- Email: [email protected]
Octadecylsilane hybrid silicas prepared by the sol–gel method: Morphological and textural aspects
Journal of Colloid and Interface Science 312 (2007) 326–332 www.elsevier.com/locate/jcis
Octadecylsilane hybrid silicas prepared by the sol–gel method: Morphological and textural aspects Rodrigo Brambilla a , Gilvan P. Pires a , João H.Z. dos Santos a,∗ , Márcia S. Lacerda Miranda b a Instituto de Química, UFRGS, Av. Bento Gonçalves, 9500, Porto Alegre, 91501-970 Brazil b Braskem S.A., III Pólo Petroquímico, Via Oeste, Lote 05, Triunfo, 95853-000 Brazil
Received 22 December 2006; accepted 2 March 2007 Available online 26 April 2007
Abstract A series of octadecylsilane-modified silicas was prepared by the sol–gel method through the hydrolysis and cocondensation of tetraethylorthosilicate (TEOS) with octadecyltriethoxysilane (ODS). The ODS:TEOS ratio was varied between 0:100 and 100:0. The resulting carbon content was between 2.5 and 53.4%. In the case of pure ODS, the resulting silica presented 68.6% of C. Hybrid silicas were characterized by scanning electron microscopy, energy dispersive X-ray spectroscopy, and 29 Si nuclear magnetic resonance spectroscopy. Spheres of ca. 0.5–1.0 µm were obtained in the case of hydrolysis of pure TEOS. The combination of ODS:TEOS ratio yielded systems combining spherical and lamellar patterns zones. Monitoring the particle growth, it seems that spherical particles grow around lamellar zones, these latter concentrating the organosilicon moieties. The degree of cross-linking of ODS moieties was shown to be dependent on the ODS addition time and stirring speed. © 2007 Elsevier Inc. All rights reserved. Keywords: Hybrid silica; Sol–gel; Octadecylsilane; ODS
1. Introduction One of the major advances of sol–gel processing is the possibility of synthesizing hybrid inorganic–organic materials. The so-called hybrid materials combine, to some extent, the properties of inorganic and organic compound in one material [1]. Recent research on hybrid materials comprises coating on carbon steel [2], synthesis of supported catalysts [3] and immobilized enzymes [4], chromatographic stationary phases [5], membrane materials [6], and nanocomposites [7], to note a few. Hybrid silicas produced by the sol–gel method can be obtained from the hydrolysis and cocondensation of tetraethylorthosilicate (TEOS) with other organosiloxanes (Rx Si(OR)4−x , where R is an alkyl groups). In these materials, TEOS functions as building blocks to construct the framework, while the organosiloxanes, with nonhydrolyzable organic groups, contribute both to the framework silica units and to the organic * Corresponding author. Fax: +55 51 3316 7304.
E-mail address: [email protected] (J.H.Z. dos Santos). 0021-9797/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2007.03.003
surface functional groups. A large number of silanes carrying different functional groups have been used for silica modification [8]. Among them, octadecyltrimethoxysilane (ODS) and octadecyltrichlorosilane (OTS) have been largely used for surface chemical modification. Examples of application of such reactions can be found in the development of self-assembled monolayers [9], of chromatographic phases [10], of hydrophobic coatings [11], and of adsorbents [12]. If multicomponent systems are prepared by hydrolysis and condensation of alkoxide precursors, the reaction rates of hydrolysis and condensation (including aggregation) become very important for the distribution of the different components. Besides, it might influence the resulting textural and chemical properties of the final product [13]. In a previous study, we investigated the effect of producing ODS-modified silica by grafting and by the sol–gel method [14]. In such materials, the conformation of the octadecyl chains, as well as the surface coverage, depended on the preparative route. One application of hybrid materials is as catalyst support [15,16], and in catalysis, morphology can be an important issue. For instance, some polymerization processes require the repli-
R. Brambilla et al. / Journal of Colloid and Interface Science 312 (2007) 326–332
cation of the catalyst (and consequently of the catalyst support) spherical morphology. In the present paper, we evaluated the effects of the TEOS/ODS ratio, the addition time, and the stirring speed on the surface coverage, and on the morphological and the textural characteristics of the resulting hybrid materials. 2. Experimental 2.1. Materials Octadecyltrimethoxysilane (Acros, 90%), octyltrimethoxysilane (Aldrich, 90%), methyltrichlorosilane (Aldrich, 90%), and tetraethylorthosilicate (TEOS) (Merck, >98%) were used without further purification. Ethanol (Merck, >99.8%) was deoxygenated and dried by standard techniques before use. Ammonium hydroxide (Merck) was purchased as a 25% solution.
327
Table 1 Carbon content for hybrid silicas System
Carbon content (wt%)
SG100A SG50A SG20A SG10A SG5A SG2A SG0A
2.5 5.8 10.8 19.9 37.7 53.4 68.6
SG10B SG10C
22.8 21.9
Zorbax C18a Partisil 10 ODS 2a Resolve C18a S-C18b Partisil-40b
15.0 15.0 12.0 18.7 20.5
a Ref. [16]. b Ref. [17].
2.2. Synthesis of xerogel by the hydrolytic alkaline route 3. Results and discussion Xerogels were synthesized in accordance with the Stöber synthesis [17]. In a typical preparation, 20 mL of ammonia solution was diluted in 100 mL of ethanol in a two-neck flask equipped with a mechanical stirrer. Five milliliters of a TEOS ethanol solution (1:4; V/V) was added to this solution and the mixture was stirred for 2 h. Then, 2.10 mmol of the organosilane (ODS) was diluted in ethanol (4 mL) and added dropwise to the solution. The addition time lasted ca. 1 h 45 min. After the addition, the mixture was stirred for two additional hours. The silica was then dried under vacuum and washed with 5 × 10 mL of ethanol. The silica was finally dried under vacuum for 16 h. 2.3. Characterization of the silicas 2.3.1. Elemental analysis (CHN) Carbon content was determined in a Perkin–Elmer MCHNSO/2400 analyzer. Measurements were made in triplicate and the results are expressed as a mean. 2.3.2. Scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDX) SEM and EDX experiments were carried out on a JEOL JSM/6060 and JEOL JSM/5800, respectively. The catalysts were initially fixed on a carbon tape and then coated with gold by conventional sputtering techniques. The employed accelerating voltage was 10 kV for SEM and 20 kV for EDX. 2.3.3. 29 Si magic angle spin nuclear magnetic resonance (29 Si MAS-NMR) Solid-state NMR measurements of 29 Si were performed on a Chemagnetics CMX-300 (Varian, USA). Samples were transferred to zirconia rotors. Measurements were performed at room temperature at 59.5 MHz. The number of scans was 25,000 for 29 Si. NMR parameters were a contact time of 3 ms and a recycle time of 2 s.
3.1. Effect on the carbon content Silicas modified with octadecylsilane (ODS) were obtained by the sol–gel method. Table 1 reports the grafted carbon content for the resulting hybrid silicas. In Table 1, SG means silica obtained by the sol–gel method. The letters A, B, and C refer to the synthetic route, which is related to the ODS addition time: 2 h after the addition of TEOS, just before the addition of TEOS (5 min), and coaddition of both reagents, respectively. The number refers to TEOS amount in the ODS:TEOS ratio, i.e., 0, 2, 5, and 50 means 1:0, 1:2, 1:5, and 1:50. For comparative reasons, data from the literature were also included. According to Table 1, the carbon content is very variable, depending, as expected, on the amount of ODS. The highest carbon content was achieved in the case of pure ODS (SG0A), while lower, in the case of pure TEOS (SG100). For the systems resulting from the cohydrolysis between ODS and TEOS, the higher the ODS content in the reaction milieu, the higher the carbon content in the resulting silica. For instance, an increase of 170% was observed when the ratio between ODS and TEOS was decreased from 10 to 2. The ODS addition time does not influence the final carbon content. It is worth noting that most of the commercial phases present roughly 15% of carbon content. In some cases, these values are very close to those obtained in the present study by the sol–gel method. 3.2. Effect on the particle morphology Fig. 1 shows a series of hybrid silicas produced at different ODS:TEOS ratios. According to Fig. 1, when the sol–gel reaction is carried out solely with TEOS, spherical particle are produced, as previously reported by Stöber et al. [15]. In the present study, the size of the resulting sphere is around 0.5–1.0 µm. Larger particles were observed for ODS:TEOS 1:20 ratio. The use of ODS in the copolymerization in the ODS:TEOS ratio 1:10 yields a
328
R. Brambilla et al. / Journal of Colloid and Interface Science 312 (2007) 326–332
(a)
(b)
(c)
(d)
(e)
(f)
Fig. 1. SEM images of silica produced at different ODS:TEOS ratios: (a) SG100A, (b) SG20A, (c) SG10A, (d) SG5A, (e) SG2A, and (f) SG0A (15,000×).
system combining spherical and lamellar patterns zones. Lowering the ODS:TEOS ratio 1 to 5, the particle size diameter is shifted down to 0.2 µm. As the amount of ODS is increased, the presence of lamellar domains is enhanced. For SGOA, which was produced by pure ODS, the presence of only lamellar patterns is observed. SEM-EDX was performed in order to evaluate the composition of both the spheres and the lamellar domains. Fig. 2 represents the 16 spots that were analyzed in terms of Si, O, and C percentage in the case of sample SG2A. Table 2 shows the elemental composition of the 16 spots, comprising both spherical and lamellar domains. According to Table 2, there are no significant differences in terms of elemental composition between spherical and lamellar domains, suggesting a homogeneous composition, when ODS was added 2 h after the addition of TEOS (route A).
3.3. Effect on the particle size The addition time of both components, TEOS and ODS, was also investigated. As discussed in Table 1, carbon content does not vary with the addition moment. The resulting hybrid systems were further investigated by SEM (Fig. 3). As shown in Fig. 3, the concomitant addition of both components affords spherical particles, with a low degree of agglomeration (Fig. 3a). When ODS is added 5 min after the complete introduction of TEOS in the reaction milieu, a less organized system is observed, combining clusters of spherical particles and lamellar domains. A more dispersed spherical silica distribution is achieved when ODS is added 2 h after the complete addition of TEOS. In order to monitor the growing process in the case of route C (concomitant addition of both reactants), samples were col-
R. Brambilla et al. / Journal of Colloid and Interface Science 312 (2007) 326–332
329
(a)
Fig. 2. SEM image of silica SG2A. Spots in which EDX measurements were performed. Table 2 Chemical composition of hybrid silica particles (SG2A) measured by SEMEDX in spherical and lamellar domains Spot
Domain
Si (%)
O (%)
C (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Spherical Spherical Spherical Spherical Spherical Lamellar Lamellar Spherical Spherical Lamellar Lamellar Lamellar Lamellar Lamellar Lamellar Spherical
17.2 16.6 17.9 18.6 16.3 17.1 16.1 16.3 16.4 17.2 15.8 17.3 14.8 17.4 17.6 18.9
65.4 65.1 66.1 66.6 64.9 65.4 64.7 64.9 65.0 65.5 64.6 65.6 63.8 65.7 65.9 66.7
17.4 18.3 16.0 14.8 18.8 17.5 19.2 18.8 18.6 17.2 19.6 17.0 21.4 16.8 16.5 14.3
lected during the polymerization reaction of sample SG10C for further analysis by SEM-EDX. The resulting micrographs are shown in Fig. 4. According to Fig. 4, after 1 min of reaction time, spherical 0.1-µm particles and lamellar domains can already be observed. As the reaction proceeds, spherical particles are grown around the lamellar domains reaching 0.5 µm in diameter after 80 min of reaction time. It seems that in the case of coaddition of both ODS and TEOS, lamellar domains are initially predominantly formed, which are covered by spherical silica along the reaction time. Data obtained by EDX-SEM, expressed in terms of C/Si ratio, are shown in Fig. 5: As the reaction proceeds, the C/Si ratio decreases, suggesting that organosilane groups are mainly located in the lamellar domains, while the inorganic silica network in the spherical particles. Such results are different from those observed in the case of route A (Table 2), suggesting that the addition time might influence local composition. The effect of the speed of stirring on the morphological properties of the hybrid silica was also evaluated. Fig. 6 shows mi-
(b)
(c) Fig. 3. SEM images of silica produced at addition time: (a) SG10C, (b) SG10B, and (c) SG10A (5000×).
crographs of two silica SG10A obtained under 200 and 14,000 rpm. According to Fig. 6, the use of 14,000 rpm did not alter the spherical morphology of the silica, nor the laminar domains. Nevertheless, the resulting particles presented mean size diameter in the range of 0.2 µm, i.e., much lower than that obtained with stirring under 200 rpm, which was between 0.4 and 1.0 µm. The effect of a terminating agent was also evaluated. Therefore, (CH3 )3 SiCl was added to the reaction milieu during the hybrid synthesis of SG10A. Fig. 7 shows the resulting micrograph. For comparative reasons, SG10A produced without terminating agent is also shown. The addition of (CH3 )3 SiCl did not alter the particle size, which remained in the micrometer range. Nevertheless, the
330
R. Brambilla et al. / Journal of Colloid and Interface Science 312 (2007) 326–332
(a)
(b)
(c)
(d)
(e) Fig. 4. SEM images of silica SG10C collected at different reaction times: (a) 1 min, (b) 10 min, (c) 20 min, (d) 40 min, and (e) 80 min.
3.4. Effect on silica structure
Fig. 5. Influence of the reaction time on the C/Si ratio for SG10C, determined by SEM-EDX.
morphology seems to be changed, since the lamellar pattern is not observed.
The 29 Si NMR has been used for the structural characterization of pure and modified silica gel [18–23]. Fig. 8 shows NMR spectra of silicas containing different amounts of ODS. Spectrum (a) of Fig. 8 shows three signals: −92, −101, and −110 ppm, assigned to Q2 , Q3 , and Q4 , which are attributed to siloxane species on a silica surface or near it (in the case of Q4 ), as shown in Scheme 1 [18]. In Scheme 1, Q represents siloxane sites and exponent is the number of oxygen atoms in a silica network. Comparing the relative intensity among the peaks in Fig. 8, that related to Q3 species is preponderant in comparison to that of Q4 . These results suggest the predominance of less condensed structures. Also, according to spectra in Fig. 8, no significant changes
R. Brambilla et al. / Journal of Colloid and Interface Science 312 (2007) 326–332
331
(a)
Fig. 8. 29 Si NMR spectra: (a) SG100, (b) SG50A, (c) SG10A, and (d) SG0.
(b) Fig. 6. SEM images of silica SG10A produced at speed of stirring: (a) 200 rpm, and (b) 14,000 rpm (15,000×).
Scheme 1. Q siloxane sites.
(a) Scheme 2. Di- and trifunctional organosilicon species.
(b) Fig. 7. SEM images of silica SG10A produced (a) in the absence, and (b) in the presence of (CH3 )3 SiCl (15,000×).
could be observed in the assignments of the nature of the Q species. The addition of ODS in low concentration (ODS:TEOS = 1:50), as in the case of SG50A (Fig. 8b), engenders a small shift in the signal assigned to Q species, and two new signals, of lower intensity, are observed: −59 and −68 ppm. Such signals can be attributed to trifunctional organosilicon species (T2 and T3 ), as shown in Scheme 2. As the amount of ODS is enhanced (SG10A), the signal assigned to T3 increases, suggesting a higher degree of crosslinking in this case, in comparison to SG50A. It is worth noting that the degree of cross-linking is an important factor for some applications such as chromatographic phases, since the hydrol-
332
R. Brambilla et al. / Journal of Colloid and Interface Science 312 (2007) 326–332
Scheme 3. Monofunctional organosilicon species.
bearing lamellar domains. Besides, other experimental parameters also influence the final grain morphology: The cohydrolysis and condensation of both reagents result in silica mainly presenting a spherical morphology. If TEOS is prepolymerized, hybrid silicas bearing lamellar domains are mainly formed as ODS is added to the system at longer reaction times. The use of a terminator reagent (CH3 SiCl) also had an influence on the particle morphology, reducing lamellar domains in the resulting hybrid silica. Particle size was shown not to be dominated by the ODS: TEOS ratio. Nevertheless, bigger particles were observed when ODS was added after letting TEOS prepolymerize. Conversely, smaller particle were obtained, as expected, with higher stirring speed. Acknowledgments R. Brambilla thanks CAPES for the grant. CNPq and FAPERGS/PRONEX are thanked for the partial financial support. References
Fig. 9. 29 Si NMR spectra: (a) SG10C and (b) SG10A.
ysis of a functional group is hindered by the multiple bonding between the organosilane and the silica surface [18]. In the case of hydrolysis of pure ODS (SG0), according to Fig. 8, the disappearance of the signal attributed to Q2 indicates that such silica has a more condensed structure. The presence of a signal assigned to Q4 suggests that some hydrolysis of ODS ligand is taking place during the sol–gel reaction. Besides, the intensity of peaks assigned to T3 increased in comparison to that of T2 . Two new signals were also observed, −43 and −51 ppm, which might be due to T1 species, bearing or not ethoxy groups, as shown in Scheme 3. According to spectrum (d) of Fig. 8, the relative high intensity signal attributed to T3 species suggests a high degree of cross-linking in comparison to the other silicas. Silicas produced at different addition times (SG10A and SG10C) present different 29 Si NMR spectra, as shown in Fig. 9. In the case of concomitant addition of both TEOS and ODS (SG10C), the signal assigned to T3 species (−67 ppm) is much more intense than that attributed to T2 species (−58 ppm). Comparing both silicas, SG10C is more prone to be chemically stable due to the higher degree of cross-linking. 4. Conclusions The morphology of hybrid silica particles bearing octadecyl groups depends on the ODS:TEOS ratio. Lower amounts of ODS yield spherical silica, while higher ones produced silica
[1] U. Schubert, N. Hüsing, Synthesis of Inorganic Materials, second ed., Wiley, Weinheim, 2005. [2] A. Pepe, M. Aparicio, S. Ceré, A. Durán, Mater. Lett. 59 (2005) 3937. [3] M. Trilla, R. Pleixats, M.W.C. Man, C. Bied, J.J.E. Moreau, Tetrahedron Lett. 47 (2006) 2399–2403. [4] Y.A. Shchipunov, T.Yu. Karpenko, I.Yu. Bakunina, Y.V. Burtseva, T.N. Zvyagintseva, J. Biochem. Biophys. Methods 58 (2004) 25. [5] L.-J. Yan, Q.-H. Zhang, Y.-Q. Feng, W.-B. Zhang, T. Li, L.-H. Zhang, Y.-K. Zhang, J. Chromatogr. A 1121 (2006) 92. [6] C.J. Cornelius, E. Marand, J. Membrane Sci. 202 (2002) 97. [7] Yu.A. Shchipunov, A. Kojima, T. Imae, J. Colloid Interface Sci. 285 (2005) 574. [8] E.F. Vansant, P. Van Der Voort, K.C. Vrancken, Characterization and Chemical Modification of the Silica Surface, Elsevier, Amsterdam, 1995. [9] S.A. Kulkarni, B.A. Kakade, I.S. Mulla, V.K. Pillai, J. Colloid Interface Sci. 299 (2006) 777. [10] E. Lesellier, C. West, A. Tchapla, J. Chromatogr. A 1111 (2006) 62. [11] S. Suzuki, A. Nakajima, M. Sakai, J.-H. Song, N. Yoshida, Y. Kameshima, K. Okada, Surf. Sci. 600 (2006) 2214. [12] M.A. Zanjanchi, H. Noei, M. Moghimi, Talanta 70 (2006) 933. [13] H. Schmidt, H. Böttner, in: H.E. Berna, W.O. Robert (Eds.), Colloidal Silica: Fundamentals and Applications, Taylor & Francis, Boca Raton, FL, 2006, p. 645. [14] R. Brambilla, G.P. Pires, J.H.Z. dos Santos, M.S.L. Miranda, B. Chornik, J. Electron Spectrosc. Relat. Phenom., in press. [15] W. Stöber, A. Fink, E. Bohn, J. Colloid Interface Sci. 26 (1968) 62. [16] R.M. Smith, Gas and Liquid Chromatography in Analytical Chemistry, Wiley, Suffolk, 1998, pp. 253–310. [17] K.D. Wyndham, J.E. O’Gara, T.H. Walter, K.H. Glose, N.L. Lawrence, B.A. Alden, G.S. Izzo, C.J. Hudalla, P.C. Iraneta, Anal. Chem. 75 (2003) 6781. [18] M. Pursch, R. Brindle, A. Ellvanger, L.C. Sander, C.M. Bell, H. Handel, K. Albert, Solid State Nucl. Magn. Reson. 9 (1997) 191–201. [19] Y. Bereznitski, M. Jaroniec, J. Chromatogr. A 828 (1998) 51–58. [20] A.P. Legrand, H. Hommel, J.B. d’Espinose de la Caillerie, Colloids Surf. A 158 (1999) 157. [21] J.J. Pesec, M.T. Matyska, R.J. Yu, J. Chromatogr. A 947 (2002) 195–203. [22] M.D. Bruch, H.O. Fatunmbi, J. Chromatogr. A 1021 (2003) 61–70. [23] M.C. Capel-Sanchez, L. Barrio, J.M. Campos-Martin, J.L.G. Fierro, J. Colloid Interface Sci. 277 (2004) 146–153.
Octadecylsilane hybrid silicas prepared by the sol–gel method: Morphological and textural aspects Rodrigo Brambilla a , Gilvan P. Pires a , João H.Z. dos Santos a,∗ , Márcia S. Lacerda Miranda b a Instituto de Química, UFRGS, Av. Bento Gonçalves, 9500, Porto Alegre, 91501-970 Brazil b Braskem S.A., III Pólo Petroquímico, Via Oeste, Lote 05, Triunfo, 95853-000 Brazil
Received 22 December 2006; accepted 2 March 2007 Available online 26 April 2007
Abstract A series of octadecylsilane-modified silicas was prepared by the sol–gel method through the hydrolysis and cocondensation of tetraethylorthosilicate (TEOS) with octadecyltriethoxysilane (ODS). The ODS:TEOS ratio was varied between 0:100 and 100:0. The resulting carbon content was between 2.5 and 53.4%. In the case of pure ODS, the resulting silica presented 68.6% of C. Hybrid silicas were characterized by scanning electron microscopy, energy dispersive X-ray spectroscopy, and 29 Si nuclear magnetic resonance spectroscopy. Spheres of ca. 0.5–1.0 µm were obtained in the case of hydrolysis of pure TEOS. The combination of ODS:TEOS ratio yielded systems combining spherical and lamellar patterns zones. Monitoring the particle growth, it seems that spherical particles grow around lamellar zones, these latter concentrating the organosilicon moieties. The degree of cross-linking of ODS moieties was shown to be dependent on the ODS addition time and stirring speed. © 2007 Elsevier Inc. All rights reserved. Keywords: Hybrid silica; Sol–gel; Octadecylsilane; ODS
1. Introduction One of the major advances of sol–gel processing is the possibility of synthesizing hybrid inorganic–organic materials. The so-called hybrid materials combine, to some extent, the properties of inorganic and organic compound in one material [1]. Recent research on hybrid materials comprises coating on carbon steel [2], synthesis of supported catalysts [3] and immobilized enzymes [4], chromatographic stationary phases [5], membrane materials [6], and nanocomposites [7], to note a few. Hybrid silicas produced by the sol–gel method can be obtained from the hydrolysis and cocondensation of tetraethylorthosilicate (TEOS) with other organosiloxanes (Rx Si(OR)4−x , where R is an alkyl groups). In these materials, TEOS functions as building blocks to construct the framework, while the organosiloxanes, with nonhydrolyzable organic groups, contribute both to the framework silica units and to the organic * Corresponding author. Fax: +55 51 3316 7304.
E-mail address: [email protected] (J.H.Z. dos Santos). 0021-9797/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2007.03.003
surface functional groups. A large number of silanes carrying different functional groups have been used for silica modification [8]. Among them, octadecyltrimethoxysilane (ODS) and octadecyltrichlorosilane (OTS) have been largely used for surface chemical modification. Examples of application of such reactions can be found in the development of self-assembled monolayers [9], of chromatographic phases [10], of hydrophobic coatings [11], and of adsorbents [12]. If multicomponent systems are prepared by hydrolysis and condensation of alkoxide precursors, the reaction rates of hydrolysis and condensation (including aggregation) become very important for the distribution of the different components. Besides, it might influence the resulting textural and chemical properties of the final product [13]. In a previous study, we investigated the effect of producing ODS-modified silica by grafting and by the sol–gel method [14]. In such materials, the conformation of the octadecyl chains, as well as the surface coverage, depended on the preparative route. One application of hybrid materials is as catalyst support [15,16], and in catalysis, morphology can be an important issue. For instance, some polymerization processes require the repli-
R. Brambilla et al. / Journal of Colloid and Interface Science 312 (2007) 326–332
cation of the catalyst (and consequently of the catalyst support) spherical morphology. In the present paper, we evaluated the effects of the TEOS/ODS ratio, the addition time, and the stirring speed on the surface coverage, and on the morphological and the textural characteristics of the resulting hybrid materials. 2. Experimental 2.1. Materials Octadecyltrimethoxysilane (Acros, 90%), octyltrimethoxysilane (Aldrich, 90%), methyltrichlorosilane (Aldrich, 90%), and tetraethylorthosilicate (TEOS) (Merck, >98%) were used without further purification. Ethanol (Merck, >99.8%) was deoxygenated and dried by standard techniques before use. Ammonium hydroxide (Merck) was purchased as a 25% solution.
327
Table 1 Carbon content for hybrid silicas System
Carbon content (wt%)
SG100A SG50A SG20A SG10A SG5A SG2A SG0A
2.5 5.8 10.8 19.9 37.7 53.4 68.6
SG10B SG10C
22.8 21.9
Zorbax C18a Partisil 10 ODS 2a Resolve C18a S-C18b Partisil-40b
15.0 15.0 12.0 18.7 20.5
a Ref. [16]. b Ref. [17].
2.2. Synthesis of xerogel by the hydrolytic alkaline route 3. Results and discussion Xerogels were synthesized in accordance with the Stöber synthesis [17]. In a typical preparation, 20 mL of ammonia solution was diluted in 100 mL of ethanol in a two-neck flask equipped with a mechanical stirrer. Five milliliters of a TEOS ethanol solution (1:4; V/V) was added to this solution and the mixture was stirred for 2 h. Then, 2.10 mmol of the organosilane (ODS) was diluted in ethanol (4 mL) and added dropwise to the solution. The addition time lasted ca. 1 h 45 min. After the addition, the mixture was stirred for two additional hours. The silica was then dried under vacuum and washed with 5 × 10 mL of ethanol. The silica was finally dried under vacuum for 16 h. 2.3. Characterization of the silicas 2.3.1. Elemental analysis (CHN) Carbon content was determined in a Perkin–Elmer MCHNSO/2400 analyzer. Measurements were made in triplicate and the results are expressed as a mean. 2.3.2. Scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDX) SEM and EDX experiments were carried out on a JEOL JSM/6060 and JEOL JSM/5800, respectively. The catalysts were initially fixed on a carbon tape and then coated with gold by conventional sputtering techniques. The employed accelerating voltage was 10 kV for SEM and 20 kV for EDX. 2.3.3. 29 Si magic angle spin nuclear magnetic resonance (29 Si MAS-NMR) Solid-state NMR measurements of 29 Si were performed on a Chemagnetics CMX-300 (Varian, USA). Samples were transferred to zirconia rotors. Measurements were performed at room temperature at 59.5 MHz. The number of scans was 25,000 for 29 Si. NMR parameters were a contact time of 3 ms and a recycle time of 2 s.
3.1. Effect on the carbon content Silicas modified with octadecylsilane (ODS) were obtained by the sol–gel method. Table 1 reports the grafted carbon content for the resulting hybrid silicas. In Table 1, SG means silica obtained by the sol–gel method. The letters A, B, and C refer to the synthetic route, which is related to the ODS addition time: 2 h after the addition of TEOS, just before the addition of TEOS (5 min), and coaddition of both reagents, respectively. The number refers to TEOS amount in the ODS:TEOS ratio, i.e., 0, 2, 5, and 50 means 1:0, 1:2, 1:5, and 1:50. For comparative reasons, data from the literature were also included. According to Table 1, the carbon content is very variable, depending, as expected, on the amount of ODS. The highest carbon content was achieved in the case of pure ODS (SG0A), while lower, in the case of pure TEOS (SG100). For the systems resulting from the cohydrolysis between ODS and TEOS, the higher the ODS content in the reaction milieu, the higher the carbon content in the resulting silica. For instance, an increase of 170% was observed when the ratio between ODS and TEOS was decreased from 10 to 2. The ODS addition time does not influence the final carbon content. It is worth noting that most of the commercial phases present roughly 15% of carbon content. In some cases, these values are very close to those obtained in the present study by the sol–gel method. 3.2. Effect on the particle morphology Fig. 1 shows a series of hybrid silicas produced at different ODS:TEOS ratios. According to Fig. 1, when the sol–gel reaction is carried out solely with TEOS, spherical particle are produced, as previously reported by Stöber et al. [15]. In the present study, the size of the resulting sphere is around 0.5–1.0 µm. Larger particles were observed for ODS:TEOS 1:20 ratio. The use of ODS in the copolymerization in the ODS:TEOS ratio 1:10 yields a
328
R. Brambilla et al. / Journal of Colloid and Interface Science 312 (2007) 326–332
(a)
(b)
(c)
(d)
(e)
(f)
Fig. 1. SEM images of silica produced at different ODS:TEOS ratios: (a) SG100A, (b) SG20A, (c) SG10A, (d) SG5A, (e) SG2A, and (f) SG0A (15,000×).
system combining spherical and lamellar patterns zones. Lowering the ODS:TEOS ratio 1 to 5, the particle size diameter is shifted down to 0.2 µm. As the amount of ODS is increased, the presence of lamellar domains is enhanced. For SGOA, which was produced by pure ODS, the presence of only lamellar patterns is observed. SEM-EDX was performed in order to evaluate the composition of both the spheres and the lamellar domains. Fig. 2 represents the 16 spots that were analyzed in terms of Si, O, and C percentage in the case of sample SG2A. Table 2 shows the elemental composition of the 16 spots, comprising both spherical and lamellar domains. According to Table 2, there are no significant differences in terms of elemental composition between spherical and lamellar domains, suggesting a homogeneous composition, when ODS was added 2 h after the addition of TEOS (route A).
3.3. Effect on the particle size The addition time of both components, TEOS and ODS, was also investigated. As discussed in Table 1, carbon content does not vary with the addition moment. The resulting hybrid systems were further investigated by SEM (Fig. 3). As shown in Fig. 3, the concomitant addition of both components affords spherical particles, with a low degree of agglomeration (Fig. 3a). When ODS is added 5 min after the complete introduction of TEOS in the reaction milieu, a less organized system is observed, combining clusters of spherical particles and lamellar domains. A more dispersed spherical silica distribution is achieved when ODS is added 2 h after the complete addition of TEOS. In order to monitor the growing process in the case of route C (concomitant addition of both reactants), samples were col-
R. Brambilla et al. / Journal of Colloid and Interface Science 312 (2007) 326–332
329
(a)
Fig. 2. SEM image of silica SG2A. Spots in which EDX measurements were performed. Table 2 Chemical composition of hybrid silica particles (SG2A) measured by SEMEDX in spherical and lamellar domains Spot
Domain
Si (%)
O (%)
C (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Spherical Spherical Spherical Spherical Spherical Lamellar Lamellar Spherical Spherical Lamellar Lamellar Lamellar Lamellar Lamellar Lamellar Spherical
17.2 16.6 17.9 18.6 16.3 17.1 16.1 16.3 16.4 17.2 15.8 17.3 14.8 17.4 17.6 18.9
65.4 65.1 66.1 66.6 64.9 65.4 64.7 64.9 65.0 65.5 64.6 65.6 63.8 65.7 65.9 66.7
17.4 18.3 16.0 14.8 18.8 17.5 19.2 18.8 18.6 17.2 19.6 17.0 21.4 16.8 16.5 14.3
lected during the polymerization reaction of sample SG10C for further analysis by SEM-EDX. The resulting micrographs are shown in Fig. 4. According to Fig. 4, after 1 min of reaction time, spherical 0.1-µm particles and lamellar domains can already be observed. As the reaction proceeds, spherical particles are grown around the lamellar domains reaching 0.5 µm in diameter after 80 min of reaction time. It seems that in the case of coaddition of both ODS and TEOS, lamellar domains are initially predominantly formed, which are covered by spherical silica along the reaction time. Data obtained by EDX-SEM, expressed in terms of C/Si ratio, are shown in Fig. 5: As the reaction proceeds, the C/Si ratio decreases, suggesting that organosilane groups are mainly located in the lamellar domains, while the inorganic silica network in the spherical particles. Such results are different from those observed in the case of route A (Table 2), suggesting that the addition time might influence local composition. The effect of the speed of stirring on the morphological properties of the hybrid silica was also evaluated. Fig. 6 shows mi-
(b)
(c) Fig. 3. SEM images of silica produced at addition time: (a) SG10C, (b) SG10B, and (c) SG10A (5000×).
crographs of two silica SG10A obtained under 200 and 14,000 rpm. According to Fig. 6, the use of 14,000 rpm did not alter the spherical morphology of the silica, nor the laminar domains. Nevertheless, the resulting particles presented mean size diameter in the range of 0.2 µm, i.e., much lower than that obtained with stirring under 200 rpm, which was between 0.4 and 1.0 µm. The effect of a terminating agent was also evaluated. Therefore, (CH3 )3 SiCl was added to the reaction milieu during the hybrid synthesis of SG10A. Fig. 7 shows the resulting micrograph. For comparative reasons, SG10A produced without terminating agent is also shown. The addition of (CH3 )3 SiCl did not alter the particle size, which remained in the micrometer range. Nevertheless, the
330
R. Brambilla et al. / Journal of Colloid and Interface Science 312 (2007) 326–332
(a)
(b)
(c)
(d)
(e) Fig. 4. SEM images of silica SG10C collected at different reaction times: (a) 1 min, (b) 10 min, (c) 20 min, (d) 40 min, and (e) 80 min.
3.4. Effect on silica structure
Fig. 5. Influence of the reaction time on the C/Si ratio for SG10C, determined by SEM-EDX.
morphology seems to be changed, since the lamellar pattern is not observed.
The 29 Si NMR has been used for the structural characterization of pure and modified silica gel [18–23]. Fig. 8 shows NMR spectra of silicas containing different amounts of ODS. Spectrum (a) of Fig. 8 shows three signals: −92, −101, and −110 ppm, assigned to Q2 , Q3 , and Q4 , which are attributed to siloxane species on a silica surface or near it (in the case of Q4 ), as shown in Scheme 1 [18]. In Scheme 1, Q represents siloxane sites and exponent is the number of oxygen atoms in a silica network. Comparing the relative intensity among the peaks in Fig. 8, that related to Q3 species is preponderant in comparison to that of Q4 . These results suggest the predominance of less condensed structures. Also, according to spectra in Fig. 8, no significant changes
R. Brambilla et al. / Journal of Colloid and Interface Science 312 (2007) 326–332
331
(a)
Fig. 8. 29 Si NMR spectra: (a) SG100, (b) SG50A, (c) SG10A, and (d) SG0.
(b) Fig. 6. SEM images of silica SG10A produced at speed of stirring: (a) 200 rpm, and (b) 14,000 rpm (15,000×).
Scheme 1. Q siloxane sites.
(a) Scheme 2. Di- and trifunctional organosilicon species.
(b) Fig. 7. SEM images of silica SG10A produced (a) in the absence, and (b) in the presence of (CH3 )3 SiCl (15,000×).
could be observed in the assignments of the nature of the Q species. The addition of ODS in low concentration (ODS:TEOS = 1:50), as in the case of SG50A (Fig. 8b), engenders a small shift in the signal assigned to Q species, and two new signals, of lower intensity, are observed: −59 and −68 ppm. Such signals can be attributed to trifunctional organosilicon species (T2 and T3 ), as shown in Scheme 2. As the amount of ODS is enhanced (SG10A), the signal assigned to T3 increases, suggesting a higher degree of crosslinking in this case, in comparison to SG50A. It is worth noting that the degree of cross-linking is an important factor for some applications such as chromatographic phases, since the hydrol-
332
R. Brambilla et al. / Journal of Colloid and Interface Science 312 (2007) 326–332
Scheme 3. Monofunctional organosilicon species.
bearing lamellar domains. Besides, other experimental parameters also influence the final grain morphology: The cohydrolysis and condensation of both reagents result in silica mainly presenting a spherical morphology. If TEOS is prepolymerized, hybrid silicas bearing lamellar domains are mainly formed as ODS is added to the system at longer reaction times. The use of a terminator reagent (CH3 SiCl) also had an influence on the particle morphology, reducing lamellar domains in the resulting hybrid silica. Particle size was shown not to be dominated by the ODS: TEOS ratio. Nevertheless, bigger particles were observed when ODS was added after letting TEOS prepolymerize. Conversely, smaller particle were obtained, as expected, with higher stirring speed. Acknowledgments R. Brambilla thanks CAPES for the grant. CNPq and FAPERGS/PRONEX are thanked for the partial financial support. References
Fig. 9. 29 Si NMR spectra: (a) SG10C and (b) SG10A.
ysis of a functional group is hindered by the multiple bonding between the organosilane and the silica surface [18]. In the case of hydrolysis of pure ODS (SG0), according to Fig. 8, the disappearance of the signal attributed to Q2 indicates that such silica has a more condensed structure. The presence of a signal assigned to Q4 suggests that some hydrolysis of ODS ligand is taking place during the sol–gel reaction. Besides, the intensity of peaks assigned to T3 increased in comparison to that of T2 . Two new signals were also observed, −43 and −51 ppm, which might be due to T1 species, bearing or not ethoxy groups, as shown in Scheme 3. According to spectrum (d) of Fig. 8, the relative high intensity signal attributed to T3 species suggests a high degree of cross-linking in comparison to the other silicas. Silicas produced at different addition times (SG10A and SG10C) present different 29 Si NMR spectra, as shown in Fig. 9. In the case of concomitant addition of both TEOS and ODS (SG10C), the signal assigned to T3 species (−67 ppm) is much more intense than that attributed to T2 species (−58 ppm). Comparing both silicas, SG10C is more prone to be chemically stable due to the higher degree of cross-linking. 4. Conclusions The morphology of hybrid silica particles bearing octadecyl groups depends on the ODS:TEOS ratio. Lower amounts of ODS yield spherical silica, while higher ones produced silica
[1] U. Schubert, N. Hüsing, Synthesis of Inorganic Materials, second ed., Wiley, Weinheim, 2005. [2] A. Pepe, M. Aparicio, S. Ceré, A. Durán, Mater. Lett. 59 (2005) 3937. [3] M. Trilla, R. Pleixats, M.W.C. Man, C. Bied, J.J.E. Moreau, Tetrahedron Lett. 47 (2006) 2399–2403. [4] Y.A. Shchipunov, T.Yu. Karpenko, I.Yu. Bakunina, Y.V. Burtseva, T.N. Zvyagintseva, J. Biochem. Biophys. Methods 58 (2004) 25. [5] L.-J. Yan, Q.-H. Zhang, Y.-Q. Feng, W.-B. Zhang, T. Li, L.-H. Zhang, Y.-K. Zhang, J. Chromatogr. A 1121 (2006) 92. [6] C.J. Cornelius, E. Marand, J. Membrane Sci. 202 (2002) 97. [7] Yu.A. Shchipunov, A. Kojima, T. Imae, J. Colloid Interface Sci. 285 (2005) 574. [8] E.F. Vansant, P. Van Der Voort, K.C. Vrancken, Characterization and Chemical Modification of the Silica Surface, Elsevier, Amsterdam, 1995. [9] S.A. Kulkarni, B.A. Kakade, I.S. Mulla, V.K. Pillai, J. Colloid Interface Sci. 299 (2006) 777. [10] E. Lesellier, C. West, A. Tchapla, J. Chromatogr. A 1111 (2006) 62. [11] S. Suzuki, A. Nakajima, M. Sakai, J.-H. Song, N. Yoshida, Y. Kameshima, K. Okada, Surf. Sci. 600 (2006) 2214. [12] M.A. Zanjanchi, H. Noei, M. Moghimi, Talanta 70 (2006) 933. [13] H. Schmidt, H. Böttner, in: H.E. Berna, W.O. Robert (Eds.), Colloidal Silica: Fundamentals and Applications, Taylor & Francis, Boca Raton, FL, 2006, p. 645. [14] R. Brambilla, G.P. Pires, J.H.Z. dos Santos, M.S.L. Miranda, B. Chornik, J. Electron Spectrosc. Relat. Phenom., in press. [15] W. Stöber, A. Fink, E. Bohn, J. Colloid Interface Sci. 26 (1968) 62. [16] R.M. Smith, Gas and Liquid Chromatography in Analytical Chemistry, Wiley, Suffolk, 1998, pp. 253–310. [17] K.D. Wyndham, J.E. O’Gara, T.H. Walter, K.H. Glose, N.L. Lawrence, B.A. Alden, G.S. Izzo, C.J. Hudalla, P.C. Iraneta, Anal. Chem. 75 (2003) 6781. [18] M. Pursch, R. Brindle, A. Ellvanger, L.C. Sander, C.M. Bell, H. Handel, K. Albert, Solid State Nucl. Magn. Reson. 9 (1997) 191–201. [19] Y. Bereznitski, M. Jaroniec, J. Chromatogr. A 828 (1998) 51–58. [20] A.P. Legrand, H. Hommel, J.B. d’Espinose de la Caillerie, Colloids Surf. A 158 (1999) 157. [21] J.J. Pesec, M.T. Matyska, R.J. Yu, J. Chromatogr. A 947 (2002) 195–203. [22] M.D. Bruch, H.O. Fatunmbi, J. Chromatogr. A 1021 (2003) 61–70. [23] M.C. Capel-Sanchez, L. Barrio, J.M. Campos-Martin, J.L.G. Fierro, J. Colloid Interface Sci. 277 (2004) 146–153.