12134
Langmuir 2008, 24, 12134-12137
Size-Dependent Shape Evolution of Silica Nanoparticles into Hollow Structures Sang-Jae Park, Yoo-Jin Kim, and So-Jung Park* Department of Chemistry, UniVersity of PennsylVania, Philadelphia, PennsylVania 19104 ReceiVed September 3, 2008 Hollow silica nanoparticles can be spontaneously generated without a template on the basis of the porous nature of silica and the high surface energy on the nanometer scale. We show that solid silica particles synthesized by the Sto¨ber and microemulsion methods initially develop small pores inside the nanoparticles under slightly basic conditions as a result of base-catalyzed etching. With further reaction, those small seed pores merge into a single void to reduce the surface energy of small pores, generating well-defined hollow nanoparticles. This behavior is unique to nanometersized porous materials, and the shape evolution is size-dependent, reinforcing the importance of evaluating the reactivity and structural changes of nanomaterials as well as their physical properties in different size ranges. The mechanism described here provides a simple way to generate uniform hollow nanoparticles of porous materials.
Introduction Silica nanostructures have been widely studied in a range of areas including catalysts, drug delivery, and biological imaging.1-5 The ability to manipulate the morphology of silica-based materials is undoubtedly important because it determines the local structure and properties closely related to those applications. Here, we report that well-defined hollow silica nanoparticles can be spontaneously generated on the basis of the unique properties of nanometer-sized silica without using a template. Hollow nanostructures are generally prepared by using soft or hard templates as sacrificial materials. For example, mesoporous silica materials have been synthesized using small surfactants and block copolymers as templates.6,7 Discrete silica nanoparticles and nanorods with hollow cores have been synthesized using various soft and hard templates such as polymer, metal, and semiconductor nanoparticles.8-10 Recently, there have been several reports on the templateless synthesis of hollow nanoparticles. Alivisatos and co-workers developed a novel synthesis method for hollow metal chalcogenide nanocrystals based on the Kirkendall effect.11 Xia et al. reported a progressive shape change of Pd nanocubes into nanoboxes and cages by the corrosion of nanocubes.12 Very recently, Yin et al. described the formation of hollow silica nanoparticles by reacting silica particles with NaBH4.13 Our study described herein shows that the spontaneous hollowing of silica * Corresponding author. E-mail:
[email protected]. (1) Bergna, H. E.; Roberts, W. O. Colloidal Silica: Fundamentals and Applications; CRC Taylor & Francis: Boca Raton, FL, 2006. (2) Thomas, J. M.; Johnson, B. F. G.; Raja, R.; Sankar, G.; Midgley, P. A. Acc. Chem. Res. 2003, 36, 20. (3) Kim, J.; Lee, J. E.; Lee, J.; Yu, J. H.; Kim, B. C.; An, K.; Hwang, Y.; Shin, C.-H.; Park, J.-G.; Kim, J.; Hyeon, T. J. Am. Chem. Soc. 2006, 128, 688. (4) Zhou, J.; Wu, W.; Caruntu, D.; Yu, M. H.; Martin, A.; Chen, J. F.; O’Connor, C. J.; Zhou, W. L. J. Phys. Chem. C 2007, 111, 17473. (5) Ow, H.; Larson, D. R.; Srivastava, M.; Baird, B. A.; Webb, W. W.; Wiesner, U. Nano Lett. 2005, 5, 113. (6) Zhao, D. Y.; Feng, J. L.; Huo, Q. S.; Melosh, N.; Fredrickson, G. H.; Chmelka, B. F.; Stucky, G. D. Science 1998, 279, 548. (7) Sayari, A.; Hamoudi, S. Chem. Mater. 2001, 13, 3151. (8) Caruso, F.; Caruso, R. A.; Mohwald, H. Science 1998, 282, 1111. (9) Obare, S. O.; Jana, N. R.; Murphy, C. J. Nano Lett. 2001, 1, 601. (10) Darbandi, M.; Thomann, R.; Nann, T. Chem. Mater. 2007, 19, 1700. (11) Yin, Y.; Rioux, R. M.; Erdonmez, C. K.; Hughes, S.; Somorjai, G. A.; Alivisatos, A. P. Science 2004, 304, 711. (12) Xiong, Y.; Wiley, B.; Chen, J.; Li, Z.-Y.; Yin, Y.; Xia, Y. Angew. Chem., Int. Ed. 2005, 44, 7913. (13) Zhang, T.; Ge, J.; Hu, Y.; Zhang, Q.; Aloni, S.; Yin, Y. Angew. Chem., Int. Ed. 2008, 47, 5806.
nanoparticles can occur in a very mild basic solution of typical bases used in the synthesis of silica nanoparticles, demonstrating that the hollowing effect is a general phenomenon that can commonly occur in silica nanoparticles. We also show that the shape evolution of silica particles is size-dependent, reinforcing the importance of evaluating chemical reactivity and structural transformation as well as physical properties for different size ranges.
Results and Discussion Silica particles with diameters ranging from 20 to 50 nm were prepared by the microemulsion method.14 In typical experiments, a nonionic surfactant, Igepal CO-520 (0.44 g, Aldrich), was dissolved in cyclohexane (9 mL) by sonication and vortex mixing. Then, ammonium hydroxide (28% in water, 85 µL) and TEOS (60 µL) were consecutively added to the solution with vigorous stirring. The nanoparticle size was controlled by simply changing the reaction time. The reaction was continued for 11 and 61 h to produce silica particles with diameters of 21.4 ( 1.6 and 33.1 ( 1.5 nm, respectively. After the reaction, synthesized nanoparticles were precipitated by methanol (45 mL), and the precipitates were collected by centrifugation (9000 rpm, 30 min) and redispersed in 10 mL of ethanol. The purification procedure was repeated one more time, and the purified particles were dispersed in 10 mL of water (Ultrapure water, 18 MΩ, Barnstard). Interestingly, the silica nanoparticles stored in water showed a spontaneous structural change from solid spheres to hollow nanoparticles over a period of a few weeks. Figure 1A-C depicts the progressive shape evolution of 33.1 ( 1.5 nm particles. Initially, multiple small pores were developed inside the nanoparticles (Figure 1B), and they eventually merged into a single cavity to form well-defined hollow nanoparticles (Figure 1C). In general, silica is known for its stability toward even reactive reagents, and many of its applications are based on this chemical stability. However, the stability of nanometer-sized materials can be quite different from that of their bulk counterparts, and conditions that would not affect the bulk material can cause a drastic change in the structures of nanomaterials. It is well known that OH- and F- ions can dissolve silica by coordinating (14) Osseoasare, K.; Arriagada, F. J. Colloids Surf. 1990, 50, 321.
10.1021/la8028885 CCC: $40.75 2008 American Chemical Society Published on Web 10/04/2008
Letters
Langmuir, Vol. 24, No. 21, 2008 12135
Figure 1. Shape evolution of silica nanoparticles synthesized by the (A-C) microemulsion and (D-F) Sto¨ber methods, monitored by taking TEM images after (A, D) 0 days, (B, E) 15 days, and (C, F) 30 days. Each scale bar is 50 nm. Scheme 1. Si-O Bond Breakage and Reformation in Silica Networks Catalyzed by Hydroxide Ions
to Si atoms and weakening Si-O bonds (Scheme 1).15 Thus, the residual amount of ammonium hydroxide, which was used in the synthesis, can remain in the nanoparticle solution after the purification and can potentially act as an etchant for silica nanoparticles, causing the structural change. Indeed, silica nanoparticles (diameter of 43.3 ( 2.2 nm) in acidic solution (0.04 M acetic acid solution, pH 4.0) remained intact for a period of months whereas particles aged in ammonium hydroxide solution (0.04 M, pH 10.0) were slowly transformed into hollow particles (Figure 2). The pH of the purified nanoparticle solutions presented in Figure 1 was 7.5, indicating that mild basic conditions are sufficient to cause profound structural changes in small silica nanoparticles. Although this observation confirms that the shape change is caused by the base-catalyzed etching of silica, it does not account for the hollowing effect. It is indeed very intriguing that the etching takes place predominantly in the core part of the particles rather than at the exterior. Note that the overall particle size remains similar before and after the core etching (Figure 2C). One possible explanation for this phenomenon is the soft templating effect by the surfactant (Igepal CO-520) used in the synthesis. To test the hypothesis, a set of silica nanoparticles were synthesized by the Sto¨ber method,16 which is based on the same chemistry as the microemulsion method described above but does not require surfactants. In a typical experiment, TEOS (750 µL) was injected into ethanol (25 mL) containing ammonium hydroxide (28 %, 1.5 mL) while stirring. The reaction was stopped after 24 and 30 h for 48.7 ( 4.4 and 123.1 ( 9.4 nm particles, respectively. The synthesized silica nanoparticles were collected by centrifugation (9000 rpm, 30 min). The precipitates were subsequently purified once more with ethanol (40 mL) and stored (15) Iler, R. K. The Chemistry of Silica: Solubility. Polymerization, Colloid and Surface Properties, and Biochemistry. John Wiley & Sons: New York, 1979. (16) Sto¨ber, W.; Fink, A.; Bohn, E. J. Colloid Interface Sci. 1968, 26, 62.
in water (5 mL). As shown in Figure 1D-F, the particles (diameter 48.7 ( 4.4 nm) prepared by the Sto¨ber method also showed similar structural changes in water, ruling out the soft template hypothesis. Again, small pores were initially developed inside the particles (Figure 1E), which were eventually merged into a larger void at the core, generating hollow nanoparticles (Figure 1F). In some cases, small silica fragments were observed inside the hollow structure, which were eventually dissolved away to form hollow nanoparticles. From the TEM analysis, it is apparent that the hollowing process takes place through two steps, the nucleation and growth of pores, as described in Figure 3. This process is similar to how nanoparticles grow; in nanoparticle synthesis, small seed particles are first formed in the nucleation step, and larger particles grow at the expense of smaller particles with higher surface energy.17,18 Similarly, in the hollow particle formation described here, small pores are initially formed inside the particles (Figure 3II), and then the higher surface energy of the small pores causes them to collapse into larger voids to reduce the surface energy (Figure 3III). The larger voids eventually merge and form a hollow core (Figure 3IV). The TEM images corresponding to each step of the process are presented in Figure 3A-F. This study demonstrates that the impact of chemical reactions can be quite different for crystalline and porous nanomaterials, which can lead to drastically different shape changes. Typically, the etching of crystalline particles should produce smaller nanoparticles.19 For porous materials, small molecules can diffuse relatively easily in and out of the particles, and chemical reactions and bond rearrangement can occur throughout the nanoparticles, generating small seed pores inside the particles. Although the etching reaction can also occur at the exterior of the nanoparticles, the surface energy of the exterior should be lower than that of small pores inside the nanoparticles, leading to site-dependent etching reactivity. In addition, previous studies on silica colloid synthesis indicated that silica nanoparticles are formed by the aggregation of subparticles (nuclei), followed by growth through monomer addition.20 Thus, the internal parts of the silica particles are presumably less dense and more porous,21 resulting in (17) Murray, C. B.; Kagan, C. R.; Bawendi, M. G. Annu. ReV. Mater. Sci. 2000, 30, 545. (18) Yin, Y.; Alivisatos, A. P. Nature 2005, 437, 664. (19) Yu, W. W.; Qu, L. H.; Guo, W. Z.; Peng, X. G. Chem. Mater. 2003, 15, 2854.
12136 Langmuir, Vol. 24, No. 21, 2008
Letters
Figure 2. Effect of pH on the shape change of silica nanoparticles monitored by taking TEM images (A) after 30 days of incubation in an acidic solution (pH 4.0) and (B) after 30 days of incubation in a basic solution (pH 10.0). Each scale bar is 50 nm. (C) Size histograms of freshly synthesized particles (red) and particles treated with base as shown in panel B (blue).
Figure 3. Proposed mechanism for the formation of hollow silica nanoparticles.
preferential pore growth in the core. It is also important to note that the reaction involved in the hollowing process is reversible (Scheme 1). A series of Si-O bond-breaking and bond-making processes by the hydroxide ions15 can rearrange high-energy bonds in the core, generating hardened fragments of silica and cracks inside the particles, and those small cracks can act as reactive sites for pore growth. This behavior is unique to nanoparticles, and the shape change is very size-dependent (Figure 4). Whereas nanoparticles in the size range of 30-50 nm were consistently transformed into welldefined hollow structures regardless of the synthetic methods used (Figure 4D-F), particles outside of this size range behaved differently. Smaller particles (diameter 21.4 ( 1.6 nm), where the internal and surface sites have similar reactivity, also initially developed small internal pores, but eventually these particles became interconnected and formed larger silica networks (Figure 4A-C). This behavior is presumably due to the reactivities on the outside and inside both being comparably high in these smaller particles. Larger silica particles with a diameter of 123.1 ( 9.4 nm also developed small pores, but hollow structures were not observed even after 2 months, likely because the larger particles are denser21 and more stable to chemical etching (Figure 4G-I). Indeed, larger particles should have properties similar to those of bulk silica and require harsher basic conditions to induce morphological changes. As briefly mentioned above, Yin and co-workers have recently reported a solid-to-hollow conversion of silica particles in the size range of a few hundred nanometers in NaBH4 solutions.13 It is worth comparing the hollowing process described here with the one reported by Yin et al. The two hollowing processes might involve similar chemical reactions, but the shape(20) Bogush, G. H.; Tracy, M. A.; Zukoski, C. F. J. Non-Cryst. Solids 1988, 104, 95. (21) Lecloux, A. J.; Bronckart, J.; Noville, F.; Dodet, C.; Marchot, P.; Pirard, J. P. Colloids Surf. 1986, 19, 359.
Figure 4. Size-dependent shape transformation (A-C, 21.4 ( 1.6 nm; D-F, 48.7 ( 4.4 nm; G-I, 123.1 ( 9.4 nm) monitored by TEM measurements right after the synthesis (A, D, G) and after 30 days (B, E, H). Scale bar: 50 nm. The 21.4 nm particles were synthesized by the microemulsion method, and 48.7 and 123.1 nm particles were synthesized by the Sto¨ber method.
transformation pathways appear to be distinct. Yin and co-workers suggested that, in NaBH4 solution, silica nanoparticles are first dissolved into smaller particles and thin shells are formed around the dissolving core particles later on. In our case, it is apparent that small seed pores are formed inside the nanoparticles, which are merged into a large cavity, generating hollow structures. We attribute the difference in the shape transformation pathways to
Letters
the more porous and less dense nature of the small silica particles used in this study. It is important to note that our study described here demonstrates that the hollowing effect is a general phenomenon that can occur in a common basic solution that can catalyze the reorganization of Si-O bonds. Yin and co-workers have suggested in their recent work13 that NaBO2 resulting from the decomposition of NaBH4 might be responsible for the shell formation and predicted that treating silica particles with a common basic solution such as aqueous NaOH solution should produce solid spheres with reduced sizes, not hollow particles. Our study clearly shows that the hollowing process can actually occur in general basic solutions and reveals that it is the unique chemistry of silica nanoparticles in basic solutions that is responsible for the hollowing behavior rather than the type of etchant. This study is important for the following reasons. First, the mechanism described herein provides a simple and efficient way to prepare hollow nanostructures of porous materials. Indeed, whereas synthetic methods for making nanoparticles with various shapes and sizes have been greatly advanced in the past decade, it is still challenging to control the internal structure without the use of templates. The ability to control the morphology and local environment of silica nanoparticles is undoubtedly important in
Langmuir, Vol. 24, No. 21, 2008 12137
many applications of silica nanostructures, including catalysts and drug delivery. Second, this study emphasizes the importance of studying material properties in different size ranges. In this study, we show that the trivial reactivity of silica in a very dilute basic solution can lead to a drastic morphology change in silica nanoparticles. Silica particles have been widely used in a variety of applications, and it is important to assess their structural integrity and chemical stability. This study also points out that better surface passivation or other synthetic methods should be implemented for applications of small silica particles (
