figs. 1–4; also, a high magnification FE-SEM micrograph and EDS spectra are shown as insets in Fig. 1. The primary particles of the calcined samples appear to be spherical in shape and closely packed together; yet they are less aggregated. They become more
Fig. 3. SEM micrograph showing titania–silica composite at Ti/Si = 6.7.
A. Hilonga et al. / Powder Technology 196 (2009) 286–291
289
should be regarded as complementary to those proposed in the previous studies [14,15,20]. The EDS result depict no other peaks except the ones expected for titania–silica composite (Ti, Si and O) indicating the high-purity of the composite obtained through this method. The ratio of Ti/Si ranges from 4.8 to 6.7 (as summarized in Table 1) confirming the formation of titania–silica composite in which the TiO2 is a main component. Furthermore, the XPS studies were performed in order to provide more evidence to substantiate the purity and the existence of Ti–O–Si bonds in titania–silica composite. The results significantly reveal the strong interaction between SiO2 and TiO2 (Fig. 5a and b). The binding energy of Ti2p3/2 appeared at 458.8 eV; while that of O1s appeared at 533 eV (Si–O–Si) and 532.2 eV (Si–O–Ti). This result is consistent with a recent report by Ren et al. [21]. 3.2. X-Ray diffraction Fig. 4. SEM micrograph showing titania–silica composite with primary particles forming large spherical aggregated. Arrows show spherical aggregates of particles.
Table 1 EDS results for the elemental composition of the titania–silica composite calcined at representative temperatures (300, 600, and 1000 °C). Atomic %
TS (As-synthesized)
TS (300 °C)
TS (600 °C)
TS (1000 °C)
O Si Ti Ratio (Ti/Si)
68.23 4.48 27.29 6.1
73.14 4.53 22.32 4.9
70.53 3.81 25.66 6.7
78.35 3.76 17.89 4.8
TS = titania–silica composite.
Na+ ions (when sodium silicate is used directly as silica source). Hydrolysis of sodium silicate by 2N HCl is pointed out in the Experimental section. The sodium ions that are generated during the hydrolysis of sodium silicate act as flocculant agent during the condensation of Si–O–Si. Thus, coagulation that favors the condensation among silica nanoparticles is induced by Na+ ions, consequently producing aggregated material as depicted in Fig. 4. On the other hand, the less aggregation behavior observed in the current study can be ascribed to the controllable hydrolysis and polymerization of silicic acid in the absence of Na+ ions. When the pH of the solution is slowly increased during the hydrolysis and condensation of silica and the reaction of silica with titania, the aggregates with less ability of forming conglomerates are formed. The mechanism suggested in the present report
Fig. 6 shows the XRD patterns of the composite together with the P-25 TiO2. The appearance of anatase and rutile peaks confirms the presence of titania particles in the silica matrix. The XRD patterns show the peaks assigned to anatase phase titania at 25.5°, 38.0°, 48.2°, 54.5° and 62.8°, while those associated with the formation of rutile phase were observed at 27.5, 36.0°, 44.1°, and 56.7°. This observation is compatible with the previous report on similar experiment based on a sol-gel process by using titanium alkoxide as titania precursor [22]. The average crystallite size D is calculated using the Debye– Scherrer formula D = Kλ/(βcosθ), where K is Scherrer constant, 0.89; λ the X-ray wavelength, 1.5406 Å; β is the full width at halfmaximum (FWHM) of the diffraction peaks, rad; and θ is the Bragg diffraction angle [19]; the summary of the results obtained is provided in Table 2. The grain size of the rutile TiO2 in the titania-silica composite is smaller (5 nm) than that of pure TiO2 (12.4 nm). Generally, the crystallite sizes increase with the increase in calcination temperatures and the pure TiO2 was completely transformed to rutile at 700 °C. Thus, the phase transformation temperature of composite is higher than that of pure titania particles. This is ascribed to the stabilization of the TiO2 by the surrounding silica matrix through the Ti–O–Si interface. It was previously report that, at the interface, the SiO2 lattice locks the Ti–O species at the interface of the TiO2 domains, preventing the nucleation that is necessary for the phase transformation to rutile. Hence, greater heat is required to drive the crystallization [23]. Therefore, even though silica is a subsidiary component in the present study, still it has effectively prevented the
Fig. 5. The XPS spectra of titania-silica composite: (a) (Ti2p), and (b) peaks showing the O 1 s lines of Si–O–Si (533 eV) and Si–O–Ti (532.2 eV).
290
A. Hilonga et al. / Powder Technology 196 (2009) 286–291
Fig. 6. XRD patterns for the representative samples of titania–silica composite calcined at 600 °C (TS — 600) and 1000 °C (TS — 1000) together with the pure TiO2 calcined at 700 °C.
Table 2 Important XRD peaks observed in titania-silica (TS) composite calcined at some representative temperatures compared to the pure TiO2. 2 Theta degree 25.5° 27.5° 36.0° 38.0° 39.0° 41.4° 44.1° 48.2° 54.5° 56.7° 62.8°
TS (1000 °C)
Pure TiO2 (700 °C)
FWHM
TS (600 °C) XS (nm)
FWHM
XS (nm)
FWHM
XS (nm)
2.08 – – 1.66 – – – 1.66 2.19 – 1.59
3 – – 5 – – – 5 4 – 5
0.77 1.59 – 1.19 – – – 1.44 1.78 1.66 1.66
10.1 5 – 7 – – – 6 5 5 5
– 0.67 0.63 – 1.94 0.64 1.89 – 0.80 0.44 0.53
– 12.4 13.4 – 4 13.4 4 – 11.3 21.5 17.6
complete transformation of anatase to rutile phase even at high calcination temperature. 3.3. Nitrogen physisorption studies The comparison for the nitrogen adsorption–desorption isotherms of the composite is presented in Fig. 7. Generally, the isotherms are similar and show typical shape for mesoporous materials. In all cases the isotherms display plateaus at P/Po < 0.4 and the adsorption step corresponding to nitrogen condensation in primary mesopores is shifted toward higher relative pressures. This indicates the occurrence of larger pores in the prepared composites. Thus, it may be suitable for various applications. The composite displays surface areas ranging from 276 to 375 m2/g. The BET surface area decreases with the increase in calcination temperature. This is due to the phase transformation of titania which lead to the grain growth, as observed and discussed in XRD analysis. It also indicates a structural collapse for the sample calcined at higher temperatures. Nevertheless, the BET surface areas obtained in this study are higher than the ones reported in literature reviewed [24,25]. Generally, all samples have got similar porosity. The pore size distribution of titania–silica composite calculated by the BJH method is shown as inset in Fig. 7. It can be deduced that all the samples presented have a similar pattern of pore size distribution and average pore diameter ranging from 5.0 to 6.1 nm. Thus calcination temperature up to 1000 °C has less effect on our present composite as also substantiated by TGA/DTGA. This advocates its potential superiority even at wider range of tempera-
Fig. 7. The adsorption–desorption isotherms for the representative samples of titania– silica composite calcined at 300 °C (TS — 300), 600 °C (TS — 600) and 1000 °C (TS — 1000). Inset: The pore size distribution of titania–silica composite calculated by the BJH method.
tures in practical applications. Moreover, the narrow pore size distribution observed indicates the absence of inter-particle voids confirming the existence of particles in less aggregated state.
3.4. TGA and DTGA analyses Thermal stability of the titania–silica composite was examined by the TGA and Differential TGA (DTGA) analyses, as presented in Fig. 8. A distinct degradation step can be clearly observed at 240 °C. A peak that can be identified in DTGA curve (inset) corresponds to the weight loss stage as demonstrated in TGA curve. The weight loss at 240 °C is associated with the loss of physisorbed water and organic solvents. It was previously reported that a sharp weight loss at the temperature lower than 270 °C is caused by the removal of physically absorbed water in the composites [26]. Moreover, a small weight loss at the range of 370 to 453 °C might be attributed to the decomposition of organic groups [27] and the development of TiO2 anatase phase. It also indicates that the OH groups on the silica network are condensing to form Si–O–Si bridges; resulting into the weight loss. There is no significant change in weight beyond 453 °C implying that the silica structure is fully developed and the TiO2 is transformed to the anatase phase. This observation concurs with the explanation on the
Fig. 8. TGA and DTGA (inset) curves for the titania–silica composite.
A. Hilonga et al. / Powder Technology 196 (2009) 286–291
291
studies reveal the formation of a nanoporous material with a large BET surface area even after calcination at higher temperatures. A simple experiment of photoreduction of methyl orange under solar radiation was attempted to demonstrate the reliability and improvement of titania–silica composite in practice. It was found out that its efficiency is high as compared to P-25 TiO2 under solar light. Our results will provide a basis for further developments of a versatile and less expensive route to synthesize the titania–silica composite with less aggregated particles for various applications.
Acknowledgements
Fig. 9. Absorbance spectra of un-degraded methyl orange (MO) in relation to the ones degraded by P-25 TiO2 and titania–silica (TiO2–TiO2) composite.
We express our gratitude to the Ministry of Commerce and Industries of the Republic of Korea for the financial support under the R & D innovation fund for “The Small and Medium Business Administration”. Project Application No. S1017370.
appearance of the XRD diffraction peaks, signifying the existence of titania phases.
References
3.5. Photocatalysis experiment result Fig. 9 shows the evolution of methyl orange absorption spectra in the absence of catalyst (un-degraded MO) titania–silica composite, and pure titania (P-25). The change in the concentration of methyl orange was determined at the range of 300 to 700 nm in the UV–vis spectra. It is clearly evident that titania–silica composite exhibited superior photocatalysis compared to the pure titania — methyl orange was highly degraded. 5. Conclusion In this study, a significant work has been undertaken to develop titania–silica composite with less aggregated particles by using a versatile and reproducible method; while utilizing relatively cheap silica precursor. The composite was synthesized by using sodium silicate, as a silica precursor, and freshly prepared TiOCl2 as a titania source. The final product has more suitable properties than the conventional materials. The EDS, XRD, and SEM investigations reveal a pure composite wherein the TiO2 is a main component; yet the particles of the composite do not aggregate significantly. The composite has high thermal stability, as revealed by TGA/DTGA, and the titania phase transformation was delayed (until 1000 °C) compared to that of pure TiO2 (below 700 °C). Nitrogen physisorption
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27]
J.K. Park, J.K. Kim, H.K. Kim, J. Mater. Proc. Technol. 186 (2007) 367. T. Kasuga, M. Hiramatsu, M. Hirano, A. Hoson, J. Mater. Res. 12 (1997) 607. H. Mori, Y. Miyamura, T. Endo, Mater. Chem. Phys. 115 (2009) 287–295. X. Gao, I.E. Wachs, Catal. Today 51 (1999) 233. P. Sen, J.T. Andrews, Solid State Commun. 120 (2001) 195. [6] S. David, B. Edward, O. Carl, US patent:WO/2008/067186. S. Ren, L. Zhao, M. Yuan, Y. Yu, Y. Guo, Z. Wang, J. Solid State Chem. 182 (2009) 312–316. E. Obuchi, T. Sakamoto, K. Nakano, Chem. Eng. Sci. 54 (1999) 1525. Y. Li, S.J. Kim, J. Phys. Chem. B 109 (25) (2005) 12309–12315. C.C. Wang, J.Y. Ying, Chem Mater. 11 (1999) 3113. C.J. Brinker, G.W. Scherer, Test article sample title placed here, Press, New York, 1989. Z. Ding, X. Hu, G.Q. Lu, P.L. Yue, P.F. Greenfield, Langmuir 16 (2000) 6216. Z.J. Li, B. Hou, Y. Xu, D. Wu, Y.H. Sun, J. Colloid Interface Sci. 288 (2005) 149. B.B.J. Aronson, C. Blanford, A. Stein, Chem. Mater. 9 (1997) 2842. B. Knoblich, Th. Gerber, J. Non-Cryst. Solids 283 (2001) 109. S.I. Seok, J.H. Kim, Mater. Chem. Phys. 86 (2004) 176. R.K. Iler, The Chemistry of Silica, Wiley, New York, 1979. R. Chatti, S.S. Rayalu, N. Dubey, N. Labhsetwar, S. Devotta, Sol. Energy Mater. Sol. Cells 91 (2007) 180–190. L. Miao, P. Jin, K. Kaneko, A. Terai, N.N. Gabain, S. Tanemura, Appl. Surf. Sci. 212 (2003) 255–263. C. Vautier-Giongo, H.O. Pastore, J. Colloid Interface Sci. 299 (2006) 874–882. S. Ren, X. Zhao, L. Zhao, M. Yuan, Y. Yu, Y. Guo, Z. Wang, J. Solid State Chem. 182 (2009) 312–316. P. Tarazona, U.M.B. Marconi, R. Evans, Mol. Phys. 60 (1987) 543. L. Zhao, Jiaguo Yu, Bei Chen, J. Solid State Chem. 178 (2005) 1818–1824. K.M.S. Khalil, M.I. Zaki, Powder Technol. 92 (1997) 233. K.M.S. Khalil, M.I. Zaki, Powder Technol. 120 (2001) 256. C.J. Brinker, G.W. Scherer, E.P. Roth, J. Non-Cryst. Solids 72 (1985) 345. J. Liu, T. Xu, X. Han, Y. Fu, Eur. Polym. J. 42 (2006) 2755.
