Article pubs.acs.org/Langmuir
Template-Assisted Synthesis of Janus Silica Nanobowls Florian Guignard and Marco Lattuada* Adolphe Merkle Institute, University of Fribourg, Chemin des Verdiers 4, CH-1700 Fribourg, Switzerland S Supporting Information *
ABSTRACT: The preparation of anisotropic nanoparticles has drawn much attention in the literature, with most of the efforts being dedicated to convex particles. In this work, instead, we present a reliable method to synthesis silica nanobowls with one well-defined opening, covering a broad range of sizes. The nanobowls have been obtained from asymmetrically functionalized silica−polymer Janus nanodumbbells, used as templates, by removing of the polymer. Polystyrene seeds having different sizes as well as surface chemistry have been used as starting material in a two-step seeded emulsion polymerization, which leads to polymer nanodumbbells. These dumbbells are also asymmetrically functionalized due to the presence of silane groups on only one of their two hemispheres. This allows us to selectively coat the silane-bearing hemisphere of the dumbbells with a silica layer by means of a Stoeber process. The silica nanobowls are eventually obtained after either calcination or dissolution of the polymeric template. Depending on the route followed to remove the polymer, nanobowls made of pure silica (from calcination) or hybrid Janus nanobowls with a silica outer shell and a covalently bound hydrophobic polymer layer inside the cavity (from dissolution) could be prepared. The difference between the two types of nanobowls has been proved by electrostatically binding oppositely charged silica nanoparticles, which adhere selectively only on the outer silica part of the nanobowls prepared by polymer dissolution, while they attach both inside and outside of nanobowls prepared by calcination. We also show that selective functionalization of the outer surface of the Janus nanobowls from dissolution is possible. This work is one of the first examples of concave objects bearing different functionalities in the inner and outer parts of their surface.
■
simplest example is given by Janus nanoparticles, first postulated by Pierre-Gilles deGennes in his 1991 Nobel lecture.5 Janus nanoparticle is a generic term, which can refer to very different types of anisotropic particles, such as surface functionalization asymmetry or a combination of both shape and functionalization asymmetry. If one considers that nanoparticles are building blocks for larger, more complex structured materials, it is straightforward to realize that Janus nanoparticles are better candidates than their isotropic counterparts. There is broad literature on the preparation of Janus particles, and a few reviews have been written that cover all of these aspects.6−9 Most of the developed methods for the preparation of Janus nanoparticles suffer from one major limitation: the quantities of Janus particles that can be produced are often quite small. Only a handful of bulk synthesis methods are currently available that allow one to prepare Janus particles in large quantities, are scalable, and are flexible enough to allow the preparation of particles covering a broad range of sizes. Asymmetry in nanoparticles can be introduced in other ways. Dipolar interactions, for example, can be induced by means of electric fields or by magnetic fields if nanoparticles are made of magnetic materials.10,11 Nonspherical nanoparticles can also
INTRODUCTION With the advent of the nanotechnology initiative in year 2000, the research activity on nanoparticles has become more and more prominent. Many research groups are dedicating considerable efforts in developing advanced procedures to prepare novel nanoparticles and to investigate their properties.1,2 For applications, the availability of large quantities of nanoparticles with well-defined properties prepared using simple synthetic approaches is an aspect of the utmost importance. So far, most of the research on nanoparticles has focused on spherical, isotropic nanoparticles. These are the easiest to produce because spheres have minimal surface area per volume, thus minimizing the high surface energy associated with nanoparticles formation. Many recipes have been available for quite some time to prepare spherical nanoparticles from a plethora of materials. Spheres, however, offer rather limited possibilities when it comes to use them as building blocks. Spherical particles can only spontaneously self-assemble into colloidal crystals, colloidal aggregates, and gels and colloidal glasses, depending on their monodispersity and their interactions. Conversely, simulations as well as some experiments, for example, with DNA-functionalized particles,3,4 have clearly demonstrated that complex structures are accessible only when asymmetric interactions are present. One approach to move beyond isotropic particles is to prepare anisotropically functionalized nanoparticles. The © XXXX American Chemical Society
Received: February 25, 2015 Revised: March 31, 2015
A
DOI: 10.1021/acs.langmuir.5b00727 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
study, either Milli-Q (Millipore) water or distilled water stripped with nitrogen (for at least 40 min) was used. Synthesis of Nanoparticles Seeds. The polystyrene seed nanoparticles used a template that has been synthesized by two different techniques. Large polystyrene nanoparticle seeds (seeds_1) have been synthesized by surfactant-free emulsion polymerization.33 A mixture of 180 mL of water and 20 mL of methanol (90v/v % water) was used a solvent in a three-necked round-bottomed flask equipped with a condenser and two septums. The solution was magnetically stirred at 500 rpm throughout the whole polymerization and was first purged with nitrogen for 60 min. Then, 24.56 g of styrene and a mixture of 0.1962 g of VBS and 5 mL of water were added to the system. After heating to 65 °C and equilibrated for 40 min, 0.135 g of KPS dissolved in 10 mL of water was added with a syringe through a rubber septum to start polymerization. The reaction was carried out overnight (16 h). Small seeds (seeds_2 and seeds_3) were prepared through a normal emulsion polymerization. Typically, 1 g of SDS was dissolved in 83.75 mL of water, which was added to a three-necked round-bottomed flask equipped with a condenser and two rubber septums, and purged with nitrogen for 40 min, with magnetic stirring at 400 rpm. Twenty grams of styrene was added to the system and mixed for 30 min. Then, the system was heated to 70 °C, and after thermal equilibration 0.1 g of KPS dissolved in 10 mL of water was added dropwise over 10 min. The system was let to react overnight. The size of the seeds obtained by emulsion polymerization can be adjusted by the amount of SDS in the system. The very small seeds (seeds_4 and seeds_5) were purchased from Magspheres. The detailed specifications of all nanoparticle seeds are listed in Table 1 in the SI. Coating with Random Copolymer of Styrene and 3Trimethoxysilyl Propyl Methacrylate (St-co-MPS). The second step consists of coating the original seeds with a more hydrophilic polymer layer.13,18 Typically, seeds were dispersed in water to reach a concentration of 9.8 wt %. The solution was stripped with nitrogen for 40 min, then transferred to a three-necked round-bottomed flask equipped with a condenser, an overhead stirrer (Premex Reactor AG), and a rubber septum. The system was mechanically stirred at 380 rpm and purged with nitrogen for 60 min. Then, an amount of monomer and initiator equal to that of seeds nanoparticles was added to the seeds solution. This composition of the monomer solutions was 80 vol % styrene and 20 vol % MPS. The amount of AIBN initiator corresponds to 3 wt % of the monomer solution weight. After mixing for 60 min, the system was heated to 70 °C to start polymerization and allowed to react overnight. The small seeds were dispersed in water to reach a concentration of 4.9 wt % to avoid coagulation. The detailed specifications of all coated seeds can be found in Table 2 in the SI. Asymmetric Janus Dumbbell Synthesis. The last step of the Janus dumbbell synthesis consists of swelling the coated seeds with additional styrene monomer. The coated seeds were further diluted (to 3.5, 1.75, or 0.875 wt %) with stripped water to prevent coagulation, and a mass of VBS equal to 0.8% of the mass of particles was added to the bottle and magnetically stirred at 250 rpm for 15 min. The solution was stripped with nitrogen for 30 min and transferred to a three-necked round-bottomed flask equipped with a condenser, an overhead stirrer (Premex Reactor AG), and a septum, and the solution was purged with nitrogen for 60 min. A swelling solution of styrene and AIBN was prepared. The volume of styrene was varied but was usually chosen as a multiple of the total volume of coated seeds (different swelling ratio), and the mass of AIBN was equal to 3 wt % of the swelling solution weight. The solution was added to the system and mixed for 60 min; then, the system was heated to 70 °C and let to react overnight. The detailed specifications of all dumbbells can be found in Table 3 in the SI. Selective Silica Coating on One Hemisphere. Before coating the Janus nanoparticles with silica, they were first transferred to EtOH. Three cycles of centrifugation (35 min at 30 000g, Beckmann Coulter Centrifuge Avanti J-26 XP)−redispersion (12 min sonication, Hielscher sonicator UP 400S, 0.5 s cycle, 60% amplitude) were performed. Typically, 3.84 mL of the nanoparticles in EtOH was diluted with 6.16 mL of EtOH, and 0.5 mL of NH3 was added. A 50/
have the tendency to interact in a more complex fashion. Preparing nanoparticles with nonspherical shapes is quite challenging and has so far been accomplished mostly with inorganic materials. Polymer colloids, instead, which represent one of the most common types of nanoparticles, are difficult to prepare with a nonspherical shape. Solomon and coworkers, for example, have developed a method to turn spherical colloids into ellipsoidal particles by means of a controlled stretching procedure.12 A few recipes have been recently developed by Mock and Zukoski13,14 by Weitz and coworkers15,16 and by Dufresne and coworkers17,18 to directly produce nonspherical polymer particles. In all of these cases, dumbbell particles, and, in some cases, trimers particles, have been targeted. However, the size of the produced particles is usually closer to the micrometer range than to the 100 nm threshold. A particularly interesting example of shape asymmetric nanoparticles that has drawn the attention of scientists in the past decade is particles with an accessible concave surface. A variety of such particles have been prepared, including nanobowls, nanocups, dimpled particles, etc.19−23 The interest in such particles is diverse. Dimpled particles have been used in self-assembly experiments to create lock and key colloids through depletion interactions.24,25 Nanocups and nanobowls have been recently prepared by Mo et al. and investigated as nanocontainers with a defined opening.26 Additionally, nanobowls made of gold were investigated for their shape-related plasmonic properties.27 Some interesting optical properties have also been observed for non-gold-based nanoparticles.28−30 Asymmetrically functionalized nanobowls made of polystyrene have also been reported.31 We present a simple method to prepare large quantities of shape-anisotropic, asymmetrically functionalized Janus nanodumbbells bearing silane groups on one hemisphere, covering a broad range of sizes.32 The dumbbells, made of polystyrene, have been asymmetrically coated by silica on one hemisphere to create hybrid polymer−silica Janus dumbbells. Then, they were used as templates for the synthesis of monodisperse Janus silica nanobowls, bearing a clearly defined hole in their wall, therefore giving an easy access to their concave interior. Two types of nanobowls have been fabricated. By removing the polymeric template by calcination, simple homogeneous silica nanobowls have been prepared. By dissolving the polymer template with an organic solvent, hybrid polymer silica Janus nanobowls have been engineered, having a hydrophobic inner surface and a hydrophilic outer surface that can be selectively functionalized, showing the Janus character of this special nanobowl.
■
EXPERIMENTAL SECTION
Materials. Styrene (99%, stabilized with 10−15 ppm 4-tbutylcathecol) was purchased from ABCR. 2-2-Azobis 2-methyl propionitrile (AIBN) and potassium peroxodisulfate (KPS) were purchased from Fluka. Sodium dodecyl sulfate (SDS) and methanol (MeOH, 99.5%) were purchased from Merck. 3-Trimethoxysilylpropyl methacrylate (MPS), sodium 4-vinylbenzenesulfonate (VBS), LUDOX-CL colloidal silica (30 wt % in water), and LUDOX HS-40 colloidal silica (40 wt % in water) were purchased from Aldrich. Tetraethylorthosilicate (TEOS, 98%) was purchased from Acros Organics. Ammonium hydroxide solution (25% NH3) was purchased from Sigma-Aldrich. Ethanol (EtOH, Absolute Reagent) was purchased from HoneyWell. N-Trimethoxysilylpropyl trimethylammonium chloride (TMSPTMAC, 50% in MeOH) was purchased from Gelest. Tetrahydrofuran (THF) was purchased from VWR. Polystrene standards of 22 and 50 nm were purchased from Magsphere, USA. All chemicals were used without any further purification. Throughout the B
DOI: 10.1021/acs.langmuir.5b00727 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir 50 vol % solution of TEOS and EtOH was prepared (typically 0.3 g each) and added to the solution with a syringe pump (World Precision Instrument ALADDIN-1000) at 1 mL/h to prevent secondary nucleation of silica. The detailed specifications of all silica-coated dumbbells can be found in Table 4 in the SI. Janus Silica Nanobowls Synthesis. To remove the polymeric template and get the silica nanobowls, we have used two strategies. The first consists of calcining the particles in an oven (Nabertherm P330) at 550 °C for 5 h, which leads to the full removal of the polymers (styrene and MPS). The second strategy consists of removing the ethanol solvent with the help of a rotavap (Heidolph Laborota 4003 Control) and replace it with THF. Then, the particles were magnetically stirred at 500 rpm for 24 h to dissolve only the polymer chains not covalently bound to silica. The so-obtained nanobowls were then retransferred to water after three centrifugation (35 min at 30 000g, Beckmann Coulter Centrifuge Avanti J-26 XP)− redispersion (12 min sonication, Hielscher sonicator UP 400S, 0.5 s cycle, 60% amplitude) cycles. Silica Nanobowls Functionalization. The dry nanobowls obtained either by calcination or by dissolution of the template (typically between 5 and 20 mg) were redispersed in 5 mL of EtOH by exposing them to 5 min of sonication (Hielscher sonicator UP 400S, 0.5 s cycle, 60% amplitude). Then, this 5 mL of the bowls in EtOH was mixed with 0.25 mL of NH3 and 0.125 mL of TMSPTMAC and magnetically stirred at 500 rpm overnight. After reaction, the samples were cleaned by three centrifugation (10 min at 15 000g, Beckmann Coulter Centrifuge Avanti J-26 XP)−redispersion (1 min sonication, Hielscher sonicator UP 400S, 0.5 s cycle, 60% amplitude) cycles in water. Heteroaggreation of Silica Nanobowls. LUDOX-CL nanoparticles (positively charged) were diluted to 0.1 wt % with HCl pH4 solution. LUDOX HS-40 nanoparticles (negatively charged) were diluted to 0.1 wt % with water. To 10 mL of the colloidal silica solution, 1 mL of the nanobowls in water was added and stirred at 500 rpm overnight. To remove the unbound LUDOX particles, the reaction mixture was smoothly centrifuged four times (5 min at 7000g, Beckmann Coulter Centrifuge Avanti J-26 XP) and redispersed in 5 mL of water (1 min sonication, Hielscher sonicator UP 400S, 0.5 s cycle, 60% amplitude). Characterization. TEM and SEM were prepared by dropping 20 μL of a highly diluted sample on a copper grid covered with a carbon film or a silica wafer, respectively. TEM analysis was done on either a Hitachi H-7100 operated at 75 kV or a Philips CM100 operated at 80 kV (used for higher magnification pictures). SEM analysis was done on a FEI XL30 Sirion FEG operated at 10 kV or a TESCAN MIRA 3 LMH operated at 10 kV, after coating the sample with 10 nm of platinum to avoid charging. Elemental analysis (EA) was done on a Thermo Scientific Flash 2000 organic elemental analyzer setup for CHNS detection, with sulfanilimide standards. Thermogravimetric analysis (TGA) was done on a Mettler Toledo TGA/DGSC 1 Stare System. EA and TGA were performed on dry samples. Both dynamic light scattering and zeta potential measurements were performed on a Brookhaven 90 plus particle size analyzer.
■
Figure 1. Sketch of the multistep synthesis to make Janus silica nanobowls.
will be propagated along the different steps and to the final product. Seeds with different sizes have been used for this work. A surfactant-free emulsion polymerization with VBS as a comonomer has been used to prepare larger seeds (seeds_1).33 The sulfate groups present in the VBS-styrene copolymer have the tendency to position at the surface of the particles and act as stabilizer in the same way as surfactant molecules. The colloidal stability of these particles is also enhanced by the residual charged moieties coming from the initiator bound to the polymer, giving the polystyrene seeds a negative zeta potential. For large particles (few hundred nanometers in diameters), this is enough to fully stabilize the nanoparticles, which are stable in water over 1 month. Nanoparticles obtained from this procedure are monodispersed down to a size of 150 nm. For smaller particles (down to 100 nm), the total surface area increases dramatically and the sulfate groups are not stabilizing enough, and the particles are not colloidally stable. This issue can be avoided by using a normal emulsion polymerization, with the use of SDS as a surfactant (seeds_2 and seeds_3). Commercial nanoparticles purchased from Magsphere were used as seeds for the smallest nanobowls (seeds_4 and seeds_5). Their size distribution is not as narrow as that of the larger particles. Even though the details of their preparation are not specified, the commercial seeds are also negatively charged, suggesting that either batch emulsion or semibatch emulsion polymerization has been used for their synthesis. The first step after preparation of these seeds is a seededpolymerization aiming at rendering their surface more vitreous and hydrophilic. This is achieved by using a random copolymer of styrene and MPS, the latter providing silane groups. These silane groups have the tendency to locate at the surface of the nanoparticles. The only measurable consequence of this step is the growth of the particles. From TEM images, it is not possible to distinguish a core−shell structure because the contrast difference between polystyrene and polystyrene-co-MPS is too small. In addition, the first swelling step is also performed to bring silane groups on the surface, which can then be specifically functionalized, to enhance the hydrophilicity of
RESULTS AND DISCUSSION
This work has two main objectives. First, we develop a simple procedure to produce hybrid polymer−silica Janus dumbbells covering a brad range of sizes. Then, we utilize these dumbbells as templates for the preparation and characterization of asymmetric silica Janus nanobowls. A schematic of the procedure followed is shown in Figure 1. The preparation starts with the synthesis of monodisperse polystyrene nanoparticles. These nanoparticles are used as seeds for two subsequent polymerization steps, with the objective to produce polymer dumbbells, following a recipe adapted from the literature.18 One of the important prerequisite for the preparation of well-defined nanobowls is to start from monodisperse seeds. Any polydispersity in the starting seeds C
DOI: 10.1021/acs.langmuir.5b00727 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Figure 2. TEM pictures of the different steps: (A,D,G,J) Monodisperse polystyrene nanoparticles used as seeds; (B,E,H,K) dumbbell nanoparticles made using a 3:1 (for panels B, E, and K) and 6:1 (for H) swelling ratio; and (C,F,I,L) dumbbells asymmetrically coated with silica. Scale bars are 500 nm in pictures A−I and are 200 nm in pictures J−L.
Figure 3. TEM and SEM pictures of Janus nanobowls: (A,D,G,J) TEM images of Janus Nanobowls obtained from the calcination of the template shown in Figure 2C,F,I,L, respectively; (B,E,H,K) TEM images of Janus nanobowls obtained from the dissolution of the template shown in Figure 2C,F,I,L, respectively; (C,F,I,L) SEM images of Janus nanobowls obtained from the calcination of the template shown in Figure 2C,F,I,L, respectively. Scale bars are 500 nm in pictures A−I and are 200 nm in pictures J−L.
the swelling solution.16 We found that a swelling ratio, defined as the amount of monomer added to that of the particles, equal to three or four to one, was optimum to get dumbbells having two lobes of the about the same size, as can be seen in Figure 2B. In this work we did not aim at tuning the relative size of the two hemispheres, so we only used recipes that lead to equalsized lobes. What is more important for this work is to be able to synthesize dumbbells starting with nanoparticles seeds having different sizes and different surface chemistry (VBS vs SDS) without affecting the final dumbbell shape. Our results, some shown in Figure 2, indicate that indeed the synthesis can
the nanoparticles surface, which is very important for the next step.13 The second swelling with a monomer and initiator solution and subsequent polymerization step leads to the formation of polymeric dumbbells. It is important to wait long enough for the monomer solution to diffuse into the coated particles before starting the polymerization by increasing the temperature. When the newly introduced monomer starts to polymerize in the particles, it will bulge out due to the presence of a more hydrophilic layer, as explained in the literature.13,18 The size of the newly formed bulge and therefore the final relative size of the two hemispheres can be tuned by changing the volume of D
DOI: 10.1021/acs.langmuir.5b00727 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Figure 4. TEM pictures of Janus nanobowls obtained by (A) calcination and (B) dissolution of the polymeric template. Scale bars are 500 nm.
Figure 5. SEM pictures of Janus nanobowls obtained by (A) calcination and (B) dissolution of the polymeric template. Scale bars are 500 nm.
be carried out starting from seeds with very different size and different surface chemistry. Once the purely polymeric dumbbells have been synthesized in high yield, the next step is to selectively coat them with silica. The selective coating restricted to one hemisphere is possible because the dumbbells have not only shape but also surface anisotropy. In fact, the first hemisphere bears some silane groups contained in the MPS molecules. The second hemisphere does not contain any silane moiety and has just surfactant on its surface. This is the crucial aspect for the rest of the synthesis. The coating with silica is done following a Stoeber-like method.34 It consists of an ammonia-catalyzed hydrolysis and condensation of a silica precursor, TEOS. Because the dumbbells are made in water, they first need to be transferred to EtOH. Three centrifugation/redispersion cycles are usually enough to transfer them with a minimal loss of matter. To minimize secondary nucleation of silica nanoparticles during the condensation reaction, TEOS was mixed with EtOH and added with a syringe pump.35 This ensures a pretty uniform coating of silica. The thickness of the newly formed silica layer could be tuned by changing the amount of silica precursor introduced to the system. Typical TEM images of silica coated dumbbells are shown in Figure 2C,F,I,L. It appears that the silica coating is quite rough and also that some secondary nucleation may still be present but only to a very low extent. In any case, the polymer silica hybrid Janus dumbbells are purified by centrifugation to remove particles generated from secondary nucleation. The evolution of the size of the nanoparticles at the different steps, as measured by DLS, is reported in Table 5 in the SI.
The last step required for the preparation of the nanobowls is the removal of the polymer. This has been achieved following two different strategies. The first one consists of simply burning the polymer via calcination. In this manner, all of the polymer can be removed, leading to a silica nanobowl with one welldefined hole. The hole comes from the asymmetric coating of the dumbbell with the silica layer. As just the first hemispheres bears silane group, the condensation of TEOS is taking place on one lobe only. The silica layer covers the hemisphere until the junction of the two lobes, and this morphology is kept after having the polymeric template removed. The second strategy, instead, consists of dissolving the polymer away by means of an organic solvent. In this manner, however, the polymer layer covalently bound to the inner surface of the shell through the MPS monomer groups remains attached to the silica. The morphology of the nanobowls can be investigated by electron microscopy, as shown in Figure 3. The combination of TEM and SEM permit us to fully visualize the 3D structure of the nanobowls. From these pictures, it appears possible to distinguish some small morphology difference between the nanobowls obtained by calcination or by dissolution from the same silica-coated dumbbells. One could expect to see a slightly thicker shell in the nanobowls obtained from dissolution of the polymer. In this case there are still polymer chains covalently bound to the silica, making the overall thickness of the final bowls larger. This effect is certainly not very visible because the thickness is only enhanced by few nanometers, but it can still be observed. As can be seen in Figure 4, the nanobowls obtained from dissolution of the template appear darker and thicker due to the polymers. For SEM, it appears from Figure 5 that the E
DOI: 10.1021/acs.langmuir.5b00727 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
the mass loss of some nanobowls is shown as a function of the temperature. The results indicate that nanobowls obtained from calcination show no mass loss upon heating, while those obtained from dissolution show a substantial weight loss (of ∼40%) in the 300−500 °C range. The loss of matter observed is due to the decomposition of the polystyrene chains attached to the nanobowls. The temperature at which it happens is typical for polystyrene decomposition, especially for the silica_dumbbells_1 and silica_dumbbells_5 curves.36−38 The small mass loss present at the beginning for dissolved bowls obtained from silica_dumbbells_1 may be due to impurities present in the sample. The difference in absolute content of polymer in the two samples (silica_dumbbells_1 and silica_dumbbells_5) can be explained by the silica content. There was more secondary nucleation of silica in the coating of dumbbells_1 with TEOS, as can be seen in Figure 7. Therefore, there is more silica in the nanobowls sample and the mass loss during the TGA is smaller. These results provide clear proof of the differences between the two types of nanobowls. TGA can be used to confirm that a thicker silica shell can be obtained by adding more TEOS in the system, as previously discussed. Taking the same recipe as for silica_dumbbells_1 in Table 4 in the SI but by adding only 0.113 g of TEOS and EtOH, the total amount of silica in silica_dumbbells_8 drops. This can be easily seen in Figure 6, where the curve for the dissolved silica_dumbbells_8 shows an amount of silica only equal to roughly 10%. The fact that the decomposition temperature for this sample is smaller than that for the others is not clear, but the pictures show that we can vary the thickness of the silica shell. This is also highlighted in Figure 8, where the calcined silica_dumbbells_8 appear very thin. The dissolved one instead, containing a polymer layer, has a larger thickness due to the polymer. From EA, instead, not only can the mass loss can be quantified but also the presence of polymer is indicated by the presence of both hydrogen and carbon signals. Table 1 shows the results of EA for different nanobowls. It can be seen that the total amount of polymer in the dissolved nanobowls is about the same as that obtained from TGA for dumbbells_1 and dumbbells_5. Additionally, the theoretical weight ratio of hydrogen to carbon in polystyrene is 1:12 (same number of carbon and hydrogen atoms in a monomer unit), while the ratio obtained by EA is between 10.25 and 7.75 for our samples, which is quite close to the theoretical one. The difference in this ratio can be explained by the presence of the MPS molecules, which also contain a lower carbon-to-hydrogen ratio,
one obtained from dissolution has a slightly smaller hole in its wall due to the thin polymer layer. One could expect this effect to be more easily visible for small nanobowls because the relative thickness of the polymer layer with respect to the silica shell is larger. Because all of the polymer dumbbells have been prepared by free radical polymerization, the average polymer length should be almost independent of the nanoparticles size, meaning that the thickness of the polymer chains covalently bound to silica does not vary. It should then be easier to see a thin polymer shells on a small silica bowl (a few tens of nanometers in diameter) as compared with a larger one (>100 nm). Even though it looks like TEM pictures of nanobowls obtained from dissolution show a darker contrast than those obtained from calcination, no conclusive evidence of the presence of a polymer layer in nanobowls prepared from dissolution can be obtained from electron microscopy. The visual differences shown in Figures 4 and 5 indicate that our hypothesis (different template removal giving birth to different types of nanobowls) is correct, but it is not enough by itself. To prove that the composition of the inner side of the nanobowls depends on the way how the polymeric template has been removed, we have performed two different types of analysis to characterize the nanobowls: Elemental Analysis (EA) and Thermo Gravimetric Analysis (TGA). Both EA and TGA are used to assess the presence of a polymer layer bound to the silica. Typical TGA curves are shown in Figure 6, where
Figure 6. Thermogravimetric analysis of calcined and dissolved Janus silica nanobowls.
Figure 7. TEM images of calcined nanobowls obtained from (A) Dumbbells 1 and (B) Dumbbells 5. The secondary nucleation in panel A is easily visible. Scale bars are 500 nm. F
DOI: 10.1021/acs.langmuir.5b00727 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Figure 8. SEM pictures of the calcined (A) and dissolved (B) silica_dumbbells_8. Scale bars are 500 nm.
To show the Janus character of our silica nanobowls, we performed heteroaggregation experiment with some small silica nanoparticles, commercially available under the trade name LUDOX. We performed two types of experiments, the results of which are shown in Figure 9. Figure 9A,D shows SEM pictures of nanobowls obtained by calcination and dissolution, respectively. First, we mixed positively charged LUDOX-CL nanoparticles (zeta potential of +24.7 mV) with the silica nanobowls obtained after dissolution and calcination of the silica_dumbbells_1 (zeta potential of −58.2 and −51.1 mV, respectively). Both of them have a negative zeta potential, as expected for bare silica. The positively charged LUDOX-CL particles aggregated both inside and outside the calcined nanobowls, as both inner and outer surfaces are identical, as shown in Figure 9B,G. Conversely, in the case of the dissolved nanobowls, they only attached to the outer surface of the nanobowls due to the presence of a polymer layer inside the nanobowls, preventing nanoparticles attachment, as shown in Figure 9E,H. The presence of this polymer layer only on the inner side of the nanobowls not only reduces the native negative charge of the silica surface but also screens it and creates a hydrophobic coating, which prevents the electrostatic binding of the positively charged LUDOX nanoparticles. This is a first proof of the Janus character of the dissolved nanobowls, which have a different surface character inside and outside and therefore different reactivity toward the LUDOX-CL colloidal silica nanoparticles. In the second experiment, we first functionalized the two types of silica nanobowls with a positively charged silane molecule (TMSPTMAC). The zeta potential of the nanobowls after functionalization switches to positive values, as expected (+20.8 mV for the dissolved ones, +31.2 mV for the calcined ones). We then mixed these silica nanobowls with some negatively charged LUDOX HS-40 colloidal silica nanoparticles (zeta potential of −23 mV). The obtained results shown in Figure 9 demonstrate that LUDOX HS-40 are present inside and outside when mixed with the calcined nanobowls (Figure 9C), while once again they only adsorb to the outer surface in the case of the nanobowls prepared from dissolution (Figure 9F). This second experiment shows that the Janus silica nanobowls can be asymmetrically functionalized because the inner polymer layer prevents reaction of the surface with TMSPTMAC. This therefore further confirms the Janus character of the silica nanobowls obtained by dissolution of the polymeric template, whose outer surface can be selectively
Table 1. Elemental Analysis Results for Calcined and Dissolved Nanobowls dumbbells used
template removal
carbon %
hydrogen %
silica_dumbbells_1
calcination dissolution calcination dissolution calcination dissolution
0.8 28.8 0.18 44.81 0.7 26.22
0.45 3.37 0.52 4.38 0.39 3.37
silica_dumbbells_5 silica_dumbbells_7
equal to 7.63. Their presence can therefore explain the experimental results. In the case of nanobowls obtained from calcination, no significant weight loss and no traces of carbon or hydrogen are detectable. These results provide additional proof of the different composition of the two types of nanobowls. In addition, the TGA presented in Figure 6 can be used to roughly calculate the thickness of the polymer layer in the dissolved nanobowls. By knowing the size and concentration of the dumbbells used and the amount of TEOS introduced in the system, one can calculate the mass of silica on each dumbbells, assuming 100% conversion of the precursor to silica, no secondary nucleation, and a spherical geometry (instead of bowl-like geometry), which are reasonable approximations in the frame of this calculation. This amount can then be converted into the thickness of the silica layer nucleated on the particle. This gives a silica layer thickness of 5 and 40 nm for dumbbells 1 and dumbbells 5, respectively. Considering the difference in size between the particles used, these values are consistent. The weight of polymer in the dissolved nanobowls can be calculated from the mass of silica per particles and the mass ratio polymer/silica obtained by the TGA curves. This mass can be converted into volume and then to a thickness of polymer by knowing the geometry of the particles. In both cases a polymer thickness of 22 nm was found. Indeed, the thickness of the polymer layer is expected to be independent of the size of particles, as free radical polymerization should give polymer chains of approximately the same average molecular weight, provided that the ratio between monomer amount and initiator is kept constant, which is the case in our experiments. Despite the approximations used in these calculations, they allow us to have a more quantitative understanding of structure of both silica and polymer layers, which turn out to be consistent with our expectations. G
DOI: 10.1021/acs.langmuir.5b00727 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Figure 9. SEM pictures of (A) the calcined nanobowls before heteroaggregation, (B) after mixing with positively charged LUDOX, and (C) after functionalization with TMSPTMAC and mixing with negatively charged LUDOX. (D) Dissolved nanobowls before heteroaggregation, (E) after mixing with positively charged LUDOX, and (F) after functionalization with TMSPTMAC and mixing with negatively charged LUDOX. (G,H) High magnification images of the nanobowls shown in panels B and E, respectively. Scale bars are 500 nm for images A−F and 100 nm for G and H.
functionalized independently of the inner surface. To show that the Janus character does not depend on the size of the nanobowls, we performed the heteroaggregation of positively charged LUDOX-CL with both dissolved and calcined nanobowls obtained from silica_dumbbells_4 and silica_dumbbells_6. We obtained analogous results, as shown in Figure S1 in the SI.
polymer template is removed. Calcination of the polymer template leads to purely hydrophilic silica nanobowls, while its dissolution in THF leaves a thin polymer layer on the inner side of the nanobowls, covalently bound to the silica via the MPS molecules. These results have been confirmed by TEM, SEM, TGA, and EA. To prove the Janus character of the nanobowls obtained by polymer dissolution, heteroaggregation experiments with positively charged silica nanoparticles have been performed. The results show that silica nanoparticles bind only to the outer surface of silica nanobowls obtained from dissolution, while they bind to both the inner and the outer surface of nanobowls obtained from calcination. Additionally, the nanobowls have also been functionalized with a cationic silane first and afterward heteroaggregated with negatively charged silica nanoparticles. The results show once more a selective adsorption of the silica nanoparticles only to the outer surface of the nanobowls obtained from dissolution, while nanoparticles attach to both the inner and outer surface of the nanobowls obtained from calcination. This result implies that the outer surface of the Janus nanobowls obtained from dissolution can be selectively functionalized without affecting their inner surface. The proposed synthetic strategy opens the way to produce nanocontainers with well-engineered shape and tunable size, where the different functionalization of their inner and outer surfaces could be used to capture and transport a cargo.
■
CONCLUSIONS In this work, we introduced a simple procedure to synthesize first hybrid polymer silica Janus dumbbells covering a broad range of sizes and then used them to prepare well-defined Janus silica nanobowls, with a different composition of their inner and outer surface. Starting with monodisperse polystyrene seed nanoparticle, we coated them with a random copolymer of styrene and MPS, bringing silane moieties on the surface. A second swelling and polymerization step gives birth to shapeanisotropic, asymmetrically functionalized dumbbells, which only bear silane groups on one hemisphere. This allows one to selectively coat only one lobe of these dumbbells with a layer of silica, as the TEOS will only condense on the hemisphere bearing the silane groups. The object produced in this manner is a hybrid polymer−silica Janus dumbbell. Upon removal of the polymeric template, Janus silica nanobowls with one welldefined hole are obtained. The size of the hole is dictated by the shape of the precursor dumbbell. Moreover, these nanobowls can be made different depending on how the H
DOI: 10.1021/acs.langmuir.5b00727 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
■
(16) Weitz, D. A.; woong Kim, J.; Lee, D.; Shum, H. C.; woong Kim, J. Colloid Surfactants for Emulsion Stabilization. Adv. Mater. 2008, 20, 3239−3243. (17) Dufresne, E. R.; gyu Park, J.; Forster, J. D. Synthesis of Colloidal Particles with the Symmetry of Water Molecules. Langmuir 2009, 25 (16), 8903−8906. (18) Dufresne, E. R.; gyu Park, J.; Forster, J. D. High-Yield Synthesis of Monodisperse Dumbbell-Shaped Polymer Nanoparticles. J. Am. Chem. Soc. 2012, 132, 5960−5961. (19) Cheng, Z.; Luo, F.; Zhang, Z.; Ma, Y.-q. Syntheses and Applications of Concave and Convex Colloids with Precisely Controlled Shapes. Soft Matter 2013, 9, 11392−11397. (20) Hyuk Im, S.; Jeong, U.; Xia, Y. Polymer Hollow Particles with Controllable Holes in Their Surfaces. Nat. Mater. 2005, 4, 671−675. (21) Deng, W.; Wang, M. Y.; Chen, G.; Kan, C. Y. Morphological Evolution of Multistage Polymer Particles in the Alkali PostTreatment. Eur. Polym. J. 2010, 46, 1210−1215. (22) Jeong, U.; Im, S. H.; Camargo, P. H. C.; Kim, J. H.; Xia, Y. Microscale Fish Bowls: A New Class of Latex Particles with Hollow Interiors and Engineered Porous Structures in Their Surfaces. Langmuir 2007, 23, 10968−10975. (23) Wang, Y.; Chen, X.; Zhang, J.; Sun, Z.; Li, Y.; Zhang, K.; Yang, B. Fabrication of Surface-Patterned and Free-Standing ZnO Nanobowls. Colloids Surf., A 2008, 329, 184−189. (24) Sacanna, S.; Irvine, W. T. M.; Chaikin, P. M.; Pine, D. J. Lock and Key Colloids. Nature 2010, 464, 575−578. (25) Wang, Y.; Wang, Y.; Zheng, X.; Yi, G.-R.; Sacanna, S.; Pine, D. J.; Weck, M. Three-Dimensional Lock and Key Colloids. J. Am. Chem. Soc. 2014, 136, 6866−6869. (26) Mo, A. H.; Landon, P. B.; Emerson, C. D.; Zhang, C.; Anzenberg, P.; Akkiraju, S.; Lal, R. Synthesis of Nano-Bowls with a Janus Template. Nanoscale 2015, 7, 771−775. (27) Liang, Z.; Susha, A.; Caruso, F. Gold Nanoparticle-Based CoreShell and Hollow Spheres and Ordered Assemblies Thereof. Chem. Mater. 2003, 15, 3176−3183. (28) Liu, J.; Cankurtaran, B.; McCredie, G.; Ford, M. J.; Wieczorek, L.; Cortie, M. B. Investigation of the Optical Properties of Hollow Aluminium ‘Nano-Caps’. Nanotechnology 2005, 16, 3023. (29) Retsch, M.; Schmelzeisen, M.; Butt, H.-J.; Thomas, E. L. Visible Mie Scattering in Nonabsorbing Hollow Sphere Powders. Nano Lett. 2011, 11, 1389−1394. (30) Fielding, L. A.; Mykhaylyk, O. O.; Schmid, A.; Pontoni, D.; Armes, S. P.; Fowler, P. W. Visible Mie Scattering from Hollow Silica Particles with Particulate Shells. Chem. Mater. 2014, 26, 1270−1277. (31) Meng, Z.; Xue, C.; Lu, L.; Yuan, B.; Yu, X.; Xi, K.; Jia, X. Synthesis of Dissymmetrical Nanoparticles with a New Hybrid Silica Template. J. Colloid Interface Sci. 2011, 356, 429−433. (32) Guignard, F.; Lattuada, M. Asymmetrically Functionalized Polymeric Dumbbells. Chimia 2013, 67, 829−829. (33) Chonde, Y.; Krieger, I. Emulsion Polymerization of Styrene with Ionic Comonomer in the Presence of Methanol. J. Appl. Polym. Sci. 1981, 26, 1819−1827. (34) Stoeber, W.; Fink, A. Controlled Growth of Monodisperse Silica Spheres in the Micron Size Range. J. Colloid Interface Sci. 1968, 26, 62−69. (35) Tissot, I.; Reymond, J. P.; Lefebvre, F.; Bourgeat-Lami, E. SiOH-Functionalized Polystyrene Latexes. A Step toward the Synthesis of Hollow Silica Nanoparticles. Chem. Mater. 2002, 14, 1325−1331. (36) Davodi, B.; Lashkenari, M. S.; Eisazadeh, H. Fabrication and Thermal Degradation Behavior of Polystyrene Nanoparticles Coated with Smooth Polyaniline. Synth. Met. 2011, 161, 1207−1210. (37) Peterson, J. D.; Vyazovkin, S.; Wight, C. A. Kinetics of the Thermal and Thermo-Oxidative Degradation of Polystyrene, Polyethylene and Poly(propylene). Macromol. Chem. Phys. 2001, 202, 775−784. (38) Pirkko, P.; Antti, Z.; Harri, V. Thermal Degradation Products of Homopolymer Polystyrene in Air. Scand. J. Work, Environ. Health 1978, 4, 22−27.
ASSOCIATED CONTENT
S Supporting Information *
Tables with the composition and DLS-measured site of nanoparticle seeds, coated seeds, swollen seeds, silica-coated dumbbells, and additional SEM pictures of the calcined and dissolved nanobowls. This material is available free of charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by the Swiss National Science Foundation, grant number PP00P2126483/1 and by the Adolphe Merkle Foundation. We acknowledge Dimitri Vanhecke for his help with the use of the TEM, Christophe Neururer for his help with the use of SEM, Anita Roulin for her help with the TGA, and Jean François Dechezelles for useful discussions.
■
REFERENCES
(1) Cushing, B. L.; Kolesnichenko, V. L.; O’Connor, C. J. Recent Advances in the Liquid-Phase Syntheses of Inorganic Nanoparticles. Chem. Rev. 2004, 104, 3893−3946. (2) Park, J.; Joo, J.; Kwon, S.; Jang, Y.; Hyeon, T. Synthesis of Monodisperse Spherical Nanocrystals. Angew. Chem., Int. Ed. 2007, 46, 4630−4660. (3) Kim, J.-W.; Kim, J.-H.; Deaton, R. DNA-Linked Nanoparticle Building Blocks for Programmable Matter. Angew. Chem., Int. Ed. 2011, 50, 9185−9190. (4) Xing, H.; Wang, Z.; Xu, Z.; Wong, N. Y.; Xiang, Y.; Liu, G. L.; Lu, Y. DNA-Directed Assembly of Asymmetric Nanoclusters Using Janus Nanoparticles. ACS Nano 2012, 6, 802−809. (5) deGennes, P.-G. Nobel Lectures in Physics 1991−1995; World Scientific Publishing Company: River Edge, NJ, 1997. (6) Lattuada, M.; Hatton, T. A. Synthesis, Properties and Applications of Janus Nanoparticles. Nano Today 2011, 6, 286−308. (7) Hu, J.; Zhou, S.; Sun, Y.; Fang, X.; Wu, L. Fabrication, Properties and Applications of Janus Particles. Chem. Soc. Rev. 2012, 41, 4356− 4378. (8) Perro, A.; Reculusa, S.; Ravaine, S.; Bourgeat-Lami, E.; Duguet, E. Design and Synthesis of Janus Micro- And Nanoparticles. J. Mater. Chem. 2005, 15, 3745−3760. (9) Wurm, F.; Kilbinger, A. Polymeric Janus Particles. Angew. Chem., Int. Ed. 2009, 48, 8412−8421. (10) Smallenburg, F.; Vutukuri, H. R.; Imhof, A.; van Blaaderen, A.; Dijkstra, M. Self-Assembly of Colloidal Particles into Strings in a Homogeneous External Electric or Magnetic Field. J. Phys.: Condens. Matter 2012, 24, 464113. (11) Furlan, M.; Brand, B.; Lattuada, M. Magnetic Gelation: A New Method for the Preparation of Polymeric Anisotropic Porous Materials. Soft Matter 2010, 6, 5636−5644. (12) Mohraz, A.; Solomon, M. J. Direct Visualization of Colloidal Rod Assembly by Confocal Microscopy. Langmuir 2005, 21, 5298− 5306. (13) Mock, E. B.; Bruyn, H.; Hawkett, B. S.; Gilbert, R. G.; Zukoski, C. F. Synthesis of Anisotropic Nanoparticles by Seeded Emulsion Polymerization. Langmuir 2006, 22, 4037−4043. (14) Mock, E. B.; Zukoski, C. F. Emulsion Polymerization Routes to Chemically Anisotropic Particles. Langmuir 2010, 26, 13747−13750. (15) Weitz, D. A.; woong Kim, J.; Larsen, R. J. Uniform Nonspherical Colloidal Particles with Tunable Shapes. Adv. Mater. 2007, 19, 2005− 2009. I
DOI: 10.1021/acs.langmuir.5b00727 Langmuir XXXX, XXX, XXX−XXX
