Preparation of hollow silica spheres with different mesostructures

Author's personal copy

Available online at www.sciencedirect.com

Journal of Non-Crystalline Solids 354 (2008) 826–830 www.elsevier.com/locate/jnoncrysol

Preparation of hollow silica spheres with different mesostructures Shiquan Liu

a,b,*

, Jiancun Rao c,d, Xueye Sui a, Pegie Cool b, Etienne F. Vansant b, Gustaaf Van Tendeloo c, Xin Cheng a

a

School of Materials Science and Engineering, University of Jinan, Jinan 250022, China Laboratory of Adsorption and Catalysis, University of Antwerp, Antwerp 2610, Belgium c EMAT, University of Antwerp, Groenenborgerlaan 171, Antwerp 2020, Belgium Institute of Advanced ceramics, Department of Materials Science, Harbin Institute of Technology (HIT),Harbin 150001, China b

d

Received 9 November 2006; received in revised form 2 August 2007 Available online 14 September 2007

Abstract Hollow silica spheres were quickly synthesized by an octylamine (OA) templating method using tetraethyl orthosilicate (TEOS) as the silica source. N2-sorption results indicate that the hollow spheres have high surface areas and pore volumes. XRD and TEM measurements reveal that the structure of the hollow spheres depends on the amount of TEOS used in the synthesis. When low amount of TEOS is added, the template-containing precursor spheres depict an XRD pattern with two peaks, which can be indexed to a lamellar phase. After the removal of the template, the obtained hollow spheres show no diffraction peaks in the XRD pattern, suggesting that the nanopores in the silica shells are disordered. If increasing the amount of TEOS, either the uncalcined or the calcined sample gives an XRD pattern with a single diffraction peak. The mesostructure of these hollow silica spheres is typically as HMS materials. TGA analyses suggest that the interaction between the silica species and surfactant is stronger in the latter case. Ó 2007 Elsevier B.V. All rights reserved. PACS: 81.20.Fw; 81.05.Rm Keywords: Microstructure; Porosity; Sol–gel, aerogel and solution chemistry

1. Introduction Hollow spheres may have potential applications in many engineering fields due to their special morphology and structure. Being spherical particles with hollow interiors, hollow spheres are light-weight compared with their solid counterparts. Therefore, they can be used as fillers in building materials for reducing the construction loading [1,2]. In some cases where only the surface properties of a material are utilized, producing it in the form of hollow particles can save the natural resources. Pores in the hollow shells can be a path for guest substances going inside or/and out of the * Corresponding author. Address: School of Materials Science and Engineering, University of Jinan, Jinan 250022, China. Tel./fax: +86 531 87974453. E-mail address: [email protected] (S. Liu).

0022-3093/$ - see front matter Ó 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jnoncrysol.2007.08.026

hollows, allowing the hollow spheres to be applied in fields of adsorption, separation, catalysis, micro-reactor, capsule, etc. [3–6]. It is also demonstrated that the hollow spheres may possess unique mechanical, acoustical, optical, electromagnetic and thermal properties [7–9], making the hollow spheres attractive as novel advanced functional materials. Except in a few reports where the hollow spheres were fabricated by physical methods [10], mostly chemical routes were applied for the preparation of the hollow spheres. Among the chemical methods, template-based syntheses were widely used [11,12]. The templates included organic or inorganic solid particles (spheres or irregular nanoparticles), emulsion droplets, vesicles, aggregates, and gas bubbles [12–16]. Shells were built around the templates by selfassembly or surface deposition. Removal of the solid templates by chemical or thermal treatments, or evaporation


S. Liu et al. / Journal of Non-Crystalline Solids 354 (2008) 826–830

and/or release of the soft or gaseous templates generated the hollow spheres. In a previous publication, we reported a fast fabrication of hollow silica spheres sized tens of micrometers by using an octylamine (OA) templating route [17]. The present work shows that the synthesized hollow silica shells may have different mesostructures, depending on the amount of TEOS used in the syntheses. The study offers an easy way to adjust the structure of the nanoporous hollow shells, so that optimal adsorption and diffusion effects might be achievable.

827

X-ray diffractograms were recorded on a Bruker D8ADVANCE powder diffractometer using a Ni-filtered Cu Ka radiation at steps of 0.02 2h degree. N2 sorption measurements were performed at 77 K on a Quantachrome Autosorb-1-MP automated gas adsorption system. The surface areas were calculated with the Brunauer– Emmett–Teller (BET) method, using the adsorption data within the range of relative pressure P/P0 from 0.05 to 0.15. The pore volumes were taken at P/P0 of 0.95. The pore size distributions were determined using the density function theory (DFT) with the silica adsorption branch kernel based on a cylindrical pore model [18].

2. Experimental 3. Results and discussion For the synthesis of the hollow spheres, TEOS and OA were first mixed for 3 min with a stirring rate of 600 rpm. Then, distilled water containing a small amount of hydrochloric acid (conc.) was quickly added. After further reaction of 5 min under stirring, the resulting precipitate was collected by a vacuum filtration and air-dried. The dried sample was then calcined to 550 °C for 6 h with a heating rate of 1 °C/min. Compositions of two representative examples, L and H, are listed in Table 1. Scanning electron microscopy (SEM) images were recorded on a JEOL JSM-5510 scanning electron microscope. The samples were sputtered with a thin film of gold. Transmission electron microscopy (TEM) measurements were performed on crushed samples using a Philips CM20 transmission electron microscope. Low electron beam intensity and low magnification were applied in order to preserve the sample structure and to avoid sample drifting. Thermogravimetric (TG) analyses were conducted on a Mettler TG-50 thermobalance connected to a TG 10A processor. The heating rate was 5 °Cmin, and an oxygen flow was supplied.

Table 1 Synthesis compositions of two representative samples Sample

TEOS/ml

Octylamine/ml

Water/ml

HCl concd./ml

L H

11.25 33.75

11.25 11.25

101.25 101.25

0.21 0.21

SEM measurements identified that hollow silica spheres could be prepared in a wide region of compositions, such as TEOS 11.25–33.75 ml, OA 3.75–11.25 ml, H2O 67.5– 135 ml, HCl 0.07–0.21 ml. However, the TEOS/OA ratio should not be smaller than 1 or larger than 4. Otherwise, spheres with craters on their surfaces or non-hollow spheres would be obtained, respectively. In the following sections, samples L and H are taken as representatives for discussions. Fig. 1(a) and (b) show the SEM images of samples L and H, respectively. It can be seen that both samples consist of micrometer-sized spheres. The broken particles (see the arrows) indicate that the spheres are hollow. However, the spheres in sample L are discrete, while the spheres in sample H are slightly aggregated and generally larger than those in sample L. The N2-sorption isotherms are presented in Fig. 2. Calculations show that sample L has a surface area of 1123 m2/g and a pore volume of 0.6291 ml/g. For sample H, the surface area and the pore volume increase to 1322 m2/g and 0.7465 ml/g, respectively. The H2 type hysteresis suggests that the hollow silica shells have interconnected networks of pores with different sizes and shapes [18]. The pore size distribution (PSD) curve (Fig. 2 inseta) of sample L, shows that the hollow silica shells have mixed micropores sized about 1.6 nm and broadly distributed mesopores centered at about 2.7 nm, respectively. In contrast, the PSD curve showed in Fig. 2 inset-b, indicates

Fig. 1. SEM images of the synthesized samples L (a) and H (b). (scale bar: 100 lm).


828


Fig. 2. Isotherms and DFT based pore size distribution curves (inset) of samples L (a) and H (b).

that sample H only has mesopores in its hollow silica shells. The size of the mesopores is the same as that in sample L, while the distribution of the mesopores is narrower. The fitting comparisons between the experimental and calculated DFT sorption curves give fitting errors of 0.342% and 0.398% for samples L and H, respectively, indicative of a good accuracy of the estimation of the pore size by the DFT model [19]. X-ray diffraction analyses were performed both on the calcined and uncalcined samples. As can be seen from Fig. 3, samples L and H show two different types of XRD patterns. Two diffraction peaks at 2.88 and 5.8 2h degrees are present in the XRD pattern of the uncalcined sample L (Fig. 3(a)). They can be indexed to a lamellar phase. Lamellar mesophases have been reported for amine templated silicas elsewhere [20,21]. After the sample is calcined, it exhibits an XRD pattern with no diffraction peak (Fig. 3(b)), indicating that the nanopores, revealed by the N2 adsorption results, are randomly distributed in the hollow silica shells. In contrast, uncalcined sample H exhibits an XRD pattern with a single peak at 2.8 2h degree (Fig. 3(c)). A similar XRD pattern is recorded for the calcined sample H, except that the peak shifts to a 2h position

of 3.16 degree. The shift of the diffraction peak is due to the further condensation and contraction of the silica skeletons during the calcination. These results for sample H are typical for HMS-type materials [22–24]. A wide investigation demonstrated that the XRD patterns of the hollow spheres synthesized with the reactant ratio falling in the region mentioned at the beginning of section 3 were only dependent on the TEOS/OA ratio and irrespective of the amounts of water or acid and the reaction time [17]. It was concluded that when the ratio of TEOS to OA was 1 or 2, the uncalcined samples, similar as the uncalcined sample L, showed XRD grams with two peaks, which were not observed in the XRD pattern for the correspondent calcined samples. However, when this ratio was larger than 2, XRD patterns with a single diffraction peak were recorded for both calcined and uncalcined samples as in the case of sample H. TEM observations further prove the structural differences revealed by the XRD results. As visible from Fig. 4(a), the uncalcined sample L depicts a lamellar structure (see the arrow pointed area). In the meantime, one can also see the hybrid segments (see the circled area) with mixed two phases, indicated by the black and white

Fig. 3. XRD patterns of uncalcined sample L (a), calcined sample L (b)uncalcined sample H (c) and calcined sample H (d).



829

Fig. 4. TEM images of uncalcined sample L (a); calcined sample L (b) and calcined sample H (c).

contrast. These segments might be the vertical view of the template-silica laminas. After the sample is calcined, the lamellar structure is fully disturbed due to the removal of the organic template. However, the silica laminas with nanopores are remained but distorted and curved (Fig. 4(b)). The results explain that although no XRD diffraction peaks are observed for calcined sample L, the hollow spheres still possesses high surface area and pore volume. The micropores may be originated from the voids between the collapsed laminas. In contrast, Fig. 4(c) shows that hollow silica spheres in sample H have a typical worm like structure as the neutral amine templated HMS-type materials [24]. (No big difference is expectable in the TEM images of uncalcined and calcined sample H, although after the calcination, the positions of the template are replaced by the nanopores.) DTG curves from the TGA measurements (Fig. 5) show that the decomposition of the template in sample H occurs at a significantly higher temperature than that in sample L, suggesting a much stronger interaction force between the silica species and the template in sample H. The difference might be explained as follows. Since more TEOS is used in the synthesis of sample H, silica acid generated by the hydrolysis of TEOS may not be well condensed in sample H due to the excess of TEOS. This results in a weak interaction between the silica species, and on the other way round, leads to a strong interaction between the silica species and the surfactant. Similar phenomenon was also observed in an assembly of titania and template [25]. Based on the presented results and discussion, we postulate a mechanism for the formation of the hollow spheres. When the mixture of TEOS and octylamine is under stir-

Fig. 5. DTG curves of uncalcined samples L (a) and H (b).

ring, due to the differences in the viscosity and the surface tension of these two components [26,27], giant octylamine vesicles are formed at the air/mixture interface [28,29], The vesicles may be multi-lamellar or unilamellar, depending on the TEOS/OA ratio. After the addition of the acidified water, a fast hydrolysis and condensation of TEOS take place. The reaction products quickly fix the morphology of the vesicles [30,31]. The enclosed cavities of these vesicles [26] would be the template for the hollows inside the hollow spheres. The subsequent removal of the organic by calcination leads to the formation of the hollow silica spheres with nanoporous shells. 4. Conclusion By changing the amount of TEOS used in the syntheses, hollow silica spheres with different mesostructures are


830


quickly synthesized by an octylamine templating route. N2sorption measurements indicate that the obtained hollow spheres have high surface areas and pore volumes. Both XRD and TEM results show that the amount of TEOS influences the mesostructure of the precursor hollow spheres. Fully disordered mixed micropores and mesopores and HMS-like mesopores are formed when the TEOS is relatively low and high, respectively. Acknowledgments This work is partly supported by the NOE project ‘Inside Pores’ and the GOA-project of the University of Antwerp. Dr S. Liu is also indebted to the PhD research foundation (B0309) in Jinan University. The authors also thank Dr V. Meynen and Dr M. Wu for their useful help and discussions. References [1] [2] [3] [4] [5] [6] [7] [8] [9]

T. Sugama, N. Carciello, Adv. Cement Based Mater. 3 (1996) 45. T. Sugama, B. Lipford, J. Mater. Sci. 32 (1997) 3523. T. Miyao, K. Minoshima, S. Naito, J. Mater. Chem. 15 (2005) 2268. H.J. Hah, J.I. Um, S.H. Han, S.M. Koo, Chem. Commun. (2004) 1012. J. Chen, H. Ding, J. Wang, L. Shao, Biomaterials 25 (2004) 723. E. Bae, S. Chah, J. Yi, J. Colloid Interf. 230 (2000) 367. H.L. Dai, X. Wang, Eur. J. Mech. A-Solid 24 (2005) 263. M. Han, W. Zhang, C. Gao, Y. Liang, Z. Xu, J. Zhu, J. He, Carbon 44 (2006) 211. P. Tartaj, T. Gonza´lez-Carrenõ, C.J. Serna, Adv. Mater. 13 (2001) 1620.

[10] Y. B. Kim, K.-S. Yoon, Macromol. Rapid Commun (2004) 25, 1643. [11] F. Caruso, R.A. Caruso, H. Mohwald, Science 282 (1998) 1111. [12] S. Schacht, Q. Huo, I.G. Voigt-Martin, G.D. Stucky, F. Schuth, Science 273 (1996) 768. [13] Y. Le, J. Chen, J. Wang, L. Shao, W. Wang, Mater. Lett. 58 (2004) 2105. [14] H. Hentze, S.R. Raghavan, C.A. McKelvey, E.W. Kaler, Langmuir 19 (2003) 1069. [15] J. Wang, Y. Xia, W. Wang, R. Mokaya, M. Poliakoff, Chem. Commun. (2005) 210. [16] P.S. Singh, K. Kosuge, Chem. Lett. (1998) 101. [17] X. Cheng, S. Liu, L. Lu, X. Sui, V. Meynen, P. Cool, E.F. Vansant, J. Jiang, Micro. Meso. Mater. 98 (2006) 41. [18] F. Rouquerol, J. Rouquerol, K. Sing, Adsorption by powders and porous solids: principles, methodology and applications, Academic press, San diego, 1999. [19] A.V. Neimark, P.I. Ravikovitch, Micro. Meso. Mater. 44-45 (2001) 697. [20] G. Cheng, C. Liu, Mater. Chem. Phys. 77 (2002) 359. [21] J. Li, Y. Xu, D. Wu, Y. Sun, Solid State Commun. 130 (2004) 619. [22] P.T. Tanev, T.J. Pinnavaia, Science 267 (1995) 865. [23] A.G.S. Prado, C. Airoldi, J. Mater. Chem. 12 (2002) 3823. [24] K. Cassiers, P. Van Der Voort, T. Linssen, E.F. Vansant, O. Lebedev, J. Van Landuyt, J. Phys. Chem. B 107 (2003) 3161. [25] E. Beyers, P. Cool, E.F. Vansant, J. Phys. Chem. B 109 (2005) 10081. [26] M. Summers, J. Eastoe, Adv. Colloid Interf. 100-102 (2003) 137. [27] M. Dubois, T. Gulik-Krzywicki, B. Cabane, Langmuir 9 (1993) 673. [28] Y.F. Zhou, D.Y. Yan, Ang. Chem-inter. Ed. 43 (2004) 896. [29] S.J. Ryhanen, V.M.J. Sally, M.J. Parry, P. Luciani, G. Mancini, J.M.I. Alakoskela, P.K.J. Kinnunen, J. Am. Chem. Soc. 128 (2006) 8659. [30] R.K. Iler, The Chemistry of Silica, John Wiley and Sons, Chichester, UK, 1979. [31] J. Du, Y. Chen, Macromolecules 37 (2004) 5710.