pubs.acs.org/Langmuir © 2009 American Chemical Society
Thermostimulable Wax@SiO2 Core-Shell Particles Mathieu Destribats, Veronique Schmitt,* and Renal Backov* Centre de Recherche Paul Pascal, Universit e Bordeaux 1, UPR 8641-CNRS, 115 Avenue du Dr Albert Schweitzer, 33600 Pessac, France Received July 31, 2009. Revised Manuscript Received September 15, 2009 We propose a new synthesis pathway without any sacrificial template to prepare original monodisperse thermoresponsive capsules made of a wax core surrounded by a silica shell. Under heating, the inner wax expands and the shell breaks, leading to the liquid oil release. Such capsules that allow triggered deliverance provoked by an external stimulus belong to the class of smart materials. The process is based on the elaboration of size-controlled emulsions stabilized by particles (Pickering emulsions) exploiting the limited coalescence phenomenon. Then the emulsions are cooled down and the obtained suspensions are mineralized by the hydrolysis and condensation of a monomer at the wax-water interface, leading to the formation of capsules. The shell break and the liquid oil release are provoked by heating above the wax melting temperature. We characterize the obtained materials and examine the effect of processing parameters and heating history. By an appropriate choice of the wax, the temperature of release can easily be tuned.
1. Introduction One of the first applications of microencapsulation was dedicated to carbon-free self-copying paper that was commercialized in 1968. As many as 110 000 tons of capsules were used for this application in the United States. This is an example of the great importance of capsules. Nowadays, encapsulation is used in classical industrial sectors such as pharmaceutics, cosmetics, food, textiles, and agriculture, and for chemical (drugs, fragrances, flavors, dyes, pesticides) trapping and delivery.1 The capsules always become more sophisticated and more controlled, particularly in the domain of pharmaceutics where addressing a specific target is required. Various morphologies of capsules have been considered, for example, protein vehicles,2 cyclodextrins,3 thermally gated liposomes,4 concentrated lamellar vesicles,5 double emulsions,6-8 colloidosomes,9,10 silica shell microcapsules,11-13 thermosensitive PNIPAM-silica nanocapsules,14 thermosensitive hydrogel micropsheres,15 PNIPAM-polylactide microspheres,16 and so forth. Numerous methods for the preparation of *Corresponding authors. E-mail:
[email protected] (V.S.);
[email protected] (R.B.). (1) Yow, Y. N.; Routh, A. F. Soft Matter 2006, 2, 940. (2) Angelova, A.; Angelov, B.; Lesieur, S.; Mutafchieva, R.; Ollivon, M.; Bourgaux, C.; Willumeit, R.; Couvreur, P. J. Drug Delivery Sci. Technol. 2008, 18, 41. (3) Daoud-Mahammed, S.; Ringard-Lefebvre, C.; Razzouq, N.; Rosilio, V.; Gillet, B.; Couvreur, P.; Amiel, C.; Gref, R. J. Colloid Interface Sci. 2007, 307, 83. (4) Chen, W.-H.; Regen, S. L. J. Am. Chem. Soc. 2005, 127, 6538. (5) Regev, O.; Backov, R.; Faure, C. Chem. Mater. 2004, 16, 5280. (6) Goubault, C.; Pays, K.; Olea, D.; Gorria, P.; Bibette, J.; Schmitt, V.; LealCalderon, F. Langmuir 2001, 17, 5184. (7) Pays, K.; Mabille, C.; Schmitt, V.; Leal-Calderon, F.; Bibette, J. J. Dispersion Sci. Technol. 2002, 23, 175. (8) Rojas, E. C.; Papadopoulos, K. D. Langmuir 2007, 23, 6911. (9) Velev, O. D.; Furusawa, K.; Nagayama, K. Langmuir 1996, 12, 2374. (10) Dinsmore, A. D.; Hsu, M. F.; Nikolaides, M. G.; Marquez, M.; Bausch, A. R.; Weitz, D. A. Science 2002, 298, 1006. (11) Goller, M. I.; Vincent, B. Colloids Surf., A 1998, 142(2), 281. (12) O’Sullivan, M.; Zhang, Z.; Vincent, B. Langmuir 2009, 25, 7962–7966. (13) Fornasieri, G.; Badaire, S.; Backov, R.; Mondain-Monval, O.; Zakri, C.; Poulin, P. Adv. Mater. 2004, 16, 1094. (14) Goa, H.; Yang, W.; Min, K.; Zha, L.; Wang, C.; Fu, S. Polymer 2005, 46, 1087. (15) Kawagushi, H. In Microspheres, microcapsules and liposomes; Arshady, R., Ed.; 1999; Vol. 1 and references therein. (16) Liu, S.-Q.; Yang, Y.-Y.; Liu, X.-M.; Tong, Y.-W. Biomacromolecules 2003, 4, 1784.
1734 DOI: 10.1021/la902828q
core-shell microcapsules have been extensively developed over the years, for example, polymer precipitation by phase separation, layer-by-layer polyelectrolyte deposition, copolymer vesicles, and polycondensation interfacial polymerization, and this list is not exhaustive. Often a slow and progressive release of the capsule content is aimed. Herein the goal is to design capsules exhibiting the ability to open and release at once their content upon application of a soft thermal treatment. Our process belongs to integrative chemistry,17-20 that is to say the coupling of soft matter and soft chemistry, since it combines emulsion science for the preparation of a monodisperse crystallizable oil-in-water emulsion stabilized by colloidal particles and sol-gel chemistry for the mineralization of the emulsion interface. More precisely, at T > Tm, where T is the temperature and Tm is the oil melting temperature, we formulate size-controlled emulsions stabilized by silica colloidal particles (Pickering emulsions)21-23 by means of limited coalescence phenomenon.24 Then the emulsion is cooled down (T < Tm) and the resulting suspension is mineralized by the hydrolysis and condensation of a monomer at the wax-water interface. Capsules, made of a wax core, surrounded by a silica shell are obtained. We shall call them wax@SiO2 because the symbol “@” is now currently used to refer to a core particle surrounded by a shell. They are thermoresponsive, since by heating them (T > Tm) the inner wax expands and breaks the shell, allowing the liquid oil to be released. The triggering temperature is fixed by the choice of the wax. These capsules also offer the advantage of avoiding a sacrificial template and a complex additional loading step. In this article, we detail the capsule elaboration, we characterize the obtained materials, and (17) Backov, R. Soft Mater 2006, 2, 452. (18) Prouzet, E.; Khani, Z.; Bertrand, M.; Tokumoto, M.; Gyuot-Ferreol, V.; Tranchant, J.-F. Microporous Mesoporous Mater. 2006, 96, 369. (19) Sanchez, C.; Boissiere, C.; Grosso, D.; Laberty, C.; Nicole, L. Chem. Mater. 2008, 20, 682. (20) Prouzet, E.; Ravaine, S.; Sanchez, C.; Backov, R. New J. Chem. 2008, 32, 1284. (21) Pickering, S. U. J. Chem. Soc. 1907, 91, 2001. (22) Binks, B. P. Curr. Opin. Colloid Interface Sci. 2003, 7, 21. (23) Leal-Calderon, F.; Schmitt, V . Curr. Opin. Colloid Interface Sci. 2007, 13, 217. (24) Arditty, S.; Whitby, C.; Schmitt, V.; Binks, B. P.; Leal-Calderon, F. Eur. Phys. J. E 2003, 11, 273.
Published on Web 10/09/2009
Langmuir 2010, 26(3), 1734–1742
Destribats et al.
Article
we show the effect of heating. In particular, we examine the release mechanisms depending on the continuous phase composition and discuss the observed phenomena.
2. Experimental Section 2.1. Chemicals. Various crystallizable oils have been used to show the generality of the process. Two wax block paraffins 42-44 and 46-48 (CAS no. 8002-74-2) were purchased from Merck, and the two alkanes eicosane (99%) and octadecane (97%) were purchased from Aldrich. The melting temperatures or narrow temperature domains of the oils are, respectively, 42-44, 46-48, 37, and 29-30 °C. The measured volume expansion due to heating, determined from room temperature to 55 °C, is equal to 13, 9, 7, and 11% for paraffin 42-44, paraffin 46-48, eicosane, and octadecane, respectively. As biological compatible oil, we used Suppocire DM, purchased from Gattefosse. This is a more complex oil made of a mixture of triglycerides, commonly used in suppository formulation, with a broad melting temperature ranging from 27 to 48 °C determined by differential scanning calorimetry (Perkin-Elmer, Pyris 1). The monomer tetraethyl orthosilane (TEOS) and the surfactant cetyltrimethylammonium bromide (CTAB) were purchased from Fluka. Aerosil silica nanoparticles A380 (diameter 7 nm) were provided by Degussa Evonik. All the chemicals were used as received without further purification. In order to examine the effect of addition of various surfactants in the continuous phase during heating, we used an ionic surfactant sodium dodecyl sulfate (SDS) purchased from Aldrich, and a nonionic surfactant Ifralan D205 (mixture of poly(oxyethylene) molecules C12E5 and C10E5) was kindly provided by IfraChem. The critical micellar concentrations of SDS, CTAB, and Ifralan D205 in pure water at ambient temperature are, respectively, 0.82, 0.92, and 0.07 mM (estimated from the data given in literature for C12E5 and C10E5).25-27 2.2. Particle Functionalization. Bare silica particles are only wetted by water; they do not exhibit any “amphiphilic” character and therefore do not adsorb at oil-water interfaces. To obtain “amphiphilic” particles that are partially wetted by both fluids, the hydrophilic surface of the silica particles is made partially hydrophobic by adsorbing a very low amount of surfactant. Aerosil particles are first dispersed into distilled water, and then cationic surfactant molecules are introduced, at a concentration well below the critical micellar concentration (∼CMC/5), in order to electrostatically adsorb at the negatively charged silica surface. The long CTAB alkyl chain allows nanoparticle hydrophobic functionalization, conferring to them an “amphiphilic” nature. This feature promotes the hybrid particles’ ability to anchor and stabilize the emulsion oil-water interfaces, as already described elsewhere.28-30 Following the same protocol, the amount of surfactant was adapted to the total mass of particles in order to maintain the same specific coverage of 13 nm2 per CTAB molecule at the silica-water interface. This value was estimated with the assumption that all the surfactant from the bulk is consumed to cover the silica surface. Oil and water were mixed without particles at the same surfactant concentration (CMC/5); the solution rapidly demixed, suggesting that such low amounts of surfactant alone are not efficient for emulsion stabilization. 2.3. Emulsification Process. Functionalized particles are dispersed in water, the mixture is heated at 65 °C, and molten (25) Weiss, J.; Herrmann, N.; McClements, D. J. Langmuir 1999, 15, 6652. (26) Gao, H.; Zhu, R.; Yang, X.; Mao, S.; Zhao, S.; Yu, J.; Du, Y. J. Colloid Interface Sci. 2004, 273, 626. (27) Schmitt, V.; Cattelet, C.; Leal-Calderon, F. Langmuir 2004, 20, 46. (28) Giermanska-Kahn, J.; Laine, V.; Arditty, S.; Schmitt, V.; Leal-Calderon, F. Langmuir 2005, 21, 4316. (29) Schmitt, V.; Kahn, J.; Reculusa, S.; Ravaine, S.; Leal-Calderon, F.; Arditty, S. Crystallizable oil compositions stabilized by colloidal particles. World Patent no. WO 2005/082507A1. (30) Perro, A.; Meunier, F.; Schmitt, V.; Ravaine, S. Colloids Surf., A 2008, 332, 57.
Langmuir 2010, 26(3), 1734–1742
wax is added drop by drop while maintaining vigorous stirring by means of an Ultra-Turrax homogenizer (T25 JANKE & KUNKEL) equipped with a S25 KV-25F rotor head. In order to homogenize the as-synthesized emulsions, final stirring is addressed through the application of strong shear induced with either an Ultraturrax apparatus operating at 9000 rpm for 1 min or a Rayneri apparatus (Turbotest 33/300P) operating at 3000 rpm during 10 min. The Ultraturrax and Rayneri apparatuses are frequently used homogenizers to produce emulsions. They are equipped with a rotor/stator head that provokes shearing and recirculation of the sample. The Rayneri apparatus allows a gentler mixing with a lower energy input than that of the Ultraturrax mixer. Such a process has already been used elsewhere, where a detailed description can be found.28 At that stage, the emulsions are allowed to come back to ambient temperature under a gentle stirring without any specific cooling rate that is imposed by the beaker inertia. Once cooled below the oil melting temperature, CTAB is added to reach its critical micellar concentration in order to avoid particle aggregation and to stabilize the wax dispersion for storage. 2.4. Mineralization of Pickering Emulsions. The wax emulsions are diluted to 7 wt % by adding a 1 wt % CTAB solution and HCl (37 vol %). In order to catalyze the TEOS hydrolysis-condensation, while optimizing heterogeneous condensation at the wax-water interface, we adjust the pH at around 0.2, below the silica isoelectric point.31 To start the mineralization, TEOS is then added drop by drop at different amounts. The emulsions are allowed to mineralize in 10 mL test tubes under continuous stirring using a wheel rotating at 25 rpm within a thermostatically controlled chamber at 25 °C. Final products are washed several times with water. This method allows precise control of the obtained material morphologies by tuning one parameter at once. The amount of TEOS into the reaction media governs the hybrid object morphologies: for low TEOS concentration, only the oil-water interface is mineralized, leading to core-shell discrete objects, while increasing the TEOS concentration allows the mineralization of the full hydrophilic continuous phase, leading to composite macroscopic monolith-type materials. Also, the amount of Aerosil particles fixes the drop average size, taking advantage of the limited coalescence phenomenon (see later in the text), while the wax type, characterized by different melting temperatures, determines the opening temperature of the silica shell upon applying a thermal treatment. 2.5. Characterization Techniques. Optical microscopy characterizations were performed using an inverted optical microscope, Zeiss Axiovert X100, equipped with a Mettler plate controlling the temperature and heating and cooling rates. The emulsions size distributions were obtained using a Malvern Mastersizer Hydro MS2000 granulometer. Granulometric measurements were performed at 25 °C in a pure water solution. The collected scattering intensity as a function of the angle is transformed into a size distribution using Mie theory. The size distribution was characterized in terms of the surface-averaged diameter D and polydispersity P (eq 1):
P N i Di 3 D ¼ Pi 2 i N i Di
P and
P ¼
1 D
i N i Di D -Di P 3 i N i Di 3
ð1Þ
where Ni is the total number of droplets with diameter Di. D is the median diameter, that is, the diameter for which the cumulative undersized volume fraction is equal to 50%. At the measuring temperature, the droplets are solid and their shape deviates little from spheres. The obtained size distribution corresponds to the
(31) Brinker, C. J.; Scherer, G. W. Sol-Gel Science: the Physics and Chemistry of Sol-Gel Processing; Academic Press: San Diego, 1990.
DOI: 10.1021/la902828q
1735
Article
Destribats et al.
Figure 1. Scheme of the whole procedure used to generate the wax@SiO2 particles. Step 1: dispersion of silica particles in water and simple functionalization by electrostatic adsorption of CTAB onto silica. Step 2: progressive incorporation of molten wax and emulsification at 65 °C. Step 3: limited coalescence leading to monodisperse Pickering emulsion at 65 °C and then cooling at ambient temperature. Step 4: silica shell mineralization around the solid wax droplets (wax@SiO2). distribution of spherical drops that scatter the light in the same way as our samples. Scanning electron microscopy (SEM) observations were performed with a SEM HITACHI TM-1000 apparatus. For better resolution, SEM was also performed with JEOL JSM-840A and JEOL 6700 F apparatuses to image the obtained capsules and to roughly estimate the shell thickness. The suspensions either were previously dried at ambient temperature and atmosphere or were lyophilized using an Alpha 2-4 LD Plus freeze-dryer. All the samples were gold coated prior to observation. Figure 2. Evolution of the inverse mean diameter (1/D) as a
3. Results and Discussion 3.1. Material Syntheses. To synthesize monodisperse temperature-stimulable core-shell particles, we first have to prepare direct emulsions with narrow size distributions (steps 1-3 of Figure 1). Formulation of an emulsion with a good monodispersity can be achieved by taking advantage of the limited coalescence phenomenon occurring in the so-called Pickering emulsions21-23 or solid-stabilized emulsions as they are stabilized by colloidal particles irreversibly adsorbed at the oil-water interface. Limited coalescence24 consists of producing a large excess of oil-water interface compared with the interfacial area that can be covered by the solid particles. Hence, for this process to occur, the systems must be formulated in the presence of a very small amount of solid particles. When the agitation is stopped, the partially unprotected droplets coalesce, thus reducing the total amount of the oil-water interface. Since the particles are irreversibly adsorbed, the coalescence process stops as soon as the oil-water interface is sufficiently covered. The resulting emulsions are characterized by narrow size distributions and are stable over months. Moreover, the size can be controlled by adjusting the amount of particles.24,28,29 Exploiting this phenomenon, we produced monodisperse emulsions over a wide range of sizes (from about 10 μm to hundreds) by controlling the amount of particles (Figures 2 and 3). The strategy of using Pickering emulsions offers a double advantage: First, as explained above, it allows obtaining narrow droplet size distributions. Second, the silica nanoparticles adsorbed at the wax-water interface will serve as nucleation sites. In the crystal growth process, one distinguishes two nucleation types, namely, homogeneous and heterogeneous ones: The homogeneous nucleation occurs within the bulk solution, and new interfaces between the native nuclei and the solution are being created, requiring thereby an additional surface energy. The heterogeneous nucleation occurs on pre-existing nuclei or interfaces where the energy associated to the creation of a new solid-liquid interface is, in this case, minimized. This is the reason why heterogeneous nucleation is always dominating the homogeneous one. Here, the nucleation enthalpy of silica is 1736 DOI: 10.1021/la902828q
function of the mass of particles with respect to the oil mass for an eicosane-in-water emulsion.
minimal at the wax-water interface, promoting thereby mainly heterogeneous nucleation rather than the homogeneous one. We mineralize the eicosane-in-water emulsions shown in Figure 3, keeping the TEOS concentration to surface ratio constant and equal to 1 M/m2 and at a pH value equal to 0.2. The obtained materials are reported in Figure 4. It is worth noticing that there is no significant widening of the distributions after mineralization, showing the absence of noticeable capsule aggregation during this step. After mineralization, the capsules are centrifuged and transferred in pure water in order to remove possible homogeneous nuclei and all traces of surfactant. The capsules can be dried by lyophilizing, and this process does not alter their integrity. This allows easy observation by SEM (Figure 5), and we checked that the silica shell is complete and fully covers the inner wax core. Hence, the capsules can be stored either as dispersed in an aqueous phase or as a dried powder. During storage, the obtained objects are stable for several months. As expected, the capsules break when heated above the oil melting temperature and the oil is released (Figures 6 and 7). For a more detailed study about capsule rupture under thermal treatment, see section 3.3. 3.2. Influence of the Oil Type. The previous protocol was applied to various oils with melting temperature varying from 29 to 48 °C. Alkanes and paraffin defined by a unique melting temperature or by a very narrow melting domain allowed production of nice capsules. Some tests have been carried out with more complex oils made of a mixture of triglycerides. These oils are interesting because of their biocompatible nature. In this latter case, capsules are also feasible. The initial emulsion size distribution is narrow, the mineralization is performed as previously, and the obtained capsules break under thermal treatment, releasing the oil (Figure 8). Oil is an important compound, since it governs the release temperature but also influences the capsule’s aspect and its surface roughness. For example, some silica aggregates are visible on the shell with paraffin (Figure 9a) whereas the capsules made Langmuir 2010, 26(3), 1734–1742
Destribats et al.
Article
Figure 3. Optical microscopy images of crystallized eicosane-in-water emulsions obtained with various amounts of particles in wt % with respect to oil: (a) 0.17 wt %, (b) 0.67 wt %, and (c) 1.4 wt %. Scale bar = 100 μm. (d) Corresponding size distributions of the three emulsions: the mean diameters are 17.4 μm (red line), 39 μm (green), and 121 μm (blue), and the polydispersity indices are, respectively, equal to 0.26, 0.26, and 0.19.
Figure 4. (a-c) Optical microscopy images of eicosane@SiO2 capsules dispersed in water produced from eicosane-in-water emulsions shown in Figure 3. Scale bar =100 μm. (d) Comparative size distributions of the emulsions before (dashed lines) and after (solid lines) mineralization.
with eicosane appear generally smoother at the same scale (Figure 7a). As expressed before, silica nanoparticles used for the oil-water interface stabilization are also important for promoting heterogeneous nucleation. These nucleation sites are favoring silica mineralization and growth, minimizing the nucleation enthalpy. As a consequence, grown silica nanoparticles are reaching sizes high enough (μm) to be observed when using scanning electron microscopy (Figures 5, 9a, and 10). 3.3. Influence of TEOS Concentration over the Material Morphologies. We examined for a fixed formulation (paraffin 42-44-in-water emulsion) the influence of the TEOS concentration, keeping the emulsion mean size constant (15 μm) and Langmuir 2010, 26(3), 1734–1742
consequently the total wax-water interface area constant. One expects an evolution of the silica shell thickness. Examples of final products can be seen in Figure 9. By varying the TEOS concentration from 0.46 to 3.7 M/m2, we have determined a critical concentration below which the core-shell particles are discrete (Figure 9a), while for higher concentrations the overall aspect is more like wax droplets embedded within a silica continuous medium (Figure 9b). This observation suggests that, below a TEOS concentration threshold of about 2.2 M/m2, the silica condensation is mostly heterogeneous and takes place at the wax-water interfaces. When increasing the TEOS amount above 2.2 M/m2, keeping the amount of HCl constant, the silica DOI: 10.1021/la902828q
1737
Article
condensation becomes dual, meaning that heterogeneous nucleation is certainly still effective but occurs with enhanced homogeneous nucleation taking place in the bulk aqueous phase. This homogeneous nucleation and growth in the bulk of the continuous phase is responsible for the formation of a silica gel that
Figure 5. SEM picture of an eicosane@SiO2 capsule after drying. Scale bar = 20 μm.
Destribats et al.
propagates totally within the continuous aqueous media. The pH in use to induce wax mineralization is acidic (pH ≈ 0.2), and this is to say not too far from the silica isoeletric point of pH 2.1.31 At this pH, mineralization leads to both a sufficiently dense (Euclidian) inorganic polymer to offer good mechanical resistance (dense shells) and sufficiently fractal character to accommodate, during the network growth, the wax drops’ curvature but also to induce propagation into the whole continuous hydrophilic media, if the silicic acid (Si(OH)4) groups are numerous enough. As a direct consequence of this dual heterogeneous/homogeneous nucleation and of the fractal character of the growing network, the full aqueous phase may be mineralized, leading to monolithtype compounds bearing continuous silica walls (Figure 9b). In order to determine the effect of TEOS concentration on the capsules, one should be able to measure precisely the shell thickness. From the observations, the capsules appear quite homogeneous on average, as can be seen in the SEM images reported in Figures 5, 7a, and 10b. This feature is even more visible when considering optical microscopy observations on capsules previously emptied by dissolution of the inner wax by tetrahydrofuran (THF) (Figure 11). However, when increasing the magnification, as already pointed out, some roughness appears. As a result, the shell thickness locally fluctuates,
Figure 6. (a) Sequenced pictures (extracted from film 1, Supporting Information) of a capsule in water releasing its wax (Tm = 42-44 °C) upon heating. Scale bar = 20 μm. Top left: capsule below the oil melting temperature. Top right: capsule above the oil melting temperature, shell rupture due to the wax volume expansion and melted oil coming out. Bottom left: shell fracture propagation. Bottom right: empty capsule after drop release. (b) Optical microscopic visualization of the effect of heating the capsules dispersed in water shown in Figure 4b at T > Tm (eicosane = 37 °C). Scale bar = 60 μm. The white arrows indicate the melted oil drops released from the broken capsules, while the black arrow indicates the final silica hollow spheres free of wax.
Figure 7. Examples of SEM pictures of dried eicosane@SiO2 capsules, previously shown in Figure 4c, (a) before and (b) after thermal treatment in an oven at 60 °C. Scale bar = 100 μm. Empty capsules have partially collapsed after oil release. 1738 DOI: 10.1021/la902828q
Langmuir 2010, 26(3), 1734–1742
Destribats et al.
Article
Figure 8. Optical microscopy images of (a) Suppocire DM emulsion, (b) Suppocire DM@SiO2 capsules at room temperature, and (c) Suppocire DM@SiO2 capsules under thermal treatment. Scale bars = 60 μm.
Figure 9. SEM images of the synthesized wax@SiO2 particles with the paraffin 42-44 as oil phase with two different amounts of TEOS introduced within the acidic hydrophilic phase for a 15 μm emulsion. The TEOS concentration equals to (a) 0.46 M/m2 and (b) 3.7 M/m2. Scale bar = 50 μm.
complicating its estimation. Despite this difficulty, the thickness can be roughly estimated by SEM on broken capsules below the critical TEOS concentration. In order to break the capsules, we use the beam focus in situ or, alternatively, we store the suspension in an oven above the melting tempertaure prior lyophilizing it (see Figure 10). By increasing the TEOS concentration from 0.46 to 2 M/m2 during mineralization, we can note a tendency to increase the silica shell thickness by about a factor 2 from roughly 200 to 400 nm (see Figure 12). Better precision is difficult to reach because of the restricted TEOS concentration range. Above 2.2 M/m2, as we are dealing with agglomerated capsules, there is no reason to consider shell thickness anymore. 3.4. Thermostimulated Rupture of the Capsules and Release Mechanisms. When raising the temperature, the inner wax Langmuir 2010, 26(3), 1734–1742
melts and expands, and as a consequence pressure is exerted on the shell causing the capsule to rupture and allowing the melted wax to be released. To determine the opening temperature, the thermal treatment consists of slowly (0.5 °C/min) increasing the applied temperature to approach quasi-static conditions. For the three systems tested (eicosan, paraffin 42-44, and paraffin 46-48), the silica shell breaking temperature, defined as the temperature where the shell breaks and the wax starts to expand out of the silica capsules, is the same as the wax melting temperature. It does not depend on the capsule size or on the shell thickness (in the considered range). The choice of wax determines the wax release temperature and allows an easy tuning from 28 to 48 °C. However, if a fast temperature ramp (>5 °C/min) is applied to the samples, the capsule size and thickness are determining DOI: 10.1021/la902828q
1739
Article
Destribats et al.
Figure 10. (a) SEM picture of paraffin 46-48@ SiO2 as an example of the fracture propagation on the silica shell under inner wax expansion
(white arrow) upon heating. Scale bar = 10 μm. (b) SEM picture of empty octadecane@ SiO2 capsules after release of their inner phase. Scale bar = 50 μm.
Figure 12. Example of a SEM picture of an empty eicosane@SiO2 capsule. The shell has been synthesized using a TEOS concentration of 1.79 M/m2. Scale bar = 0.7 μm.
Figure 11. Optical microscopy image showing the silica shell homogeneity. The capsule cores have been dissolved using THF. Scale bar = 30 μm.
parameters: for a given capsule size, the thinner shells break first, and for a given TEOS concentration/interfacial area (i.e., for a given shell thickness) the smaller capsules break before the larger capsules do. We determined that these observations are only due to kinetic effects. Here, two scenarios can explain this behavior: either the range of the silica shell thickness proposed herein is not wide enough to assess its effect over the breaking temperature, or the shell opening is favored by defects of the silica shells, thereby avoiding the shell thickness effect. Whatever the chosen oil and capsule size, one always observes a fracture in the shell that propagates and the liquefied wax is released from one side of the capsule (Figure 6). The capsules do not open isotropically, but all the liquid releases from the same point, suggesting that a fracture occurs and propagates in the silica shell under the pressure of the wax expansion. This can also be seen by SEM observations (Figure 10). Another interesting aspect to point out is the way the oil is released after rupture of the shell. As can be seen in the previously shown images, the oil exits and forms a spherical oil drop that does not spontaneously detach from the capsule. The presence of a local flow favors the detachment, and the oil drop is then dispersed as a whole in the surrounding water. This may be due to 1740 DOI: 10.1021/la902828q
dewetting because the silica shell is bearing a silanol group (-OH) and possesses some negative charges at its surface.31 In order to increase the oil affinity toward the continuous phase, we performed experiments of release in the presence of various surfactants at a concentration of about 60CMC: anionic SDS, cationic CTAB and nonionic Ifralan D205 (mixture of poly(oxyethylene) molecules C12E5 and C10E5). Moreover, as surfactants are small molecules diffusing very fast and decreasing the interfacial tension between oil and water, they should all promote drop rupture (fragmentation) and as a consequence favor the oil dispersion in the continuous phase. The release scenarios are illustrated in Figure 13, in the absence or presence of various surfactants. SDS almost does not modify the release mechanism compared to the release in the absence of surfactant even if some smaller drops are present in water (Figure 13a and b). Expulsion of the oil outside the capsules is greatly favored by the presence of CTAB or Ifralan D205, and the oil is fragmented in small droplets when exiting the capsules (Figure 13c and d). As examples of the different release modes, films are shown in the Supporting Information (films 2-5). Moreover, one can notice that the droplets are smaller for Ifralan D205 compared to CTAB. We can compare the interfacial tensions between water and dodecane, a liquid oil at ambient temperature, in the presence of SDS, CTAB, and Ifralan D205 above their critical micellar concentrations; they are, respectively, Langmuir 2010, 26(3), 1734–1742
Destribats et al.
Article
Figure 13. Optical microscopy images of the wax release from eicosane@SiO2 capsules under thermal treatment and depending on the type of surfactant in the aqueous phase: (a) no surfactant, (b) SDS (60CMC), (c) CTAB (60CMC), and (d) Ifralan D205 (60CMC). We can distinguish two modes of release. The first one is the formation of big drops (a and b), the detachment of which is enhanced by the use of surfactant. The second one is the release of wax with the formation of many droplets (c and d). This last kind of release throws out almost completely the inner wax. Scale bars = 60 μm.
equal to 7.5, 5.3, and 0.75 mN/m.25,27,32 So, the difference in interfacial tensions could explain the following observation: release and emulsification are increasingly easier from SDS to CTAB to Ifralan D205. In the two latter cases, the oil exits through small fractures produced by expansion of the wax and enlargement to a big shell fracture is not needed for the oil to be released. As a consequence, small droplets are rapidly distributed around the capsules (Figure 13c and d) that can be completely emptied (Figure 14 and Supporting Information film 6). In addition to the decrease in drop size, one can distinguish an increase of the resulting drop polydispersity from Figure 13a to d. One can hypothesize that in all cases there is a large distribution of the “pore” sizes when the silica shell breaks. However, since drop formation is more difficult in cases a and b, the thinner fractures are inefficient and thereby only larger pores allow the oil release. On the contrary, in cases c and d, drops easily rupture even through very small fractures so that very small drops coexist with very large ones probably originating from large holes. It is worth noticing that all drops exiting from the same pore have very identical sizes as is the case with microfluidic devices. This can easily be seen in Figure 13c and Supporting Information film 5. Once dispersed in water, the drops interact attractively through depletion because of the presence of surfactant micelles (surfactant concentration=60CMC). This phenomenon can be observed in Figure 13c or Supporting Information film 6. We checked that the mechanisms are determined by the surfactant, and for a given surfactant it is independent of the encapsulated oil. (32) Medrzycka, K.; Zwierzykowski, W. J. Colloid Interface Sci. 2000, 230, 67.
Langmuir 2010, 26(3), 1734–1742
Figure 14. Optical microscopy images of final capsules. The low contrast shows that the capsules are almost empty after release of the inner wax into the Ifralan D205 concentrated aqueous phase. Scale bar = 60 μm.
The release scenario is the same for octadecane@SiO2 and eicosane@SiO2 capsules. The difference of release could be exploited for various applications. One could take advantage of the release of large drops (in the absence of surfactant) for sudden release as soon as a given temperature threshold is reached. Applications such as perfume release or temperature history control are based on such an approach. On the contrary, the formation of very small drops containing a water insoluble monomer or initiator could be used in order to provoke a controlled and spatially uniform DOI: 10.1021/la902828q
1741
Article
Destribats et al.
release to start polymerization. All the experiments of release were done at rest (quasi-static conditions with a slow temperature increase) or with a soft hydrodynamic flux. We also investigated the effects of hydrodynamic flux and rapid increase of temperature. These two nonequilibrium external fields allow faster oil release but do not change the release mechanism; only the kinetics is affected.
4. Conclusion A novel process is proposed combining Pickering emulsions (emulsions stabilized by solid particles) and sol-gel chemistry in order to elaborate for the first time stable wax@SiO2 core-shell capsules able to open and release their content upon heating. These capsules can hence be stored and used in both dispersed and dried states. The protocol is easy and very general, as it can be applied to various oils: alkanes, block paraffin, and triglycerides. The generalization with this latter kind of oil opens a large field of applications in pharmaceutics, cosmetics, and foods, since triglycerides are the basis of many biocompatible and edible oils. Such an easy protocol presents the advantage of a direct adaptation to mass production, since no specific equipment is necessary.33 A very good control over the release temperature can be achieved, as it is directly related to the nature of the wax in use. We have demonstrated that the temperature of release can easily be controlled by the choice of the oil, that the way the oil is released can be tuned by the choice of the continuous phase, and that the kinetics of release can be accelerated by an external hydrodynamic field. In a further study, we aim to take advantage of emulsion systems to elaborate more complex capsules comprising multiple compartments with a high potential for facilitating multitherapy. Such capsules will be built from multiple emulsions as precursors, (33) Schmitt, V.; Destribats, M.; Backov, R. French Patent no. FR0955417. (34) Guery, J.; Bertrand, E.; Rouzeau, C.; Levitz, P.; Weitz, D. A.; Bibette, J. Phys. Rev. Lett. 2006, 96, 198301.
1742 DOI: 10.1021/la902828q
and as examples one can mention double emulsions (water-inwax-in-water)34 or even more complex emulsions obtained with millifluidic tools where the number of droplets and the number of different fluids can be tuned on demand.35-39 Supporting Information Available: Films of different capsule types releasing their content in the aqueous phase under thermal treatment. Film 1: Example of a paraffin 4244@SiO2 capsule releasing its content in the aqueous phase, heating rate 5 °C/min from 20 to 100 °C, accelerated 24. Film 2: Eicosane@SiO2 capsules releasing their content in the aqueous phase (SDS, 60 CMC), heating rate 5 °C/min from 33 to 53 °C, accelerated 12. Film 3: Eicosane@SiO2 capsules releasing their content in the aqueous phase (CTAB, 60 CMC), heating rate 5 °C/min from 33 to 53 °C, accelerated 12. Film 4: Eicosane@SiO2 capsules releasing their content in the aqueous phase (Ifralan D205, 60 CMC), heating rate 5 °C/min from 33 to 53 °C, at rest without hydrodynamic flux, accelerated 12. Film 5: Eicosane@ SiO2 capsules releasing their content in the aqueous phase (Ifralan D205, 60 CMC), heating rate 5 °C/min from 33 to 53 °C, with a soft hydrodynamic flux, accelerated 12. Film 6: Example of an eicosane@SiO2 capsule releasing quickly and completely its content in the aqueous phase (Ifralan D205, 60 CMC), heating rate 5 °C/min from 33 to 53 °C, accelerated 12. This material is available free of charge via the Internet at http://pubs.acs.org. (35) Panizza, P.; Engl, W.; Hany, C.; Backov, R. Colloids Surf., A 2008, 312, 24. (36) Engl, W.; Backov, R.; Panizza, P. Curr. Opin. Colloid Interface Sci. 2008, 13, 206. (37) Tachibana, M.; Engl, W.; Usushi, H.; Panizza, P.; Lecommandoux, S.; Backov, R. Chem. Eng. Process. 2008, 47, 1323. (38) Engl, W.; Tachibana, M.; Ushiki, H.; Panizza, P.; Backov, R. Int. J. Multiphase Flow 2007, 33, 897. (39) Utada, A. S.; Lorenceau, E.; Link, D. R.; Kaplan, P. D.; Stone, H. A.; Weitz, D. A. Science 2005, 308, 537.
Langmuir 2010, 26(3), 1734–1742
