- Email: [email protected]
Fabrication of colloidal arrays by self-assembly of sub-100 nm silica particles
Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
Contents lists available at ScienceDirect
Colloids and Surfaces A: Physicochemical and Engineering Aspects journal homepage: www.elsevier.com/locate/colsurfa
Fabrication of colloidal arrays by self-assembly of sub-100 nm silica particles Yuan Huang 1 , Jeanne E. Pemberton ∗ Department of Chemistry and Biochemistry, University of Arizona, 1306 East University Boulevard, Tucson, AZ 85721, United States
a r t i c l e
i n f o
Article history: Received 3 September 2010 Received in revised form 30 November 2010 Accepted 10 December 2010 Available online 23 December 2010 Keywords: Silica nanoparticles Colloidal array Self-assembly
a b s t r a c t Self-assembly of sub-100 nm spherical silica nanoparticles into ordered, tightly packed three- and twodimensional arrays was studied. Self-assembly by vertical evaporation was investigated for particles made by two methods: an optimized Stöber method recently reported from this laboratory and a modified reverse micelle method. Ordered, close-packed, two- and three-dimensional structures were formed with spherical nanoparticles made by the optimized Stöber method. Fast Fourier transforms of top-view scanning electron microscopy images document close-packed hexagonal packing for three-dimensional arrays consisting of particles as small as 50 nm. Ranges over which evaporation temperature and suspension particle concentration can be altered as strategies for improving packing quality of 50 nm particles have been defined. Self-assembly behavior that is distinct from that of larger particles (>200 nm) is observed for these sub-100 nm particles in that the ranges over which these variables can be altered to affect array packing quality are much smaller than for larger particles. In contrast, for particles made by the reverse micelle method, only structures with poor packing quality were obtained despite the fact that such particles are typically more uniform than those made by the Stöber method. These results provide clear evidence that, in addition to particle uniformity, other particle properties deriving from fabrication method play important roles in directing self-assembly of sub-100 nm particles. Finally, a rapid selfassembly method based on horizontal evaporation was used to produce close-packed three-dimensional structures of these sub-100 nm particles spanning several millimeters. Although these arrays are not as ordered as those made by vertical evaporation, the strategy reported herein allows tightly packed, crackfree arrays up to microns in thickness to be fabricated. A mechanism for self-assembly by this process is proposed. © 2011 Elsevier B.V. All rights reserved.
1. Introduction Colloidal arrays of close-packed structures of monodisperse, sub-100 nm spherical silica nanoparticles are important in various applications including as templates for meso-porous carbon [1], polymers with ordered structures [2], and planar wave-guide materials [3]. Such arrays are also model systems for the study of fundamental phenomena such as molecular adsorption [4] and mass transport in mesopores [5–8]. Despite their potential importance, however, little systematic understanding of the variables that control self-assembly of sub-100 nm particles into ordered, three-dimensional arrays is currently in hand. Nonetheless, the elusive challenge of creating such arrays has captured the attention of several research groups recently resulting in a menu of recipes for creating arrays from nanoparticles of specific sizes without due attention to variables that might be systematically
∗ Corresponding author. Tel.: +1 520 621 8245; fax: +1 520 621 8248. E-mail address: [email protected] (J.E. Pemberton). 1 Current address: Department of Chemistry, University of Minnesota, Minneapolis, MN 55455, United States. 0927-7757/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfa.2010.12.021
exploited to create arrays from particles in the sub-100 nm size regime. Many methods have been reported for the fabrication of ordered colloidal arrays of nanoparticles [9]. Although several papers report fabrication of silica colloid arrays using direct pressures greater than several MPa [1,2], most methods involve self-assembly of particles due to the large area of ordered packing and control of array thickness that can be achieved [9–24]. It is generally believed that in addition to van der Waals interactions, the self-assembly behavior of silica particles is affected by electrostatic interactions and hydrogen bonding between particles [3,25–29]. Among the selfassembly methods for making silica nanoparticle colloidal arrays, one simple method for fabrication is sedimentation [10], but this method is relatively slow and usually takes weeks for completion of a single sample [11]. Wang and Gu [12] reported a method for fabrication of closepacked particle arrays through manipulation of silica nanoparticle hydrophobicity. This strategy requires particle surface modification and must be carefully controlled in order to obtain appropriate surface properties for self-assembly. The Langmuir–Blodgett method has also been used for formation of both monolayers and multilayers of silica spheres [13]. This method also requires particle surface
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
modification, and in addition, multilayer structures are obtained that are not well ordered. Okuba and co-workers introduced a rapid method for forming large area monolayers of uniform silica particles ranging in size from 25 to 100 nm using a simple wet coating process in which a suspension of silica nanoparticles is deposited onto a substrate and allowed to dry [14,15]. Hexagonal close-packed domains consisting of tens of nanoparticles can be obtained with this approach, but the arrays still contain point and line defects between these domains and do not possess long-range registry. Self-assembly by solvent evaporation is based on attractive capillary forces operative during solvent evaporation. This method is simple, relatively fast, and can be used for particles over a wide range of sizes [16–26]. Although numerous studies describe successful self-assembly of particles >200 nm [3,16–20], relatively few reports describe self-assembly of three-dimensional colloidal arrays of silica particles <100 nm [21–23]. Indeed, to date, no systematic study of sub-100 nm silica nanoparticle self-assembly into three-dimensional structures has been undertaken. Successful self-assembly of particles by solvent evaporation requires a monodisperse size distribution. According to Jiang et al. [24], self-assembly of three-dimensional silica nanoparticle arrays possessing long-range order requires particle size relative standard deviations (RSD) <8%. However, Wang et al. [3] demonstrated that self-assembled, two-dimensional, sub-100 nm silica particle monolayers possessing a significant fraction of close-packed hexagonal structures could still be observed using 60 nm particles with a RSD of 13%. In contrast, arrays from 40 nm particles with a RSD of 32% were not well packed [3]. Micheletto’s results confirm that packing order is greatly affected by particle uniformity [16]. An alternate method for fabrication of uniform silica particles in the sub-100 nm range is the reverse micelle method, which can be used to make very uniform particles with perfect spherical shapes in the range of 30–70 nm [30,31]. The RSDs of the particles can be controlled to within 4% [30,31], which would appear to satisfy the uniformity requirement [3,16,24] for successful self-assembly of ordered, close-packed structures over large areas. Significantly, however, systematic study of the self-assembly behavior of such particles has not been reported. Indeed, three-dimensional, closepacked structures consisting of nanoparticles made by the reverse micelle method have only been successfully formed as pellets requiring high pressure [2]. The lack of successful self-assembly of these particles into ordered, three-dimensional arrays despite their excellent monodispersity makes them interesting systems to compare with particles formed by the Stöber method. The work reported here involves self-assembly of sub-100 nm silica particles into three- and two-dimensional arrays. Using a fast (<2 h) batch Stöber method with reaction conditions optimized by a protocol developed previously in this laboratory, monodisperse, spherical silica nanoparticles with controlled size in the range from 50 to 120 nm with RSD < 9.5% can be produced [32]. This strategy provides large quantities of particles of sufficient quality for systematic study of their macroscopic self-assembly behavior. The effects of particle properties, including particle size distribution, fabrication method, evaporation temperature, and particle concentration, on the packing quality of the resulting arrays are discussed. To the best of the authors’ knowledge, the effects of these variables on the self-assembly of nanoparticles in the sub-100 nm size range has not been previously reported. This work seeks to answer two questions: First, do sub100 nm particles exhibit the same self-assembly behavior as larger (>200 nm) particles? Secondly, how do operational parameters associated with the self-assembly process affect packing quality of the resulting arrays? Answers to these questions are obviously critical for applications and technologies requiring self-assembly of silica nanoparticles in this size range. Finally, array formation
77
using a fast self-assembly method based on horizontal evaporation is discussed. This method not only provides a rapid protocol for self-assembly as an alternative to the widely used vertical evaporation method, but also provides new insights into the self-assembly behavior of sub-100 nm spherical silica nanoparticles. 2. Materials and methods 2.1. Fabrication and purification of silica particles An optimization protocol for synthesis of sub-100 nm silica particles by the Stöber method has been described recently from this laboratory [32]. With the exception of a single study in which packing of particles made by this optimized Stöber procedure were compared with packing of particles made by a conventional Stöber approach, all particles herein designated as Stöber particles were fabricated by this optimized method. After this synthesis, the particles so obtained were washed with ethanol with typically 4 cycles of centrifugation at 8000 rpm followed by ultrasonic redispersion in ethanol, and then finally sintered at 600 ◦ C for 4 h prior to selfassembly. A modified reverse micelle method based on those proposed by Arriagada and Osseo-Asare [30,31] was used to scale reactions from 5 to 50 mL to produce silica nanoparticles in sufficient quantity for self-assembly. Specifically, 47 mL hexane, 1 mL ammonium hydroxide and 4.31 mL polyoxyethylene (5) nonylphenyl ether (C9 H19 -C6 H4 -(OCH2 CH2 )n OH (n ∼ = 5, NP-5, Sigma–Aldrich) were transferred to a 120 mL glass bottle and mixed well by mechanical stirring. Then, 1 mL TEOS was added to the mixture and stirred at moderate rate for 3 min. The system was then allowed to react for 24 h without stirring. By this procedure, a suspension containing silica colloids of ∼50 nm dia was obtained. After synthesis, the silica nanoparticles were purified by centrifugation and ultrasonic dispersion in acetone, ethanol and water to remove the surfactant and other unreacted materials. 2.2. Self-assembly by vertical evaporation Silica nanoparticles were self-assembled using the vertical evaporation procedure reported by Jiang et al. [24] In this procedure, 0.005–0.05 g (0.05–0.5 wt%) of purified silica particles were dispersed by sonication in 10 mL ethanol in a clean scintillation vial. A clean microslide (1 × 10 cm, Fisher Scientific) was then placed in a vertical position in the scintillation vial for thin film development. The scintillation vial and microslide were covered by a 1200 mL crystallizing dish to eliminate external airflow and contamination [24]. To evaluate particle concentration effects on packing quality, Stöber silica particle suspensions with the same particle size (53 nm ± 9.5%) but different weight percentages (0.25%, 0.3%, and 0.4%) were used. All samples were placed under the same crystallizing dish to ensure identical ethanol evaporation rates. Films were allowed to develop at room temperature, measured to be 22 ◦ C. Stöber silica particle suspensions of the same particle size (53 nm ± 9.5%) with a 0.2 wt% particle percentage were used to study the effect of temperature on packing quality. These suspensions were developed at different temperatures (22 ◦ C, 40 ◦ C, and 50 ◦ C) while keeping all other conditions the same. 2.3. Characterization of silica colloidal arrays The resulting silica particle thin films were characterized by field emission scanning electron microscopy (SEM, Hitachi S-4800 and S-4500) in the backscattering mode after coating with a thin layer of Au–Pd. Images were acquired using an electron accelerating voltage of 15 kV. Cross sectional images were collected at 45◦ . Fast Fourier
78
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
Fig. 1. SEM images (top view) of self-assembled structures made by different methods: (a) array from 0.2 wt% ethanol suspension of 53 nm ± 9.5% particles made by the optimized Stöber method; (b) array from 0.2 wt% ethanol suspension of 58 nm ± 18.7% particles made by the standard Stöber method in ethanol at room temperature; (c) array from 0.15 wt% ethanol suspension of 55 nm ± 4% particles made by the reverse micelle method.
transformation of the SEM images was accomplished using Image-J (National Institutes of Health). 3. Results and discussion 3.1. Effect of particle properties on packing quality The self-assembly behavior of three nanoparticle samples made by different methods with different size distributions but approximately the same particle size was investigated. These particles were self-assembled on glass micro-slides using the vertical evaporation method described above, and the results are shown in Fig. 1. The sample whose image is shown in Fig. 1a was synthesized by an optimized Stöber method [32] resulting in a narrow size distribution (53 nm ± 9.5%). The sample whose image is shown in Fig. 1b was synthesized in ethanol at room temperature (58 nm ± 18.7%) using a non-optimized Stöber method. Particles for the image in Fig. 1c were made by the reverse micelle method (55 nm ± 4%). These images clearly document different morphologies and packing quality for the resulting films even though the particle sizes are approximately the same. Fig. 1a exhibits close-packed structures over a large area while the image in Fig. 1b only exhibits closepacked structures over small areas. In contrast to the close-packed structures formed by these particles, Fig. 1c shows poorly packed structures with small aggregated structures distributed throughout the film. The different packing quality of the samples in Fig. 1a and b must be due to the different size distributions of the particles, with more uniform particles forming close-packed structures over larger areas. This observation is consistent with results from other studies on the self-assembly of nanoparticles [3,16,24]. However, the images in Fig. 1a and c show the opposite trend: the more uniform, perfectly spherical particles made by the reverse micelle method exhibit much worse packing than the particles made by an optimized Stöber method despite their more favorable monodispersity and spherical shape. Indeed, to the best of the author’s knowledge, ordered arrays of this size scale have not been successfully produced by self-assembly using silica nanoparticles fabricated by the reverse micelle method. This result indicates that particle uniformity is not the only variable affecting self-assembly. Factors [33–36] such as particle surface chemistry, porosity, and density also contribute to the quality of the self-assembly process. Cortalezzi et al. [35] showed that by controlling properties of the suspension solvent system, and hence, surface chemistry of the silica particles, either well-ordered arrays or disordered aggregates of silica colloids were obtained. Organic and low ionic strength solvent systems result in ordered arrays. In contrast, high ionic strength aqueous suspensions lead to disorganized aggregate structures, presumably due to screen-
ing of the repulsive interactions between particles. Additionally, high ionic strength reduces the magnitude of the particle surface zeta potential. Both effects reduce electrostatic repulsive interactions between particles, leading to aggregation of particles during self-assembly. Wang et al. [36] studied the self-assembly of silica particles synthesized in the presence of surfactants at low concentration. Their results showed that anionic surfactants greatly improve packing quality while non-ionic and cationic surfactants lead to poor packing quality. These observations were rationalized in terms of the change in surface charge density caused by surfactant adsorption on the silica nanoparticles. Anionic surfactants increase the negative charge density on the particle surface, which reduces particle aggregation and results in good packing. In contrast, cationic and non-ionic surfactants decrease the magnitude of the particle surface zeta potential by screening the negative charge, leading to poor packing quality [36]. It should be noted that non-ionic surfactants were used for the synthesis of particles made by the reverse-micelle method in the work reported here. Even after extensive cleaning, residual surfactant molecules must still be strongly adsorbed on the particle surfaces, causing poor packing. In addition to particle surface chemistry, the density of the silica nanoparticles (i.e. particle porosity) should also affect packing quality. For two suspensions containing the same silica weight percent and particle diameter, but in particles with different intraparticle porosities, it is obvious that the suspension containing more porous particles will contain more particles per unit volume (i.e. a higher particle number density.). A high particle number density in suspension is unfavorable for high quality packing for several reasons. First, the average distance between particles is smaller and the total surface area of the particles is large. Consequently, particles have more opportunity to interact, leading to particle aggregation. Secondly, at a given temperature, the number of the particles carried to the three-phase line per unit time (particle flux) is proportional to the particle number density in the suspension [18]. Therefore, at a given evaporation rate as determined by evaporation temperature, suspensions with a higher particle number density will result in greater particle flux, making it more difficult for the particles to find appropriate lattice positions for perfect packing and resulting in more defects in the resulting array [37]. In fact, studies based on >200 nm silica particles did demonstrate a significant degradation of packing quality when particle concentration was higher than 4% (v/v) [24]. Therefore, detailed characterization of particles made by different methods, including surface zeta potential [35], amount of adsorbed surfactant, density, and intraparticle porosity, is necessary for further investigation of the self-assembly behavior of particles made by different methods. More importantly, these results also imply that detailed information about particle fabrication method is necessary in order to understand the resulting
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
79
Fig. 2. SEM images (side view) of close-packed three-dimensional structures made from 0.2 wt% Stöber particle suspensions with particle sizes of: (a) 81 nm ± 9.5%; (b) 67 nm ± 9.8%; (c) 53 nm ± 9.5%; (d) 38 nm ± 12.8%.
self-assembly behavior of a given batch of particles, which has not been explicitly recognized in previous work [14,15]. 3.2. Three-dimensional arrays from sub-100 nm particles Fig. 2 shows side views of close-packed structures formed at room temperature from several silica nanoparticle samples with relatively narrow size distributions. These images document close packing over multiple tightly packed layers. To further evaluate packing of the colloidal arrays, the top-view SEM images of four colloidal arrays made from silica particles between 50 and 250 nm were evaluated by fast Fourier transformation (FFT). The top-view SEM images and corresponding FFT images are shown in Figs. 3 and 4, respectively. All images in Fig. 4 show hexagonal close-packed patterns, although with a decrease in particle size, the long-range order of hexagonal packing decreases, as indicated by an increasing diffuseness of the FFT image. This effect indicates poorer packing due to an increase in size RSD with a decrease in particle size. 3.3. Effect of temperature on packing quality Although ordered colloidal arrays can be formed by using vertical evaporation at room temperature as shown in Figs. 2–4, this process usually takes several days to complete. One approach to increasing the rate of assembly is to increase the temperature during vertical evaporation [33,37,38]. Increasing temperature has also been reported as necessary to counter-balance gravity in the selfassembly of large silica particles (800 nm) into three-dimensional arrays [39]. Ye et al. [37] studied the effects of temperature on the packing quality of 310 nm polystyrene particles within the range of 45–65 ◦ C using 0.5 wt % suspensions. Their results show that the best packing quality was obtained at 55 ◦ C, while at 45 and
65 ◦ C, structures with poorer quality were obtained. These authors ascribed this to the balance between two opposing forces on packing quality. At higher temperature, particles have more kinetic energy to explore possible lattice sites for the lowest energy positions. Therefore, defects such as vacancies, dislocations and plane stacking faults were reduced, and structures with high packing quality were obtained when the temperature was increased from 45 to 55 ◦ C. However, a further increase in temperature results in a greater particle flux and faster array formation. As a result, particles have insufficient time to shift to the lowest energy lattice sites because of more rapid array formation, leading to greater defect formation [37]. SEM images from colloidal arrays formed by self-assembly of 53 nm Stöber silica nanoparticles at different evaporation temperatures along with their fast Fourier transforms are shown in Fig. 5. These images document that colloidal arrays of poorer packing quality are formed when temperature is increased above 40 ◦ C relative to those formed at room temperature (22 ◦ C, see Fig. 3c). This result is different from the behavior observed by Ye et al. who report an optimum temperature for packing of 45 ◦ C for 310 nm polystyrene particles. After considering the density difference between silica (2.04 g/mL) [24] and polystyrene (1.04 g/mL) [40], one can determine that the particle number density of our silica nanoparticle suspensions is more than 50 times greater than that used by Ye. In our experiments at room temperature (22 ◦ C), the particle flux is not large and array growth is relatively slow. Therefore, particles have enough time to shift to the most favorable lattice positions for high quality packing. With an increase in temperature, however, both particle flux and rate of array growth increase, making it more difficult for particles to shift to the most favorable lattice positions [37]. When such effects dominate over the favorable effect for ordered packing of the increased kinetic energy at elevated temperature, poor packing quality will result [37]. Since particle flux should increase monotonically with parti-
80
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
Fig. 3. SEM images (top view) of close-packed three-dimensional structures made from Stöber particle suspensions of size: (a) 250 nm ± 5.0% at 1.0 wt%; (b) 120 nm ± 6.8% at 0.5 wt%; (c) 81 nm ± 9.5% at 0.2 wt%; (d) 53 nm ± 9.5% at 0.2 wt%.
cle number density [18], the change in particle flux as a function of temperature is also proportional to particle number density. Due to the much higher particle number density used in our experiments, particle flux increases much faster with temperature, and hence, greater temperature effects on packing quality are observed in our experiments. As a result, packing quality deteriorates at a lower temperature in our experiments. This result indicates that although increasing temperature might be an effective way to improve packing quality and packing speed for large particles, such a strategy is considerably more limited for particles as small as 50 nm as used in our experiments. 3.4. Effect of particle concentration on packing quality In addition to evaporation temperature, particle concentration in the assembly suspension also affects packing quality. Fig. 6 shows top-view images of arrays formed by self-assembly of 53 nm particles from suspensions with particle wt% varying from 0.25 to 0.4 wt % along with their fast Fourier transforms. These films were developed at room temperature (22 ◦ C), as this temperature was shown to be optimal according to the experiments described above. Comparison of these images to that in Fig. 3c indicates that, with an increase in particle concentration, packing quality deteriorates. This trend is consistent with previous work using larger particles [24]. Significantly, the particle concentration regime in which deleterious effects on packing quality are observed for larger (e.g. ∼300 nm) silica particles is almost an order of magnitude higher than for particles in the 50 nm size range [24]. The much lower particle concentration at which packing quality deteriorates with 50 nm particles is also attributed to the higher particle number density, which leads to greater particle flux and more rapid array formation. The much lower particle concentration that can be used for high-quality self-assembly of small particles seriously restricts the maximum thickness of the arrays that can be obtained with these particles using vertical evaporation.
3.5. Self-assembly of monolayer structures Finally, fabrication of monolayers of these nanoparticles by vertical evaporation from very dilute solutions was investigated. To the best of our knowledge, the successful self-assembly of high surface coverage, homogenous, close-packed monolayers from particles with sub-100 nm diameters has not been heretofore reported and is challenging. Fig. 7a shows monolayer structures formed by vertical evaporation from a 0.05 wt % ethanol sol of 38 nm silica particles. With this decrease in particle concentration, well-packed monolayer structures are formed. Fig. 7b is a higher-magnification image of one edge of an isolated area of this monolayer. This image shows that well-ordered monolayers are formed over regions spanning several m in area. However, these are not perfect monolayers as evidenced by the observation of a partial second layer of loosely packed particles on parts of the monolayer (see arrows in Fig. 7d). The discontinuous nature of the monolayer and the formation of regions of multilayer structures were also observed by Micheletto et al. [16] and Wang et al. [3] for monolayers formed from sub100 nm particles by solvent evaporation. Fig. 7c shows a high magnification top-view image of the same monolayer. This image shows that, even though some particles are not strictly spherical, close-packing over relatively large areas is still observed. This observation is consistent with other results that show that monodispersity is more important than particle shape irregularity for large-area hexagonal packing of monolayers [3]. Fig. 7e shows a close-up view of the monolayer edge. From this image, formation of only a monolayer from these 38 nm silica nanoparticles is confirmed. 3.6. Fast self-assembly by horizontal evaporation As pointed out by Jiang and McFarland [41], since most current techniques for making colloidal crystalline arrays are only for low
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
81
Fig. 4. Fast Fourier transforms of the full top-view SEM images of close-packed three-dimensional structures formed from Stöber particles shown in Fig. 3 with sizes of: (a) 250 nm; (b) 120 nm; (c) 81 nm; (d) 53 nm.
volume, laboratory-scale production and take days to weeks, they are not amenable to scale-up for industrial-scale mass-fabrication, which requires fast and simple procedures. Restrictions of these approaches thus also impact mass fabrication of other materials, such as macroporous polymers and polymeric nanocomposites that use colloidal crystals as structural scaffolds [2,41–43]. The restrictions in both array formation rate and array thickness imposed by the vertical evaporation approach used here inspired us to pursue a simpler and more rapid self-assembly method for production of large-scale arrays from sub-100 nm silica particles. Horizontal evaporation is an alternate procedure that takes only several hours and has been shown to result in close-packed monolayer films of high quality [3,16,20]. However, large area closepacked three-dimensional arrays of more than eight layers have not been successfully fabricated using this approach. In fact, even for monolayer fabrication, surface coverage of only 60–85% is typically obtained by horizontal evaporation for sub-100 nm particles [3,16]. In our hands, initial experiments on horizontal evaporation of purified sub-100 nm silica particles from pure water or ethanol produced only randomly packed multilayer structures. This is consistent with reported results for particle self-assembly on the micron scale using the horizontal evaporation method [20]. To our surprise, however, closely packed multilayer structures were observed by horizontal evaporation when the original reaction solutions of unpurified 53-nm silica particle suspensions made by the Stöber method were used.
The experiments were performed as follows: after the Stöber reaction process is complete, 2 drops of the original reaction suspension are transferred onto a clean, 1 cm2 silicon wafer using a plastic pipet and the ethanol is rapidly (<30 s) evaporated with a gentle stream of N2 gas. An image of the resulting film (Fig. 8a) shows that over small areas, particles are well packed. However, in contrast to films formed by vertical evaporation, the low magnification image in Fig. 8c shows that films formed by this approach are not homogeneous and consist of well-packed regions of ∼5 m × 10 m in size separated by observable cracks. The close-packed local structure of each region implies that a slow evaporation rate may indeed be critical for forming homogeneous crystalline films over large areas. However, under our experimental conditions, self-assembly of these particles is a fast process, provided that the particles have a sufficiently narrow size distribution. In fact, such fast self-assembly of sub-100 nm particles to close-packed local structure has been also observed by other groups [21,44,45]. To get uniform arrays over larger areas, cracks between the well-packed regions must be eliminated. This was achieved by appropriate reduction of the evaporation rate [34,42,43] by increasing ambient vapor pressure of the ethanol solvent during drying. This was achieved by including a small vial containing ethanol in the evaporation chamber. The system was then allowed to evaporate for 2–3 h and the resulting films characterized by SEM.
82
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
Fig. 5. SEM images (a and c, top view) and full-image fast Fourier transforms (b and d) of three-dimensional structures made from 0.2 wt% suspensions of 53 nm ± 9.5% Stöber particles at 40 ◦ C (a and b) and 50 ◦ C (c and d).
Fig. 6. SEM images (a, c and e, top view) and full-image fast Fourier transforms (b, d and f) of three-dimensional structures made from 53 nm ± 9.5% Stöber particle suspensions of 0.25 wt% (a and b), 0.3 wt% (c and d), and 0.4 wt% (e and f).
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
83
Fig. 7. SEM images of two-dimensional monolayer structures formed from 0.05 wt% ethanol suspensions of 38 nm ± 12.8% Stöber particles. (a) 80 m × 60 m image; (b) 5 m × 3.7 m image; (c) 2 m × 1.5 m image; (d) 1 m × 0.7 m image; (e) close-up image of monolayer edge, 2.5 m × 2 m.
Images of one film produced in this manner are shown in Fig. 9. Fig. 9a and b is the top and side views of the array, respectively. Fig. 9a shows the closed-packed structure of the top layer. Fig. 9b shows a tightly packed layered structure in this film. Scanning along the edges of the film by SEM shows a uniform thickness of ∼2 m in
the film. Fig. 9c is the low magnification image of the film, showing a uniform film without any observable cracks on a mm length scale. In these experiments, the SEM images document smooth, flat arrays covering the entire substrate area. Significantly, these films are distinctly different from those produced by horizontal evapora-
84
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
Fig. 8. SEM images of close-packed arrays from 53 nm ± 9.5% Stöber particles by fast evaporation using a nitrogen gas stream. (a) High magnification image, 4 m × 3 m; (b) cross-section of image in (a) showing close-packing; (c) low magnification image, 24 m × 18 m; (d) FFT of the image in (a).
Fig. 9. SEM images of arrays from 53 nm ± 9.5% Stöber particles by horizontal fast self-assembly method: (a) top view; (b) side view; (c) low magnification image, 3 mm × 2 mm; (d) FFT of image of the image in (a).
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
tion reported earlier [19,20,25,26,46–48]. In previous attempts to create well-ordered arrays by horizontal evaporation, multilayer structures were observed only at the edges of the evaporation area with only a small number of particles remaining in the center of the drop evaporation region. Such an uneven particle distribution and the presence of free space in the film is a direct consequence of outward solvent flow to replenish evaporating solvent at the edges of the evaporation area that removes particles from the center region of the evaporating array [19,20,46–48]. Based on the absence of this uneven particle distribution in our films, one can conclude that such convective flow does not contribute significantly to particle packing during evaporation and that the different regions of the array remain homogenous during the evaporation process. In contrast to our results, we note that larger silica particles (∼300 nm) exhibit different horizontal self-assembly behavior [36]. According to Wang et al., when anionic surfactants were used in the synthesis of ∼300 nm silica nanoparticles, ordered, tightly packed three-dimensional structures were readily formed from the original Stöber reaction suspensions. This was attributed to the enhanced negative charges on anionic surfactant-adsorbed silica particles [36]. In contrast, however, for particles synthesized without surfactants or for particles synthesized with cationic or non-ionic surfactants, similar ordered self-assembled structures failed to form. On the basis of Wang’s results and the results presented here, several important conclusions can be drawn regarding the requirements for successful horizontal evaporation of ordered arrays of spherical silica nanoparticles. First, although enhanced electrostatic repulsive interactions are essential for self-assembly by horizontal evaporation for large particles (e.g. ∼300 nm), such enhanced repulsive interactions are not as necessary for maintaining the suspension integrity for the small particles used in our work when basic ammonia-containing reaction solutions are used for self-assembly. Secondly, Wang’s experiments show the importance of electrostatic repulsive interactions in the self-assembly of larger particles; our experiments on horizontal evaporation of small particles from water compared to the Stöber reaction solution also indicates that the negative charge that develops on the silica surfaces in the basic ammonia environment of the Stöber reaction solution is important. Both of these conclusions strongly suggest that the unique horizontal evaporation packing behavior of 53 nm particles observed here is due to the small size of the particles and the ammonia-ethanol solvent used in our Stöber reaction solution. The much smaller contribution from gravity for these small particles allows them to remain dispersed during the self-assembly period rendering sedimentation a minor contributor to array formation. In addition, due to electrostatic repulsive interactions between particles under basic conditions, particles may be farther away from each other in solution, forming ordered suspensions, with disruption of these ordered suspensions by gravity-induced sedimentation effectively prevented. In fact, ordered colloidal arrays of silica nanoparticles in aqueous solutions have been reported by many groups when strong electrostatic repulsive interactions are maintained [49–51]. It has also been reported that the evaporation of water from such colloidal array solutions can result in ordered, well-packed, dry colloidal arrays [52]. As further support for our explanation regarding the unique array formation mechanism for horizontal evaporation of these small particles, we note that our suspensions are stable for up to one week without phase separation caused by aggregation. Further studies, such as optical attenuation spectral and diffraction measurements [50,51] are needed to investigate the structures of the sub-100 nm silica particle suspensions used here to better understand the detailed mechanism of this unique, rapid self-assembly behavior.
85
In addition to ionizing silanol groups on the silica particle surfaces to strengthen electrostatic repulsion, as proposed by other groups [3,27,53], the presence of ammonium ions in solution also provides additional attractive interactions between particles, which consequently helps maintain an ordered sol structure during evaporation. During solvent evaporation, particles near the surface of three-dimensional arrays will be dewet and packed by lateral capillary forces. With further solvent evaporation, particles inside of the array will be sequentially packed by dewetting. Movement of the particles packed later during self-assembly will inevitably cause disruption of the structures already packed. It is noted that, although capillary forces can drag particles together for close packing, once the solvent is completely evaporated from the inter-particle pores, with the disappearance of the meniscus among the particles, the attraction between the particles caused by lateral capillary forces will accordingly disappear. The decrease in attractive forces between dry, close-packed particles increases the possibility of disruption of the three-dimensional structure by later-packed particles. This may be one reason why horizontal evaporation has only been successfully implemented for monolayer arrays [3,15–17]. The solvent composition used in the Stöber method (ammoniaethanol) is helpful in resisting this disruptive effect. It has been reported that in low dielectric solvents in the presence of ammonium ions, silica particles aggregate to form gels [27,53] presumably as the result of attractive interactions between ammonium ions and deprotonated silanols on the silica particles. Consequently, in the dry colloidal arrays, due to the low dielectric constant of air, ammonium cations will act as bridges to hold silica particles together, enhancing the attractive interactions between the particles, leading to robust, close-packed structures of silica particles that can effectively resist the disruptive effects caused by later-packed particles. Ammonium ions have been reported to strongly bind to silica particles synthesized by the Stöber method. Several wash cycles or even calcination at 550 ◦ C does not completely eliminate ammonia from the particles [54]. X-ray photoelectron spectroscopy results have also confirmed the existence of residual ammonium ions in silica particles made by the Stöber method after solvent is removed by rotary evaporation [3]. According to this proposed ammonium ion bridging mechanism, the enhanced attraction between particles is a short-range interaction [3]. As a result, attractive forces between particles are only enhanced when particles are dragged together by capillary forces. This is consistent with the observation that our silica particle suspensions are stable without aggregation. In addition, it is noted that, although it is apparent from the high magnification SEM images that the structures are close-packed over short distances, the long range order of the colloidal arrays made by the fast horizontal evaporation method is worse than that made by the vertical evaporation method. This can be clearly seen by comparison of Figs. 9d and 4d, which show the FFT images of the structures formed under vertical and horizontal evaporation conditions, respectively. Fig. 4d shows that the array formed from 53 nm particles using vertical evaporation has a hexagonal pattern, while Fig. 9d shows that the array formed by fast horizontal evaporation for the identical particles does not exhibit any clearly discernible pattern. This must be the result of the fact that, in fast horizontal evaporation, arrays are formed on a much shorter time scale than in the vertical method. Due to the much larger number of particles involved in self-assembly per unit time in horizontal evaporation, particles do not have sufficient time to find the lowest energy sites, and consequently, do not develop long range order. Finally, as shown in Fig. 9c, rather than ring-shaped films, a uniform film over several millimeters was formed in our horizontal evaporation method. This result shows that neither outward flow as reported by Deegan et al. [46–48] nor any pattern of convec-
86
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
tive flow due to the Marangoni effect [44,45] develops during array formation. Since our films are formed in a small confined space saturated with ethanol vapor, it is not surprising that uniform films are formed due to the slow evaporation rate and long evaporation time, similar to reports by Maillard et al. [44,45]. Acknowledgements The authors gratefully acknowledge support of this research by the Department of Energy (DE-FG03-95ER14546) and the National Science Foundation (CHE-0848624). References [1] Z.B. Lei, Y. Xiao, L.Q. Dang, M. Lu, W.S. You, Fabrication of ultra-large mesoporous carbon with tunable pore size by monodisperse silica particles derived from seed growth process, Micropor. Mesopor. Mater. 96 (2006) 127–134. [2] S.A. Johnson, P.J. Ollivier, T.E. Mallouk, Ordered mesoporous polymers of tunable pore size from colloidal silica templates, Science 283 (1999) 963–965. [3] C. Wang, Y.H. Zhang, L. Dong, L.M. Fu, Y.B. Bai, T.J. Li, J.G. Xu, Y. Wei, Twodimensional ordered arrays of silicon nanoparticles, Chem. Mater. 12 (2000) 3662–3666. [4] D. Rivera, P.E. Poston, R.H. Uibel, J.M. Harris, In situ adsorption studies at silica/solution interfaces by attenuated total internal reflection Fourier transform infrared spectroscopy: examination of adsorption models in normal phase liquid chromatography, Anal. Chem. 72 (2000) 1543–1554. [5] K.S. McCain, D.C. Hanley, J.M. Harris, Single-molecule fluorescence trajectories for investigating molecular transport in thin silica sol–gel films, Anal. Chem. 75 (2003) 4351–4359. [6] K.S. McCain, J.M. Harris, Total internal reflection fluorescence-correlation spectroscopy study of molecular transport in thin sol–gel films, Anal. Chem. 75 (2003) 3616–3624. [7] K.S. McCain, P. Schluesche, J.M. Harris, Poly(amidoamine) dendrimers as nanoscale diffusion probes in sol–gel films investigated by total internal reflection fluorescence spectroscopy, Anal. Chem. 76 (2004) 939–946. [8] D. Rivera, J.M. Harris, In situ ATR-FT-IR kinetic studies of molecular transport and surface binding in thin sol–gel films: Reactions of chlorosilane reagents in porous silica materials, Anal. Chem. 73 (2001) 411–423. [9] Y.N. Xia, B. Gates, Y.D. Yin, Y. Lu, Monodispersed colloidal spheres: old materials with new applications, Adv. Mater. 12 (2000) 693–713. [10] H. Miguez, F. Meseguer, C. Lopez, A. Blanco, J.S. Moya, J. Requena, A. Mifsud, V. Fornes, Control of the photonic crystal properties of fcc-packed submicrometer SiO2 spheres by sintering, Adv. Mater. 10 (1998) 480. [11] M.D. Sacks, T.Y. Tseng, Preparation of SiO2 glass from model powder compacts. 1. Formation and characterization of powders, suspensions, and green compacts, J. Am. Ceram. Soc. 67 (1984) 526–532. [12] W. Wang, B.H. Gu, Self-assembly of two- and three-dimensional particle arrays by manipulating the hydrophobicity of silica nanospheres, J. Phys. Chem. B 109 (2005) 22175–22180. [13] M. Szekeres, O. Kamalin, R.A. Schoonheydt, K. Wostyn, K. Clays, A. Persoons, I. Dekany, Ordering and optical properties of monolayers and multilayers of silica spheres deposited by the Langmuir–Blodgett method, J. Mater. Chem. 12 (2002) 3268–3274. [14] H. Nishikawa, K. Morozumi, M. Hu, T. Okubo, M. Fujita, Y. Yamaguchi, Effects of particle size on the monolayer structure of nanoparticles formed via a wetcoating process, J. Chem. Eng. Japan 38 (2005) 564–570. [15] M.H. Hu, S. Chujo, H. Nishikawa, Y. Yamaguchi, T. Okubo, Spontaneous formation of large-area monolayers of well-ordered nanoparticles via a wet-coating process, J. Nanopart. Res. 6 (2004) 479–487. [16] R. Micheletto, H. Fukuda, M. Ohtsu, A simple method for the production of a 2-dimensional, ordered array of small latex-particles, Langmuir 11 (1995) 3333–3336. [17] S Rakers, L.F. Chi, H. Fuchs, Influence of the evaporation rate on the packing order of polydisperse latex monofilms, Langmuir 13 (1997) 7121–7124. [18] A.S. Dimitrov, K. Nagayama, Continuous convective assembling of fine particles into two-dimensional arrays on solid surfaces, Langmuir 12 (1996) 1303–1311. [19] N.D. Denkov, O.D. Velev, P.A. Kralchevsky, I.B. Ivanov, H. Yoshimura, K. Nagayama, Two-dimensional crystallization, Nature 361 (1993) 26. [20] N.D. Denkov, O.D. Velev, P.A. Kralchevsky, I.B. Ivanov, H. Yoshimura, K. Nagayama, Mechanism of formation of two-dimensional crystals from latexparticles on substrates, Langmuir 8 (1992) 3183–3190. [21] K.D. Hartlen, A.P.T. Athanasopoulos, V. Kitaev, Facile preparation of highly monodisperse small silica spheres (15 to >200 nm) suitable for colloidal templating and formation of ordered arrays, Langmuir 24 (2008) 1714–1720. [22] T. Yokoi, Y. Sakamoto, O. Terasaki, Y. Kubota, T. Okubo, T. Tatsumi, Periodic arrangement of silica nanospheres assisted by amino acids, J. Am. Chem. Soc. 128 (2006) 13664–13665.
[23] M.A. Snyder, J.A. Lee, T.M. Davis, L.E. Scriven, M. Tsapatsis, Silica nanoparticle crystals and ordered coatings using lys-sil and a novel coating device, Langmuir 23 (2007) 9924–9928. [24] P. Jiang, J.F. Bertone, K.S. Hwang, V.L. Colvin, Single-crystal colloidal multilayers of controlled thickness, Chem. Mater. 11 (1999) 2132–2140. [25] J. Xu, J.F. Xia, Z.Q. Lin, Evaporation-induced self-assembly of nanoparticles from a sphere-on-flat geometry, Angew. Chem. Int. Ed. 46 (2007) 1860–1863. [26] J. Xu, J.F. Xia, S.W. Hong, Z.Q. Lin, F. Qiu, Y.L. Yang, Self-assembly of gradient concentric rings via solvent evaporation from a capillary bridge, Phys. Rev. Lett. 96 (2006) 066104. [27] J.C. Brinker, G.W. Scherer, Sol–Gel Science, San Diego, CA, Academic press, 1990. [28] R.K. Iler, The Chemistry of Silica, Wiley-Interscience, New York, 1979. [29] M. Kondo, K. Shinozaki, L. Bergstrom, N. Mizutani, Preparation of colloidal monolayers of alkoxylated silica particles at the air–liquid interface, Langmuir 11 (1995) 394–397. [30] F.J. Arriagada, K. Osseo-Asare, Synthesis of nanosize silica in a nonionic waterin-oil microemulsion: effects of the water/surfactant molar ratio and ammonia concentration, J. Colloid Interface Sci. 211 (1999) 210–220. [31] K. Osseoasare, F.J. Arriagada, Preparation of SiO2 nanoparticles in a nonionic reverse micellar system, Colloids Surf. 50 (1990) 321–339. [32] Y. Huang, J.E. Pemberton, Synthesis of uniform, spherical sub-100 nm silica particles using a conceptual modification of the classic LaMer model, Colloids Surf. A 360 (2010) 175–183. [33] L.M. Goldenberg, J. Wagner, J. Stumpe, B.R. Paulke, E. Gornitz, Ordered Arrays of large latex particles organized by vertical deposition, Langmuir 18 (2002) 3319–3323. [34] S. Veerapaneni, M.R. Wiesner, Particle deposition on an infinitely permeable surface -dependence of deposit morphology on particle size, J. Colloid Interface Sci. 162 (1994) 110–122. [35] M.M. Cortalezzi, V. Colvin, M.R. Wiesner, Controlling submicron particle template morphology: effect of solvent chemistry, J. Colloid Interface Sci. 283 (2005) 366–372. [36] W. Wang, B.H. Gu, L.Y. Liang, Effect of anionic surfactants on synthesis and selfassembly of silica colloidal nanoparticles, J. Colloid Interface Sci. 313 (2007) 169–173. [37] Y.H. Ye, F. LeBlanc, A. Hache, V.V. Truong, Self-assembling three-dimensional colloidal photonic crystal structure with high crystalline quality, Appl. Phys. Lett. 78 (2001) 52–54. [38] T. Van Le, E.E. Ross, T.R.C. Velarde, M.A. Legg, M.J. Wirth, Sintered silica colloidal crystals with fully hydroxylated surfaces, Langmuir 23 (2007) 8554– 8559. [39] Y.A. Vlasov, X.Z. Bo, J.C. Sturm, D.J. Norris, On-chip natural assembly of silicon photonic bandgap crystals, Nature 414 (2001) 289–293. [40] S.H. Im, Y.T. Lim, D.J. Suh, O.O. Park, Three-dimensional self-assembly of colloids at a water-air interface: a novel technique for the fabrication of photonic bandgap crystals, Adv. Mater. 14 (2002) 1367–1369. [41] P. Jiang, M.J. McFarland, Large-scale fabrication of wafer-size colloidal crystals, macroporous polymers and nanocomposites by spin-coating, J. Am. Chem. Soc. 126 (2004) 13778–13786. [42] P. Jiang, J.F. Bertone, V.L. Colvin, A lost-wax approach to monodisperse colloids and their crystals, Science 291 (2001) 453–457. [43] P. Jiang, K.S. Hwang, D.M. Mittleman, J.F. Bertone, V.L. Colvin, Template-directed preparation of macroporous polymers with oriented and crystalline arrays of voids, J. Am. Chem. Soc. 121 (1999) 11630–11637. [44] M. Maillard, L. Motte, A.T. Ngo, M.P. Pileni, Rings and hexagons made of nanocrystals: a Marangoni effect, J. Phys. Chem. B 104 (2000) 11871–11877. [45] M. Maillard, L. Motte, M.P. Pileni, Rings and hexagons made of nanocrystals, Adv. Mater. 13 (2001) 200–204. [46] R.D. Deegan, Pattern formation in drying drops, Phys. Rev. E 61 (2000) 475– 485. [47] R.D. Deegan, O. Bakajin, T.F. Dupont, G. Huber, S.R. Nagel, T.A. Witten, Capillary flow as the cause of ring stains from dried liquid drops, Nature 389 (1997) 827–829. [48] R.D. Deegan, O. Bakajin, T.F. Dupont, G. Huber, S.R. Nagel, T.A. Witten, Contact line deposits in an evaporating drop, Phys. Rev. E 62 (2000) 756–765. [49] R. Kesavamoorthy, S. Tandon, S. Xu, S. Jagannathan, S.A. Asher, Self-assembly and ordering of electrostatically stabilized silica suspensions, J. Colloid Interface Sci. 153 (1992) 188–198. [50] L. Liu, P.S. Li, S.A. Asher, Fortuitously superimposed lattice plane secondary diffraction from crystalline colloidal arrays, J. Am. Chem. Soc. 119 (1997) 2729–2732. [51] L. Liu, P.S. Li, S.A. Asher, Entropic trapping of macromolecules by mesoscopic periodic voids in a polymer hydrogel, Nature 397 (1999) 141–144. [52] W. Wang, B.H. Gu, L.Y. Liang, W.A. Hamilton, Fabrication of near-infrared photonic crystals using highly-monodispersed submicrometer SiO2 spheres, J. Phys. Chem. B 107 (2003) 12113–12117. [53] G.W. Scherer, J.C. Luong, Glasses from colloids, J. Non-Cryst. Solids 63 (1984) 163–172. [54] G. Herbert, Synthesis of monodispersed silica powders I. Particle properties and reaction kinetics, J. Eur. Ceram. Soc. 14 (1994) 189–204.
Contents lists available at ScienceDirect
Colloids and Surfaces A: Physicochemical and Engineering Aspects journal homepage: www.elsevier.com/locate/colsurfa
Fabrication of colloidal arrays by self-assembly of sub-100 nm silica particles Yuan Huang 1 , Jeanne E. Pemberton ∗ Department of Chemistry and Biochemistry, University of Arizona, 1306 East University Boulevard, Tucson, AZ 85721, United States
a r t i c l e
i n f o
Article history: Received 3 September 2010 Received in revised form 30 November 2010 Accepted 10 December 2010 Available online 23 December 2010 Keywords: Silica nanoparticles Colloidal array Self-assembly
a b s t r a c t Self-assembly of sub-100 nm spherical silica nanoparticles into ordered, tightly packed three- and twodimensional arrays was studied. Self-assembly by vertical evaporation was investigated for particles made by two methods: an optimized Stöber method recently reported from this laboratory and a modified reverse micelle method. Ordered, close-packed, two- and three-dimensional structures were formed with spherical nanoparticles made by the optimized Stöber method. Fast Fourier transforms of top-view scanning electron microscopy images document close-packed hexagonal packing for three-dimensional arrays consisting of particles as small as 50 nm. Ranges over which evaporation temperature and suspension particle concentration can be altered as strategies for improving packing quality of 50 nm particles have been defined. Self-assembly behavior that is distinct from that of larger particles (>200 nm) is observed for these sub-100 nm particles in that the ranges over which these variables can be altered to affect array packing quality are much smaller than for larger particles. In contrast, for particles made by the reverse micelle method, only structures with poor packing quality were obtained despite the fact that such particles are typically more uniform than those made by the Stöber method. These results provide clear evidence that, in addition to particle uniformity, other particle properties deriving from fabrication method play important roles in directing self-assembly of sub-100 nm particles. Finally, a rapid selfassembly method based on horizontal evaporation was used to produce close-packed three-dimensional structures of these sub-100 nm particles spanning several millimeters. Although these arrays are not as ordered as those made by vertical evaporation, the strategy reported herein allows tightly packed, crackfree arrays up to microns in thickness to be fabricated. A mechanism for self-assembly by this process is proposed. © 2011 Elsevier B.V. All rights reserved.
1. Introduction Colloidal arrays of close-packed structures of monodisperse, sub-100 nm spherical silica nanoparticles are important in various applications including as templates for meso-porous carbon [1], polymers with ordered structures [2], and planar wave-guide materials [3]. Such arrays are also model systems for the study of fundamental phenomena such as molecular adsorption [4] and mass transport in mesopores [5–8]. Despite their potential importance, however, little systematic understanding of the variables that control self-assembly of sub-100 nm particles into ordered, three-dimensional arrays is currently in hand. Nonetheless, the elusive challenge of creating such arrays has captured the attention of several research groups recently resulting in a menu of recipes for creating arrays from nanoparticles of specific sizes without due attention to variables that might be systematically
∗ Corresponding author. Tel.: +1 520 621 8245; fax: +1 520 621 8248. E-mail address: [email protected] (J.E. Pemberton). 1 Current address: Department of Chemistry, University of Minnesota, Minneapolis, MN 55455, United States. 0927-7757/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfa.2010.12.021
exploited to create arrays from particles in the sub-100 nm size regime. Many methods have been reported for the fabrication of ordered colloidal arrays of nanoparticles [9]. Although several papers report fabrication of silica colloid arrays using direct pressures greater than several MPa [1,2], most methods involve self-assembly of particles due to the large area of ordered packing and control of array thickness that can be achieved [9–24]. It is generally believed that in addition to van der Waals interactions, the self-assembly behavior of silica particles is affected by electrostatic interactions and hydrogen bonding between particles [3,25–29]. Among the selfassembly methods for making silica nanoparticle colloidal arrays, one simple method for fabrication is sedimentation [10], but this method is relatively slow and usually takes weeks for completion of a single sample [11]. Wang and Gu [12] reported a method for fabrication of closepacked particle arrays through manipulation of silica nanoparticle hydrophobicity. This strategy requires particle surface modification and must be carefully controlled in order to obtain appropriate surface properties for self-assembly. The Langmuir–Blodgett method has also been used for formation of both monolayers and multilayers of silica spheres [13]. This method also requires particle surface
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
modification, and in addition, multilayer structures are obtained that are not well ordered. Okuba and co-workers introduced a rapid method for forming large area monolayers of uniform silica particles ranging in size from 25 to 100 nm using a simple wet coating process in which a suspension of silica nanoparticles is deposited onto a substrate and allowed to dry [14,15]. Hexagonal close-packed domains consisting of tens of nanoparticles can be obtained with this approach, but the arrays still contain point and line defects between these domains and do not possess long-range registry. Self-assembly by solvent evaporation is based on attractive capillary forces operative during solvent evaporation. This method is simple, relatively fast, and can be used for particles over a wide range of sizes [16–26]. Although numerous studies describe successful self-assembly of particles >200 nm [3,16–20], relatively few reports describe self-assembly of three-dimensional colloidal arrays of silica particles <100 nm [21–23]. Indeed, to date, no systematic study of sub-100 nm silica nanoparticle self-assembly into three-dimensional structures has been undertaken. Successful self-assembly of particles by solvent evaporation requires a monodisperse size distribution. According to Jiang et al. [24], self-assembly of three-dimensional silica nanoparticle arrays possessing long-range order requires particle size relative standard deviations (RSD) <8%. However, Wang et al. [3] demonstrated that self-assembled, two-dimensional, sub-100 nm silica particle monolayers possessing a significant fraction of close-packed hexagonal structures could still be observed using 60 nm particles with a RSD of 13%. In contrast, arrays from 40 nm particles with a RSD of 32% were not well packed [3]. Micheletto’s results confirm that packing order is greatly affected by particle uniformity [16]. An alternate method for fabrication of uniform silica particles in the sub-100 nm range is the reverse micelle method, which can be used to make very uniform particles with perfect spherical shapes in the range of 30–70 nm [30,31]. The RSDs of the particles can be controlled to within 4% [30,31], which would appear to satisfy the uniformity requirement [3,16,24] for successful self-assembly of ordered, close-packed structures over large areas. Significantly, however, systematic study of the self-assembly behavior of such particles has not been reported. Indeed, three-dimensional, closepacked structures consisting of nanoparticles made by the reverse micelle method have only been successfully formed as pellets requiring high pressure [2]. The lack of successful self-assembly of these particles into ordered, three-dimensional arrays despite their excellent monodispersity makes them interesting systems to compare with particles formed by the Stöber method. The work reported here involves self-assembly of sub-100 nm silica particles into three- and two-dimensional arrays. Using a fast (<2 h) batch Stöber method with reaction conditions optimized by a protocol developed previously in this laboratory, monodisperse, spherical silica nanoparticles with controlled size in the range from 50 to 120 nm with RSD < 9.5% can be produced [32]. This strategy provides large quantities of particles of sufficient quality for systematic study of their macroscopic self-assembly behavior. The effects of particle properties, including particle size distribution, fabrication method, evaporation temperature, and particle concentration, on the packing quality of the resulting arrays are discussed. To the best of the authors’ knowledge, the effects of these variables on the self-assembly of nanoparticles in the sub-100 nm size range has not been previously reported. This work seeks to answer two questions: First, do sub100 nm particles exhibit the same self-assembly behavior as larger (>200 nm) particles? Secondly, how do operational parameters associated with the self-assembly process affect packing quality of the resulting arrays? Answers to these questions are obviously critical for applications and technologies requiring self-assembly of silica nanoparticles in this size range. Finally, array formation
77
using a fast self-assembly method based on horizontal evaporation is discussed. This method not only provides a rapid protocol for self-assembly as an alternative to the widely used vertical evaporation method, but also provides new insights into the self-assembly behavior of sub-100 nm spherical silica nanoparticles. 2. Materials and methods 2.1. Fabrication and purification of silica particles An optimization protocol for synthesis of sub-100 nm silica particles by the Stöber method has been described recently from this laboratory [32]. With the exception of a single study in which packing of particles made by this optimized Stöber procedure were compared with packing of particles made by a conventional Stöber approach, all particles herein designated as Stöber particles were fabricated by this optimized method. After this synthesis, the particles so obtained were washed with ethanol with typically 4 cycles of centrifugation at 8000 rpm followed by ultrasonic redispersion in ethanol, and then finally sintered at 600 ◦ C for 4 h prior to selfassembly. A modified reverse micelle method based on those proposed by Arriagada and Osseo-Asare [30,31] was used to scale reactions from 5 to 50 mL to produce silica nanoparticles in sufficient quantity for self-assembly. Specifically, 47 mL hexane, 1 mL ammonium hydroxide and 4.31 mL polyoxyethylene (5) nonylphenyl ether (C9 H19 -C6 H4 -(OCH2 CH2 )n OH (n ∼ = 5, NP-5, Sigma–Aldrich) were transferred to a 120 mL glass bottle and mixed well by mechanical stirring. Then, 1 mL TEOS was added to the mixture and stirred at moderate rate for 3 min. The system was then allowed to react for 24 h without stirring. By this procedure, a suspension containing silica colloids of ∼50 nm dia was obtained. After synthesis, the silica nanoparticles were purified by centrifugation and ultrasonic dispersion in acetone, ethanol and water to remove the surfactant and other unreacted materials. 2.2. Self-assembly by vertical evaporation Silica nanoparticles were self-assembled using the vertical evaporation procedure reported by Jiang et al. [24] In this procedure, 0.005–0.05 g (0.05–0.5 wt%) of purified silica particles were dispersed by sonication in 10 mL ethanol in a clean scintillation vial. A clean microslide (1 × 10 cm, Fisher Scientific) was then placed in a vertical position in the scintillation vial for thin film development. The scintillation vial and microslide were covered by a 1200 mL crystallizing dish to eliminate external airflow and contamination [24]. To evaluate particle concentration effects on packing quality, Stöber silica particle suspensions with the same particle size (53 nm ± 9.5%) but different weight percentages (0.25%, 0.3%, and 0.4%) were used. All samples were placed under the same crystallizing dish to ensure identical ethanol evaporation rates. Films were allowed to develop at room temperature, measured to be 22 ◦ C. Stöber silica particle suspensions of the same particle size (53 nm ± 9.5%) with a 0.2 wt% particle percentage were used to study the effect of temperature on packing quality. These suspensions were developed at different temperatures (22 ◦ C, 40 ◦ C, and 50 ◦ C) while keeping all other conditions the same. 2.3. Characterization of silica colloidal arrays The resulting silica particle thin films were characterized by field emission scanning electron microscopy (SEM, Hitachi S-4800 and S-4500) in the backscattering mode after coating with a thin layer of Au–Pd. Images were acquired using an electron accelerating voltage of 15 kV. Cross sectional images were collected at 45◦ . Fast Fourier
78
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
Fig. 1. SEM images (top view) of self-assembled structures made by different methods: (a) array from 0.2 wt% ethanol suspension of 53 nm ± 9.5% particles made by the optimized Stöber method; (b) array from 0.2 wt% ethanol suspension of 58 nm ± 18.7% particles made by the standard Stöber method in ethanol at room temperature; (c) array from 0.15 wt% ethanol suspension of 55 nm ± 4% particles made by the reverse micelle method.
transformation of the SEM images was accomplished using Image-J (National Institutes of Health). 3. Results and discussion 3.1. Effect of particle properties on packing quality The self-assembly behavior of three nanoparticle samples made by different methods with different size distributions but approximately the same particle size was investigated. These particles were self-assembled on glass micro-slides using the vertical evaporation method described above, and the results are shown in Fig. 1. The sample whose image is shown in Fig. 1a was synthesized by an optimized Stöber method [32] resulting in a narrow size distribution (53 nm ± 9.5%). The sample whose image is shown in Fig. 1b was synthesized in ethanol at room temperature (58 nm ± 18.7%) using a non-optimized Stöber method. Particles for the image in Fig. 1c were made by the reverse micelle method (55 nm ± 4%). These images clearly document different morphologies and packing quality for the resulting films even though the particle sizes are approximately the same. Fig. 1a exhibits close-packed structures over a large area while the image in Fig. 1b only exhibits closepacked structures over small areas. In contrast to the close-packed structures formed by these particles, Fig. 1c shows poorly packed structures with small aggregated structures distributed throughout the film. The different packing quality of the samples in Fig. 1a and b must be due to the different size distributions of the particles, with more uniform particles forming close-packed structures over larger areas. This observation is consistent with results from other studies on the self-assembly of nanoparticles [3,16,24]. However, the images in Fig. 1a and c show the opposite trend: the more uniform, perfectly spherical particles made by the reverse micelle method exhibit much worse packing than the particles made by an optimized Stöber method despite their more favorable monodispersity and spherical shape. Indeed, to the best of the author’s knowledge, ordered arrays of this size scale have not been successfully produced by self-assembly using silica nanoparticles fabricated by the reverse micelle method. This result indicates that particle uniformity is not the only variable affecting self-assembly. Factors [33–36] such as particle surface chemistry, porosity, and density also contribute to the quality of the self-assembly process. Cortalezzi et al. [35] showed that by controlling properties of the suspension solvent system, and hence, surface chemistry of the silica particles, either well-ordered arrays or disordered aggregates of silica colloids were obtained. Organic and low ionic strength solvent systems result in ordered arrays. In contrast, high ionic strength aqueous suspensions lead to disorganized aggregate structures, presumably due to screen-
ing of the repulsive interactions between particles. Additionally, high ionic strength reduces the magnitude of the particle surface zeta potential. Both effects reduce electrostatic repulsive interactions between particles, leading to aggregation of particles during self-assembly. Wang et al. [36] studied the self-assembly of silica particles synthesized in the presence of surfactants at low concentration. Their results showed that anionic surfactants greatly improve packing quality while non-ionic and cationic surfactants lead to poor packing quality. These observations were rationalized in terms of the change in surface charge density caused by surfactant adsorption on the silica nanoparticles. Anionic surfactants increase the negative charge density on the particle surface, which reduces particle aggregation and results in good packing. In contrast, cationic and non-ionic surfactants decrease the magnitude of the particle surface zeta potential by screening the negative charge, leading to poor packing quality [36]. It should be noted that non-ionic surfactants were used for the synthesis of particles made by the reverse-micelle method in the work reported here. Even after extensive cleaning, residual surfactant molecules must still be strongly adsorbed on the particle surfaces, causing poor packing. In addition to particle surface chemistry, the density of the silica nanoparticles (i.e. particle porosity) should also affect packing quality. For two suspensions containing the same silica weight percent and particle diameter, but in particles with different intraparticle porosities, it is obvious that the suspension containing more porous particles will contain more particles per unit volume (i.e. a higher particle number density.). A high particle number density in suspension is unfavorable for high quality packing for several reasons. First, the average distance between particles is smaller and the total surface area of the particles is large. Consequently, particles have more opportunity to interact, leading to particle aggregation. Secondly, at a given temperature, the number of the particles carried to the three-phase line per unit time (particle flux) is proportional to the particle number density in the suspension [18]. Therefore, at a given evaporation rate as determined by evaporation temperature, suspensions with a higher particle number density will result in greater particle flux, making it more difficult for the particles to find appropriate lattice positions for perfect packing and resulting in more defects in the resulting array [37]. In fact, studies based on >200 nm silica particles did demonstrate a significant degradation of packing quality when particle concentration was higher than 4% (v/v) [24]. Therefore, detailed characterization of particles made by different methods, including surface zeta potential [35], amount of adsorbed surfactant, density, and intraparticle porosity, is necessary for further investigation of the self-assembly behavior of particles made by different methods. More importantly, these results also imply that detailed information about particle fabrication method is necessary in order to understand the resulting
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
79
Fig. 2. SEM images (side view) of close-packed three-dimensional structures made from 0.2 wt% Stöber particle suspensions with particle sizes of: (a) 81 nm ± 9.5%; (b) 67 nm ± 9.8%; (c) 53 nm ± 9.5%; (d) 38 nm ± 12.8%.
self-assembly behavior of a given batch of particles, which has not been explicitly recognized in previous work [14,15]. 3.2. Three-dimensional arrays from sub-100 nm particles Fig. 2 shows side views of close-packed structures formed at room temperature from several silica nanoparticle samples with relatively narrow size distributions. These images document close packing over multiple tightly packed layers. To further evaluate packing of the colloidal arrays, the top-view SEM images of four colloidal arrays made from silica particles between 50 and 250 nm were evaluated by fast Fourier transformation (FFT). The top-view SEM images and corresponding FFT images are shown in Figs. 3 and 4, respectively. All images in Fig. 4 show hexagonal close-packed patterns, although with a decrease in particle size, the long-range order of hexagonal packing decreases, as indicated by an increasing diffuseness of the FFT image. This effect indicates poorer packing due to an increase in size RSD with a decrease in particle size. 3.3. Effect of temperature on packing quality Although ordered colloidal arrays can be formed by using vertical evaporation at room temperature as shown in Figs. 2–4, this process usually takes several days to complete. One approach to increasing the rate of assembly is to increase the temperature during vertical evaporation [33,37,38]. Increasing temperature has also been reported as necessary to counter-balance gravity in the selfassembly of large silica particles (800 nm) into three-dimensional arrays [39]. Ye et al. [37] studied the effects of temperature on the packing quality of 310 nm polystyrene particles within the range of 45–65 ◦ C using 0.5 wt % suspensions. Their results show that the best packing quality was obtained at 55 ◦ C, while at 45 and
65 ◦ C, structures with poorer quality were obtained. These authors ascribed this to the balance between two opposing forces on packing quality. At higher temperature, particles have more kinetic energy to explore possible lattice sites for the lowest energy positions. Therefore, defects such as vacancies, dislocations and plane stacking faults were reduced, and structures with high packing quality were obtained when the temperature was increased from 45 to 55 ◦ C. However, a further increase in temperature results in a greater particle flux and faster array formation. As a result, particles have insufficient time to shift to the lowest energy lattice sites because of more rapid array formation, leading to greater defect formation [37]. SEM images from colloidal arrays formed by self-assembly of 53 nm Stöber silica nanoparticles at different evaporation temperatures along with their fast Fourier transforms are shown in Fig. 5. These images document that colloidal arrays of poorer packing quality are formed when temperature is increased above 40 ◦ C relative to those formed at room temperature (22 ◦ C, see Fig. 3c). This result is different from the behavior observed by Ye et al. who report an optimum temperature for packing of 45 ◦ C for 310 nm polystyrene particles. After considering the density difference between silica (2.04 g/mL) [24] and polystyrene (1.04 g/mL) [40], one can determine that the particle number density of our silica nanoparticle suspensions is more than 50 times greater than that used by Ye. In our experiments at room temperature (22 ◦ C), the particle flux is not large and array growth is relatively slow. Therefore, particles have enough time to shift to the most favorable lattice positions for high quality packing. With an increase in temperature, however, both particle flux and rate of array growth increase, making it more difficult for particles to shift to the most favorable lattice positions [37]. When such effects dominate over the favorable effect for ordered packing of the increased kinetic energy at elevated temperature, poor packing quality will result [37]. Since particle flux should increase monotonically with parti-
80
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
Fig. 3. SEM images (top view) of close-packed three-dimensional structures made from Stöber particle suspensions of size: (a) 250 nm ± 5.0% at 1.0 wt%; (b) 120 nm ± 6.8% at 0.5 wt%; (c) 81 nm ± 9.5% at 0.2 wt%; (d) 53 nm ± 9.5% at 0.2 wt%.
cle number density [18], the change in particle flux as a function of temperature is also proportional to particle number density. Due to the much higher particle number density used in our experiments, particle flux increases much faster with temperature, and hence, greater temperature effects on packing quality are observed in our experiments. As a result, packing quality deteriorates at a lower temperature in our experiments. This result indicates that although increasing temperature might be an effective way to improve packing quality and packing speed for large particles, such a strategy is considerably more limited for particles as small as 50 nm as used in our experiments. 3.4. Effect of particle concentration on packing quality In addition to evaporation temperature, particle concentration in the assembly suspension also affects packing quality. Fig. 6 shows top-view images of arrays formed by self-assembly of 53 nm particles from suspensions with particle wt% varying from 0.25 to 0.4 wt % along with their fast Fourier transforms. These films were developed at room temperature (22 ◦ C), as this temperature was shown to be optimal according to the experiments described above. Comparison of these images to that in Fig. 3c indicates that, with an increase in particle concentration, packing quality deteriorates. This trend is consistent with previous work using larger particles [24]. Significantly, the particle concentration regime in which deleterious effects on packing quality are observed for larger (e.g. ∼300 nm) silica particles is almost an order of magnitude higher than for particles in the 50 nm size range [24]. The much lower particle concentration at which packing quality deteriorates with 50 nm particles is also attributed to the higher particle number density, which leads to greater particle flux and more rapid array formation. The much lower particle concentration that can be used for high-quality self-assembly of small particles seriously restricts the maximum thickness of the arrays that can be obtained with these particles using vertical evaporation.
3.5. Self-assembly of monolayer structures Finally, fabrication of monolayers of these nanoparticles by vertical evaporation from very dilute solutions was investigated. To the best of our knowledge, the successful self-assembly of high surface coverage, homogenous, close-packed monolayers from particles with sub-100 nm diameters has not been heretofore reported and is challenging. Fig. 7a shows monolayer structures formed by vertical evaporation from a 0.05 wt % ethanol sol of 38 nm silica particles. With this decrease in particle concentration, well-packed monolayer structures are formed. Fig. 7b is a higher-magnification image of one edge of an isolated area of this monolayer. This image shows that well-ordered monolayers are formed over regions spanning several m in area. However, these are not perfect monolayers as evidenced by the observation of a partial second layer of loosely packed particles on parts of the monolayer (see arrows in Fig. 7d). The discontinuous nature of the monolayer and the formation of regions of multilayer structures were also observed by Micheletto et al. [16] and Wang et al. [3] for monolayers formed from sub100 nm particles by solvent evaporation. Fig. 7c shows a high magnification top-view image of the same monolayer. This image shows that, even though some particles are not strictly spherical, close-packing over relatively large areas is still observed. This observation is consistent with other results that show that monodispersity is more important than particle shape irregularity for large-area hexagonal packing of monolayers [3]. Fig. 7e shows a close-up view of the monolayer edge. From this image, formation of only a monolayer from these 38 nm silica nanoparticles is confirmed. 3.6. Fast self-assembly by horizontal evaporation As pointed out by Jiang and McFarland [41], since most current techniques for making colloidal crystalline arrays are only for low
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
81
Fig. 4. Fast Fourier transforms of the full top-view SEM images of close-packed three-dimensional structures formed from Stöber particles shown in Fig. 3 with sizes of: (a) 250 nm; (b) 120 nm; (c) 81 nm; (d) 53 nm.
volume, laboratory-scale production and take days to weeks, they are not amenable to scale-up for industrial-scale mass-fabrication, which requires fast and simple procedures. Restrictions of these approaches thus also impact mass fabrication of other materials, such as macroporous polymers and polymeric nanocomposites that use colloidal crystals as structural scaffolds [2,41–43]. The restrictions in both array formation rate and array thickness imposed by the vertical evaporation approach used here inspired us to pursue a simpler and more rapid self-assembly method for production of large-scale arrays from sub-100 nm silica particles. Horizontal evaporation is an alternate procedure that takes only several hours and has been shown to result in close-packed monolayer films of high quality [3,16,20]. However, large area closepacked three-dimensional arrays of more than eight layers have not been successfully fabricated using this approach. In fact, even for monolayer fabrication, surface coverage of only 60–85% is typically obtained by horizontal evaporation for sub-100 nm particles [3,16]. In our hands, initial experiments on horizontal evaporation of purified sub-100 nm silica particles from pure water or ethanol produced only randomly packed multilayer structures. This is consistent with reported results for particle self-assembly on the micron scale using the horizontal evaporation method [20]. To our surprise, however, closely packed multilayer structures were observed by horizontal evaporation when the original reaction solutions of unpurified 53-nm silica particle suspensions made by the Stöber method were used.
The experiments were performed as follows: after the Stöber reaction process is complete, 2 drops of the original reaction suspension are transferred onto a clean, 1 cm2 silicon wafer using a plastic pipet and the ethanol is rapidly (<30 s) evaporated with a gentle stream of N2 gas. An image of the resulting film (Fig. 8a) shows that over small areas, particles are well packed. However, in contrast to films formed by vertical evaporation, the low magnification image in Fig. 8c shows that films formed by this approach are not homogeneous and consist of well-packed regions of ∼5 m × 10 m in size separated by observable cracks. The close-packed local structure of each region implies that a slow evaporation rate may indeed be critical for forming homogeneous crystalline films over large areas. However, under our experimental conditions, self-assembly of these particles is a fast process, provided that the particles have a sufficiently narrow size distribution. In fact, such fast self-assembly of sub-100 nm particles to close-packed local structure has been also observed by other groups [21,44,45]. To get uniform arrays over larger areas, cracks between the well-packed regions must be eliminated. This was achieved by appropriate reduction of the evaporation rate [34,42,43] by increasing ambient vapor pressure of the ethanol solvent during drying. This was achieved by including a small vial containing ethanol in the evaporation chamber. The system was then allowed to evaporate for 2–3 h and the resulting films characterized by SEM.
82
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
Fig. 5. SEM images (a and c, top view) and full-image fast Fourier transforms (b and d) of three-dimensional structures made from 0.2 wt% suspensions of 53 nm ± 9.5% Stöber particles at 40 ◦ C (a and b) and 50 ◦ C (c and d).
Fig. 6. SEM images (a, c and e, top view) and full-image fast Fourier transforms (b, d and f) of three-dimensional structures made from 53 nm ± 9.5% Stöber particle suspensions of 0.25 wt% (a and b), 0.3 wt% (c and d), and 0.4 wt% (e and f).
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
83
Fig. 7. SEM images of two-dimensional monolayer structures formed from 0.05 wt% ethanol suspensions of 38 nm ± 12.8% Stöber particles. (a) 80 m × 60 m image; (b) 5 m × 3.7 m image; (c) 2 m × 1.5 m image; (d) 1 m × 0.7 m image; (e) close-up image of monolayer edge, 2.5 m × 2 m.
Images of one film produced in this manner are shown in Fig. 9. Fig. 9a and b is the top and side views of the array, respectively. Fig. 9a shows the closed-packed structure of the top layer. Fig. 9b shows a tightly packed layered structure in this film. Scanning along the edges of the film by SEM shows a uniform thickness of ∼2 m in
the film. Fig. 9c is the low magnification image of the film, showing a uniform film without any observable cracks on a mm length scale. In these experiments, the SEM images document smooth, flat arrays covering the entire substrate area. Significantly, these films are distinctly different from those produced by horizontal evapora-
84
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
Fig. 8. SEM images of close-packed arrays from 53 nm ± 9.5% Stöber particles by fast evaporation using a nitrogen gas stream. (a) High magnification image, 4 m × 3 m; (b) cross-section of image in (a) showing close-packing; (c) low magnification image, 24 m × 18 m; (d) FFT of the image in (a).
Fig. 9. SEM images of arrays from 53 nm ± 9.5% Stöber particles by horizontal fast self-assembly method: (a) top view; (b) side view; (c) low magnification image, 3 mm × 2 mm; (d) FFT of image of the image in (a).
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
tion reported earlier [19,20,25,26,46–48]. In previous attempts to create well-ordered arrays by horizontal evaporation, multilayer structures were observed only at the edges of the evaporation area with only a small number of particles remaining in the center of the drop evaporation region. Such an uneven particle distribution and the presence of free space in the film is a direct consequence of outward solvent flow to replenish evaporating solvent at the edges of the evaporation area that removes particles from the center region of the evaporating array [19,20,46–48]. Based on the absence of this uneven particle distribution in our films, one can conclude that such convective flow does not contribute significantly to particle packing during evaporation and that the different regions of the array remain homogenous during the evaporation process. In contrast to our results, we note that larger silica particles (∼300 nm) exhibit different horizontal self-assembly behavior [36]. According to Wang et al., when anionic surfactants were used in the synthesis of ∼300 nm silica nanoparticles, ordered, tightly packed three-dimensional structures were readily formed from the original Stöber reaction suspensions. This was attributed to the enhanced negative charges on anionic surfactant-adsorbed silica particles [36]. In contrast, however, for particles synthesized without surfactants or for particles synthesized with cationic or non-ionic surfactants, similar ordered self-assembled structures failed to form. On the basis of Wang’s results and the results presented here, several important conclusions can be drawn regarding the requirements for successful horizontal evaporation of ordered arrays of spherical silica nanoparticles. First, although enhanced electrostatic repulsive interactions are essential for self-assembly by horizontal evaporation for large particles (e.g. ∼300 nm), such enhanced repulsive interactions are not as necessary for maintaining the suspension integrity for the small particles used in our work when basic ammonia-containing reaction solutions are used for self-assembly. Secondly, Wang’s experiments show the importance of electrostatic repulsive interactions in the self-assembly of larger particles; our experiments on horizontal evaporation of small particles from water compared to the Stöber reaction solution also indicates that the negative charge that develops on the silica surfaces in the basic ammonia environment of the Stöber reaction solution is important. Both of these conclusions strongly suggest that the unique horizontal evaporation packing behavior of 53 nm particles observed here is due to the small size of the particles and the ammonia-ethanol solvent used in our Stöber reaction solution. The much smaller contribution from gravity for these small particles allows them to remain dispersed during the self-assembly period rendering sedimentation a minor contributor to array formation. In addition, due to electrostatic repulsive interactions between particles under basic conditions, particles may be farther away from each other in solution, forming ordered suspensions, with disruption of these ordered suspensions by gravity-induced sedimentation effectively prevented. In fact, ordered colloidal arrays of silica nanoparticles in aqueous solutions have been reported by many groups when strong electrostatic repulsive interactions are maintained [49–51]. It has also been reported that the evaporation of water from such colloidal array solutions can result in ordered, well-packed, dry colloidal arrays [52]. As further support for our explanation regarding the unique array formation mechanism for horizontal evaporation of these small particles, we note that our suspensions are stable for up to one week without phase separation caused by aggregation. Further studies, such as optical attenuation spectral and diffraction measurements [50,51] are needed to investigate the structures of the sub-100 nm silica particle suspensions used here to better understand the detailed mechanism of this unique, rapid self-assembly behavior.
85
In addition to ionizing silanol groups on the silica particle surfaces to strengthen electrostatic repulsion, as proposed by other groups [3,27,53], the presence of ammonium ions in solution also provides additional attractive interactions between particles, which consequently helps maintain an ordered sol structure during evaporation. During solvent evaporation, particles near the surface of three-dimensional arrays will be dewet and packed by lateral capillary forces. With further solvent evaporation, particles inside of the array will be sequentially packed by dewetting. Movement of the particles packed later during self-assembly will inevitably cause disruption of the structures already packed. It is noted that, although capillary forces can drag particles together for close packing, once the solvent is completely evaporated from the inter-particle pores, with the disappearance of the meniscus among the particles, the attraction between the particles caused by lateral capillary forces will accordingly disappear. The decrease in attractive forces between dry, close-packed particles increases the possibility of disruption of the three-dimensional structure by later-packed particles. This may be one reason why horizontal evaporation has only been successfully implemented for monolayer arrays [3,15–17]. The solvent composition used in the Stöber method (ammoniaethanol) is helpful in resisting this disruptive effect. It has been reported that in low dielectric solvents in the presence of ammonium ions, silica particles aggregate to form gels [27,53] presumably as the result of attractive interactions between ammonium ions and deprotonated silanols on the silica particles. Consequently, in the dry colloidal arrays, due to the low dielectric constant of air, ammonium cations will act as bridges to hold silica particles together, enhancing the attractive interactions between the particles, leading to robust, close-packed structures of silica particles that can effectively resist the disruptive effects caused by later-packed particles. Ammonium ions have been reported to strongly bind to silica particles synthesized by the Stöber method. Several wash cycles or even calcination at 550 ◦ C does not completely eliminate ammonia from the particles [54]. X-ray photoelectron spectroscopy results have also confirmed the existence of residual ammonium ions in silica particles made by the Stöber method after solvent is removed by rotary evaporation [3]. According to this proposed ammonium ion bridging mechanism, the enhanced attraction between particles is a short-range interaction [3]. As a result, attractive forces between particles are only enhanced when particles are dragged together by capillary forces. This is consistent with the observation that our silica particle suspensions are stable without aggregation. In addition, it is noted that, although it is apparent from the high magnification SEM images that the structures are close-packed over short distances, the long range order of the colloidal arrays made by the fast horizontal evaporation method is worse than that made by the vertical evaporation method. This can be clearly seen by comparison of Figs. 9d and 4d, which show the FFT images of the structures formed under vertical and horizontal evaporation conditions, respectively. Fig. 4d shows that the array formed from 53 nm particles using vertical evaporation has a hexagonal pattern, while Fig. 9d shows that the array formed by fast horizontal evaporation for the identical particles does not exhibit any clearly discernible pattern. This must be the result of the fact that, in fast horizontal evaporation, arrays are formed on a much shorter time scale than in the vertical method. Due to the much larger number of particles involved in self-assembly per unit time in horizontal evaporation, particles do not have sufficient time to find the lowest energy sites, and consequently, do not develop long range order. Finally, as shown in Fig. 9c, rather than ring-shaped films, a uniform film over several millimeters was formed in our horizontal evaporation method. This result shows that neither outward flow as reported by Deegan et al. [46–48] nor any pattern of convec-
86
Y. Huang, J.E. Pemberton / Colloids and Surfaces A: Physicochem. Eng. Aspects 377 (2011) 76–86
tive flow due to the Marangoni effect [44,45] develops during array formation. Since our films are formed in a small confined space saturated with ethanol vapor, it is not surprising that uniform films are formed due to the slow evaporation rate and long evaporation time, similar to reports by Maillard et al. [44,45]. Acknowledgements The authors gratefully acknowledge support of this research by the Department of Energy (DE-FG03-95ER14546) and the National Science Foundation (CHE-0848624). References [1] Z.B. Lei, Y. Xiao, L.Q. Dang, M. Lu, W.S. You, Fabrication of ultra-large mesoporous carbon with tunable pore size by monodisperse silica particles derived from seed growth process, Micropor. Mesopor. Mater. 96 (2006) 127–134. [2] S.A. Johnson, P.J. Ollivier, T.E. Mallouk, Ordered mesoporous polymers of tunable pore size from colloidal silica templates, Science 283 (1999) 963–965. [3] C. Wang, Y.H. Zhang, L. Dong, L.M. Fu, Y.B. Bai, T.J. Li, J.G. Xu, Y. Wei, Twodimensional ordered arrays of silicon nanoparticles, Chem. Mater. 12 (2000) 3662–3666. [4] D. Rivera, P.E. Poston, R.H. Uibel, J.M. Harris, In situ adsorption studies at silica/solution interfaces by attenuated total internal reflection Fourier transform infrared spectroscopy: examination of adsorption models in normal phase liquid chromatography, Anal. Chem. 72 (2000) 1543–1554. [5] K.S. McCain, D.C. Hanley, J.M. Harris, Single-molecule fluorescence trajectories for investigating molecular transport in thin silica sol–gel films, Anal. Chem. 75 (2003) 4351–4359. [6] K.S. McCain, J.M. Harris, Total internal reflection fluorescence-correlation spectroscopy study of molecular transport in thin sol–gel films, Anal. Chem. 75 (2003) 3616–3624. [7] K.S. McCain, P. Schluesche, J.M. Harris, Poly(amidoamine) dendrimers as nanoscale diffusion probes in sol–gel films investigated by total internal reflection fluorescence spectroscopy, Anal. Chem. 76 (2004) 939–946. [8] D. Rivera, J.M. Harris, In situ ATR-FT-IR kinetic studies of molecular transport and surface binding in thin sol–gel films: Reactions of chlorosilane reagents in porous silica materials, Anal. Chem. 73 (2001) 411–423. [9] Y.N. Xia, B. Gates, Y.D. Yin, Y. Lu, Monodispersed colloidal spheres: old materials with new applications, Adv. Mater. 12 (2000) 693–713. [10] H. Miguez, F. Meseguer, C. Lopez, A. Blanco, J.S. Moya, J. Requena, A. Mifsud, V. Fornes, Control of the photonic crystal properties of fcc-packed submicrometer SiO2 spheres by sintering, Adv. Mater. 10 (1998) 480. [11] M.D. Sacks, T.Y. Tseng, Preparation of SiO2 glass from model powder compacts. 1. Formation and characterization of powders, suspensions, and green compacts, J. Am. Ceram. Soc. 67 (1984) 526–532. [12] W. Wang, B.H. Gu, Self-assembly of two- and three-dimensional particle arrays by manipulating the hydrophobicity of silica nanospheres, J. Phys. Chem. B 109 (2005) 22175–22180. [13] M. Szekeres, O. Kamalin, R.A. Schoonheydt, K. Wostyn, K. Clays, A. Persoons, I. Dekany, Ordering and optical properties of monolayers and multilayers of silica spheres deposited by the Langmuir–Blodgett method, J. Mater. Chem. 12 (2002) 3268–3274. [14] H. Nishikawa, K. Morozumi, M. Hu, T. Okubo, M. Fujita, Y. Yamaguchi, Effects of particle size on the monolayer structure of nanoparticles formed via a wetcoating process, J. Chem. Eng. Japan 38 (2005) 564–570. [15] M.H. Hu, S. Chujo, H. Nishikawa, Y. Yamaguchi, T. Okubo, Spontaneous formation of large-area monolayers of well-ordered nanoparticles via a wet-coating process, J. Nanopart. Res. 6 (2004) 479–487. [16] R. Micheletto, H. Fukuda, M. Ohtsu, A simple method for the production of a 2-dimensional, ordered array of small latex-particles, Langmuir 11 (1995) 3333–3336. [17] S Rakers, L.F. Chi, H. Fuchs, Influence of the evaporation rate on the packing order of polydisperse latex monofilms, Langmuir 13 (1997) 7121–7124. [18] A.S. Dimitrov, K. Nagayama, Continuous convective assembling of fine particles into two-dimensional arrays on solid surfaces, Langmuir 12 (1996) 1303–1311. [19] N.D. Denkov, O.D. Velev, P.A. Kralchevsky, I.B. Ivanov, H. Yoshimura, K. Nagayama, Two-dimensional crystallization, Nature 361 (1993) 26. [20] N.D. Denkov, O.D. Velev, P.A. Kralchevsky, I.B. Ivanov, H. Yoshimura, K. Nagayama, Mechanism of formation of two-dimensional crystals from latexparticles on substrates, Langmuir 8 (1992) 3183–3190. [21] K.D. Hartlen, A.P.T. Athanasopoulos, V. Kitaev, Facile preparation of highly monodisperse small silica spheres (15 to >200 nm) suitable for colloidal templating and formation of ordered arrays, Langmuir 24 (2008) 1714–1720. [22] T. Yokoi, Y. Sakamoto, O. Terasaki, Y. Kubota, T. Okubo, T. Tatsumi, Periodic arrangement of silica nanospheres assisted by amino acids, J. Am. Chem. Soc. 128 (2006) 13664–13665.
[23] M.A. Snyder, J.A. Lee, T.M. Davis, L.E. Scriven, M. Tsapatsis, Silica nanoparticle crystals and ordered coatings using lys-sil and a novel coating device, Langmuir 23 (2007) 9924–9928. [24] P. Jiang, J.F. Bertone, K.S. Hwang, V.L. Colvin, Single-crystal colloidal multilayers of controlled thickness, Chem. Mater. 11 (1999) 2132–2140. [25] J. Xu, J.F. Xia, Z.Q. Lin, Evaporation-induced self-assembly of nanoparticles from a sphere-on-flat geometry, Angew. Chem. Int. Ed. 46 (2007) 1860–1863. [26] J. Xu, J.F. Xia, S.W. Hong, Z.Q. Lin, F. Qiu, Y.L. Yang, Self-assembly of gradient concentric rings via solvent evaporation from a capillary bridge, Phys. Rev. Lett. 96 (2006) 066104. [27] J.C. Brinker, G.W. Scherer, Sol–Gel Science, San Diego, CA, Academic press, 1990. [28] R.K. Iler, The Chemistry of Silica, Wiley-Interscience, New York, 1979. [29] M. Kondo, K. Shinozaki, L. Bergstrom, N. Mizutani, Preparation of colloidal monolayers of alkoxylated silica particles at the air–liquid interface, Langmuir 11 (1995) 394–397. [30] F.J. Arriagada, K. Osseo-Asare, Synthesis of nanosize silica in a nonionic waterin-oil microemulsion: effects of the water/surfactant molar ratio and ammonia concentration, J. Colloid Interface Sci. 211 (1999) 210–220. [31] K. Osseoasare, F.J. Arriagada, Preparation of SiO2 nanoparticles in a nonionic reverse micellar system, Colloids Surf. 50 (1990) 321–339. [32] Y. Huang, J.E. Pemberton, Synthesis of uniform, spherical sub-100 nm silica particles using a conceptual modification of the classic LaMer model, Colloids Surf. A 360 (2010) 175–183. [33] L.M. Goldenberg, J. Wagner, J. Stumpe, B.R. Paulke, E. Gornitz, Ordered Arrays of large latex particles organized by vertical deposition, Langmuir 18 (2002) 3319–3323. [34] S. Veerapaneni, M.R. Wiesner, Particle deposition on an infinitely permeable surface -dependence of deposit morphology on particle size, J. Colloid Interface Sci. 162 (1994) 110–122. [35] M.M. Cortalezzi, V. Colvin, M.R. Wiesner, Controlling submicron particle template morphology: effect of solvent chemistry, J. Colloid Interface Sci. 283 (2005) 366–372. [36] W. Wang, B.H. Gu, L.Y. Liang, Effect of anionic surfactants on synthesis and selfassembly of silica colloidal nanoparticles, J. Colloid Interface Sci. 313 (2007) 169–173. [37] Y.H. Ye, F. LeBlanc, A. Hache, V.V. Truong, Self-assembling three-dimensional colloidal photonic crystal structure with high crystalline quality, Appl. Phys. Lett. 78 (2001) 52–54. [38] T. Van Le, E.E. Ross, T.R.C. Velarde, M.A. Legg, M.J. Wirth, Sintered silica colloidal crystals with fully hydroxylated surfaces, Langmuir 23 (2007) 8554– 8559. [39] Y.A. Vlasov, X.Z. Bo, J.C. Sturm, D.J. Norris, On-chip natural assembly of silicon photonic bandgap crystals, Nature 414 (2001) 289–293. [40] S.H. Im, Y.T. Lim, D.J. Suh, O.O. Park, Three-dimensional self-assembly of colloids at a water-air interface: a novel technique for the fabrication of photonic bandgap crystals, Adv. Mater. 14 (2002) 1367–1369. [41] P. Jiang, M.J. McFarland, Large-scale fabrication of wafer-size colloidal crystals, macroporous polymers and nanocomposites by spin-coating, J. Am. Chem. Soc. 126 (2004) 13778–13786. [42] P. Jiang, J.F. Bertone, V.L. Colvin, A lost-wax approach to monodisperse colloids and their crystals, Science 291 (2001) 453–457. [43] P. Jiang, K.S. Hwang, D.M. Mittleman, J.F. Bertone, V.L. Colvin, Template-directed preparation of macroporous polymers with oriented and crystalline arrays of voids, J. Am. Chem. Soc. 121 (1999) 11630–11637. [44] M. Maillard, L. Motte, A.T. Ngo, M.P. Pileni, Rings and hexagons made of nanocrystals: a Marangoni effect, J. Phys. Chem. B 104 (2000) 11871–11877. [45] M. Maillard, L. Motte, M.P. Pileni, Rings and hexagons made of nanocrystals, Adv. Mater. 13 (2001) 200–204. [46] R.D. Deegan, Pattern formation in drying drops, Phys. Rev. E 61 (2000) 475– 485. [47] R.D. Deegan, O. Bakajin, T.F. Dupont, G. Huber, S.R. Nagel, T.A. Witten, Capillary flow as the cause of ring stains from dried liquid drops, Nature 389 (1997) 827–829. [48] R.D. Deegan, O. Bakajin, T.F. Dupont, G. Huber, S.R. Nagel, T.A. Witten, Contact line deposits in an evaporating drop, Phys. Rev. E 62 (2000) 756–765. [49] R. Kesavamoorthy, S. Tandon, S. Xu, S. Jagannathan, S.A. Asher, Self-assembly and ordering of electrostatically stabilized silica suspensions, J. Colloid Interface Sci. 153 (1992) 188–198. [50] L. Liu, P.S. Li, S.A. Asher, Fortuitously superimposed lattice plane secondary diffraction from crystalline colloidal arrays, J. Am. Chem. Soc. 119 (1997) 2729–2732. [51] L. Liu, P.S. Li, S.A. Asher, Entropic trapping of macromolecules by mesoscopic periodic voids in a polymer hydrogel, Nature 397 (1999) 141–144. [52] W. Wang, B.H. Gu, L.Y. Liang, W.A. Hamilton, Fabrication of near-infrared photonic crystals using highly-monodispersed submicrometer SiO2 spheres, J. Phys. Chem. B 107 (2003) 12113–12117. [53] G.W. Scherer, J.C. Luong, Glasses from colloids, J. Non-Cryst. Solids 63 (1984) 163–172. [54] G. Herbert, Synthesis of monodispersed silica powders I. Particle properties and reaction kinetics, J. Eur. Ceram. Soc. 14 (1994) 189–204.