908
Langmuir 2001, 17, 908-918
Preparation of Polymeric Nanocapsules by Miniemulsion Polymerization Franca Tiarks, Katharina Landfester,* and Markus Antonietti Max Planck Institute of Colloids and Interfaces, Am Mu¨ hlenberg, 14476 Golm, Germany Received September 6, 2000 This paper reports on the convenient one-step synthesis of hollow polymer nanocapsules by miniemulsion polymerization of different monomers in the presence of larger amounts of a hydrophobe. The idea of the procedure is that the hydrophobe and monomer form a common miniemulsion, whereas the polymer is immiscible with the hydrophobe and demixes throughout polymerization to form the hollow polymer structure surrounding the hydrophobe. The primary objective of this work was to study the effect of different monomers and monomer mixtures, of the type and amount of surfactant, and of the hydrophobe on the morphological characteristics of the polymer/oil composite particle, using dynamic light scattering, scanning and transmission electron microscopies, and atomic force microscopy. It was found that the structure can be adjusted to cover the whole range from independent particles over partially engulfed structures to structurally integer nanocapsules of high uniformity.
Introduction The control of the morphology of latex particles has been an important area in polymer science. Technology has advanced such that a variety of structured particles are nowadays accessible including core-shell, microdomain, and interpenetrating network latexes. Synthetic methods leading to polymer particles having cavities and voids have also been extensively investigated. Here, we are interested in the special case of a hollow sphere or nanocapsule structure. Hollow latex particles can serve as synthetic pigments, which contribute to the opacity of architectural coatings by scattering light.1,2 Additionally these particles can enhance the gloss of paper coatings by influencing the mechanical properties of these high pigment volume concentration formulations during the calendering of the coating.3,4 The potential value of polyalkylcyanoacrylate nanocapsules has been discussed for a variety of pharmaceutical applications, such as a medium for controlled release of calcitonin,5 for peroral administration of insulin,6 or as a part of blood substitutes.7 Nanocapsules are generally considered as spherical, hollow structures with an average diameter smaller than 1 µm.8 Typically the capsule consists of a polymeric wall with a thickness in the nanometer region, filled with an oil which can dissolve lipophilic agents. To enable a stable dispersion, the capsule surface is stabilized by surface charges or by adsorption of an amphiphile. Up to now, predominantly the formation of capsules with a size of 1 µm and larger is described. However, for many applications, especially in medicine and high(1) Strauss, J. Surf. Coat. Aust. 1987, 24, 6. (2) Dowling, D.; Grange, B. Lestarquit Congr. FATIPEC 1986, 1/A, 117. (3) Kaji, T.; Kami, P. Gikyoshi 1992, 46, 271. (4) Kowalski, A.; Vogel, M.; Blankenship, R. M. US Patent 4,427,836; 1984. (5) Tasset, C.; Barrete, N.; Thysman, S.; Ketelegers, J. M.; Lemoine D.; Preat, V. J. Controlled Release 1995, 33, 23. (6) Damge, C.; Michel, C.; Aprahamian, M.; Couvreuer, P.; Devissaguet, J. J. Controlled Release 1990, 13, 233. (7) Chang, T. M. S. Eur. J. Pharm. Biopharm. 1998, 45, 3. (8) (a) Al Khouri Fallouh, N.; Roblot-Treupel, L.; Fessi, H.; Devissaguet, J. P.; Puisieux, F. Int. J. Pharm. 1986, 28, 125-132. (b) Chouinard, F.; Kann, F. W.; Leroux, J.-C.; Foucher, C.; Lenaerts, V. Int. J. Pharm. 1991, 72, 211. (c) Gallardo, M.; Couarraze, G; Denizot, B.; Treupel, J.; Couvreur, P.; Puisex, F. Int. J. Pharm. 1993, 100, 55.
resolution electronic inks, smaller capsules between 50 and 300 nm are of high interest. The approach for the synthesis of nanocapsules described below is based on the principle of miniemulsion using the differences of interfacial tension and the phase separation process during polymerization to obtain a nanocapsule morphology. Direct miniemulsions are aqueous dispersions of relatively stable oil droplets with a size in the 50-500 nm region prepared by shearing a system containing oil (e.g., a water-insoluble monomer), water, a surfactant, and a highly water-insoluble compound (hydrophobe) to suppress Ostwald ripening of the droplets.9 The polymerization of such miniemulsions extends the possibilities of the widely applied emulsion polymerization and provides advantages with respect to incorporation of hydrophobic compounds and composition of the oil phase.10 As only a small surfactant concentration is needed to stabilize the miniemulsion droplets efficiently, no free micelles are existing. In this paper, a miniemulsion polymerization is described that yields an encapsulation of a nonsolvent hydrocarbon by the polymer being formed. By this process, it is possible to prepare latex particles having voids with facile control of the particle diameter, void fraction, and structure. The process initially involves polymerizing a monomer in a dispersed hydrocarbon-monomer mixture which phase separates during the polymerization. This phase-separated polymer subsequently serves as a locus for polymerization. The morphology of the demixing structure is determined by the type of surfactant chosen, the polarity of the monomer, and the choice of hydrophobe which modify either the polymer/hydrocarbon or water/ polymer interface and determine the degree of engulfing. The principle of the encapsulation process is shown in Figure 1. Theory of Droplets Composed of Binary Mixtures. Pioneer work considering the theoretical prediction of three-phase interactions in shear and electrical fields was (9) Sudol, E. D.; El-Aasser, M. S. In Emulsion Polymerization and Emulsion Polymers; Lovell, P. A., El-Aasser, M. S., Eds.; Chichester, 1997; p 699. (10) Landfester, K.; Bechthold, N.; Tiarks, F.; Antonietti, M. Macromolecules 1999, 32, 5222.
10.1021/la001276n CCC: $20.00 © 2001 American Chemical Society Published on Web 01/06/2001
Polymeric Nanocapsules
Langmuir, Vol. 17, No. 3, 2001 909
Figure 1. Schematic view of the encapsulation reaction.
published by Torza and Mason.11 They proposed that the resulting equilibrium configuration of two immiscible liquid drops, designated as phase 1 and phase 3, suspended in a third immiscible liquid, phase 2, and brought into contact is readily predicted from the interfacial tensions σij and spreading coefficients Si ) σjk - (σij + σik). In terms of the convention σ12 > σ23 (S1 < 0) phase 1 is completely engulfed by phase 3 when S2 < 0 and S3 > 0, no engulfing occurs when S2 > 0 and S3 < 0; and S1, S2, S3 < 0 leads to partial engulfing and formation of two-phase droplets with three interfaces, the shapes of which can be calculated. The mechanism of engulfing was established with the aid of high-speed cinematography and shown to involve two competitive processes: penetration and spreading. For simplicity it was assumed that the final equilibrium state is determined solely by the three interfacial tensions σ12, σ13, and σ23. The equilibrium state of three phases of equal density will be that which has a minimum free surface energy Gs ) ∑σijAij, where Aij is the area of the ij interface in the configuration. A simple analysis which yields the same results as that based on minimizing Gs can be made from the consideration of the three spreading coefficients
Si ) σjk - (σij + σik)
(1)
If one adopts the convention of designating phase 1 to be that for which σ12 > σ23, it follows that S1 < 0. Inspection of eq 1 shows that there are only three possible sets of values of Si corresponding to the three different equilibrium configurations illustrated in Figure 2 which are (a) complete engulfing, (b) partial engulfing, and (c) nonengulfing. The description of the expected particle morphology is a system with a complex parameter field. Recognizing the dramatic effect that common emulsifiers have on the interfacial tension between water and organic liquids or solids, it is not surprising to find that the preferred particle (11) Torza, S.; Mason, S. G. J. Colloid Interface Sci. 1970, 33, 6783.
Figure 2. Possible equilibrium configurations corresponding to the three sets of relations for Si. The medium is phase 2.
morphology reacts sensitively on the chemical natures of the emulsifier, the polymer, and the oil. It is obvious that the development of the final morphology in polymer microparticles involves the mobility or diffusion of at least two molecular species influenced by some driving force to attain the phase-separated structure. The ease of movement may be related to the phase viscosity, but in this approach the main emphasis is on the driving force which is the Gibbs free energy change of the process. An analysis of the thermodynamics of two-stage particle formation has been developed by Sundberg et al.12 in which the system was considered simply in terms of the free energy changes at the interface of a three-phase system (i.e., in their case polymers 1 and 2 and water) based on G ) ∑σijAij, where G is the Gibbs’ free energy of the system. According to this analysis, each particular morphological configuration will have a different value for G, and the arrangement with the minimal free energy will be the one which is thermodynamically favored. Sundberg also discussed13 that the influence of the surfactant and the nature of the incompatible polymers on polymer particle (12) (a) Berg, J.; Sundberg, D.; Kronberg, B. Polym. Mater. Sci. Eng. 1986, 64, 367. (b) Berg, J.; Sundberg, D.; Kronberg, B. Microencapsulation 1989, 6, 327.
910
Langmuir, Vol. 17, No. 3, 2001
morphology are seen through their individual and collective effects upon the interfacial tensions. Several apparently different morphologies (hemispherical, sandwich, multiple lobes) have been found to coexist at the same time within a single emulsion, suggesting that they may be simply different states of phase separation and not thermodynamically stable, unique morphologies. Chen and co-workers developed a thermodynamically based model to describe the free energy differences between different possible particle structures.14 For the preparation of polymer particles in aqueous media consisting of two phases, the interfacial energies can for example be influenced by the following parameters: the differences in the hydrophilicity of the monomers and the polymers, and the solubility of the monomers and polymers in the aqueous phase;15,16 the compatibility of the formed polymers;17 type and amount of initiator;18 temperature.19 Current Status of Nanocapsule Synthesis by Polymerization Techniques. The formation of nanocapsules was achieved by a variety of approaches. One of the earliest processes for making hollow latex particles was developed in the research laboratories of the Rohm and Haas Co.20 Their concept involved making a structured particle with a carboxylated core polymer and one or more outer shells. The ionization of the carboxylated core with base under the appropriate temperature conditions expands the core by osmotic swelling to produce hollow particles with water and polyelectrolyte in the interior. In addition to this approach, a number of alternative processes have also been patented that are complex in terms of process stages and chemistry.21 McDonald et al. found that the modification of an emulsion polymerization with a water-miscible alcohol and a hydrocarbon nonsolvent for the polymer can influence the morphology and enables the formation of monodisperse particles with a hollow structure or diffuse microvoids.22 Both kinetic and thermodynamic aspects of the polymerization dictate particle morphology. Complete encapsulation of the hydrocarbon occurs provided low molecular weight polymer is formed initially in the process. Monodisperse hollow particles with diameters from 0.2 to 1 µm were obtainable, and void fractions as high as 50% are feasible. Another well-established method is the preparation of alkylcyanoacrylate nanocapsules, the special choice of monomer yielding in thinner capsule walls and generally in a more reproducible capsule structure. The sizes of capsules prepared in the described manner depend on the (13) Sundberg, D. C.; Casassa, A. P.; Pantazopoulos, J.; Muscato, M.; Kronberg, B.; Berg, J. J. Appl. Polym. Sci. 1990, 41, 1425. (14) Chen, Y.-C.; Dimonie, V.; El-Aasser, M. S. J. Appl. Polym. Sci. 1991, 42, 1049. (15) Okubo, M.; Yamada, A.; Matsumoto, T. J. Polym. Sci., Polym. Chem. Ed. 1980, 16, 3219. (16) Muroi, S.; Hashimoto, H.; Hosoi, K. J. Polym. Sci., Polym. Chem. Ed. 1984, 22, 1365. (17) Okubo, M.; Kanaida, K.; Matsumoto, T. Colloid Polym. Sci. 1987, 265, 876. (18) Okubo, M.; Katsuta, Y.; Matsumoto, T. J. Polym. Sci., Polym. Lett. Ed. 1982, 20, 45. (19) Lee, S.; Rudin, A. J. Polym. Sci., Polym. Chem. Ed. 1992, 30, 2211. (20) (a) Kowalski, A.; Blankenship, R. US Patent 4,468,498, 1984. (b) Kowalski, A.; Vogel, M. US Patent 4,469,825, 1984. (c) Blankenship, R. M.; Kowalski, A. US Patent 4,594,363, 1986. (d) Kowalski, A.; Vogel, M. US Patent 4,880,842, 1989. (21) (a) Nippon Zeon KK, Japanese Patent 052779409 A, 1993. (b) Kaino, M.; Takagishi, Y.; Toda, H. US Patent, 5,360,827, 1994. (c) Nippon Zeon KK, Japanes Patents 07021011, 1995. (22) McDonald, C. J.; Bouck, K. J.; Chaput, A. B.; Stevens, C. J. Macromolecules 2000, 33, 1593.
Tiarks et al.
concentrations of the oil and the monomer components.23 Capsules from polyalkylcyanoacrylates synthesized by interfacial polymerization were initially described by Florence, Whateley, and Wood.24 In the meantime, various techniques for the preparation of polyalkylcyanoacrylate nanocapsules have been studied.25a-c Berg et al. reported the preparation and evaluation of microcapsules formed by the polymerization of methyl methacrylate in the presence of an oil/water macroemulsion. The oil phase was composed of an alkane (e.g., decane or hexadecane), and the oil/water emulsions were stabilized by a variety of emulsifiers.26a,b Both oil- and watersoluble initiators were used, and the monomer was introduced by either dissolving in the oil or feeding it through the water phase. The authors view this system to be an opportunity to study morphological characteristics of polymeric microparticles in the 1-100 µm size range. Okubo et al. examined the penetration/release behavior of various solvents into/from the interior of micrometersized monodisperse cross-linked polystyrene/polydivinylbenzene composite particles.27 The hollow particles were produced by the seeded polymerization utilizing the dynamic swelling method.28 Itou et al. prepared crosslinked hollow polymer particles in the submicrometer size by means of a seeded emulsion polymerization.29 The morphology of the particles depends on the composition of divinylbenzene and methyl methacrylate. Nanocapsules made by the polymerization of rather complex entities were made by Lui et al.,30 who stabilized block copolymer vesicles made of polyisoprene-b-poly(2cinnamoylethyl methacrylate) diblock copolymers by UV cross-linking. A similar approach using block copolymer building blocks was used by Meier who generated nanocapsules by cross-linking polymerization of ABA triblock copolymer vesicles,31 the size of which can be controlled in the range of 50 nm up to 500 nm. Due to their crosslinked structure, both the Liu and the Meier nanocapsules are shape persistent even after their isolation from the aqueous solution. Feldheim et al. used gold particles as templates for the synthesis of hollow polypyrolle capsules.32 Etching of the gold leaves a structurally intact hollow polymer capsule with a shell thickness governed by polymerization time (5-100 nm) and a hollow core diameter dictated by the diameter of the template particle (5-200 nm). Microencapsulation of peptides and proteins is achieved by preparing microcapsules by using a double emulsion technique.33 For the induced phase separation method (23) Wohlgemuth, M.; Ma¨chtle, W.; Mayer, C. J. Microencapsulation 2000, 17, 437. (24) Florence, A. T.; Whateley, T. L.; Wood, D. A. J. Pharm. Pharmacol. 1979, 31, 422. (25) (a) Al Khouri Fallouh, N.; Roblot-Treupel, L.; Fessi, H.; Devissaguet, J. P.; Puisieux, F. Int. J. Pharm. 1986, 28, 125. (b) El-Samaligy, M. S.; Rohdewald, P.; Mahmoud, H. A. J. Pharm. Pharmcol. 1986, 38, 216. (c) Lescure, F.; Zimmer, C.; Roy, D.; Couvreur, P. J. Colloid Interface Sci. 1992, 154, 77. (26) (a) Berg, J.; Sundberg, D.; Kronberg, B. Polym. Mater. Sci. Eng. 1986, 54, 367. (b) Berg, J.; Sundberg, D.; Kronberg, B. J. Microencapsulation 1989, 3, 327. (27) Okubo, M.; Minami, H.; Ynamoto, Y. Colloids Surf., A 1999, 153, 405. (28) Okubo, M.; Shiozaki, M.; Tsujihiro, M.; Tsukuda, Y. Colloid Polym. Sci. 1991, 269, 222. (29) Itou, N.; Masukawa, T.; Ozaki, I.; Hattori, M.; Kasai, K. Colloids Surf., A 1999, 153, 311. (30) Stewart, S.; Liu, G. J. Chem. Mater. 1999, 11, 1048. (31) Nardin, C.; Hirt, T.; Leukel, J.; Meier, W. Langmuir 2000, 16, 1035. (32) Marinakos, S. M.; Novak, J. P.; Brousseau, L. C.; Blaine House, A.; Edeki, E. M.; Feldhaus, J. C.; Feldheim, D. L. J. Am. Chem. Soc. 1999, 121, 8518. (33) Hildebrand, G. E.; Tack, J. W. Int. J. Pharm. 2000, 196, 173.
Polymeric Nanocapsules
Langmuir, Vol. 17, No. 3, 2001 911 Table 1. Characteristics of MMA/HD Latexes with Varying Ratioa
latex JEME 2 JEME 10 JEME 11 JEME 12 JEME 13 JEME 14 JEME 1 JEME 15 JEME 16 JEME 17 JEME 18
monomer (g) MMA MMA MMA MMA MMA MMA MMA MMA MMA MMA MMA
6.01 5.90 5.00 4.50 4.03 3.49 3.00 2.50 1.99 1.50 1.00
hydrophobe (g) HD HD HD HD HD HD HD HD HD HD HD
0.25 0.10 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00
solid content (%)
dl (nm)
Mw (g/mol)
16.09 16.18 15.26 14.86 13.35 9.84 8.37 6.45 5.48 5.12 2.29
73 71 87 107 114 124 142 139 156 147 162
1.06 × 106 8.7 × 105 8.0 × 105 7.3 × 105 7.7 × 105 6.3 × 105 5.5 × 105 3.5 × 105 3.5 × 105 2.9 × 105 8.5 × 104
a The concentration of the initiator AIBN in MMA was kept constant at 22 mg of AIBN/g of MMA. 30 mL of water and 250 mg of SDS were used.
the aqueous drug solution was intensively mixed with the organic polymer solution while an aqueous surfactant solution is added slowly to the oil-in-water (O/W) emulsion. The obtained water-in-oil-in-water (W/O/W) emulsion is stirred under partial vacuum conditions until the organic solvent was removed and the microcapsule was built up. Experimental Part Materials. The monomers styrene and methyl methacrylate (MMA) (both Fluka) and acrylic acid (AA) (Merck) were distilled before use. Sodium dodecyl sulfate (SDS), potassium persulfate (KPS), and 2,2′-azobisisobutyronitrile (AIBN) (all Fluka), hexadecane (HD) (Aldrich), and ethylene glycol dimethacrylate (EGDMA) (Merck) were used as received. SE 3030 (Goldschmidt) is a block copolymer consisting of a polystyrene block (3000 g‚mol-1) and a poly(ethylene oxide) block (3000 g‚mol-1). Lutensol AT 50 (BASF) is a hexadecyl-modified poly(ethylene glycol) (C16H33)(EO)50. PEGA 20034 represents a poly(ethylene oxide)azo initiator (azodi(poly(ethylene glycol) isobutyrate)) with an ethylene oxide molecular weight of 200 g‚mol-1. Synthesis of the Latexes. Six grams of the monomer including the respective amount of hydrophobe was mixed with the initiator (in the case of the oil-soluble AIBN) and added to a solution of different amounts of a surfactant in 30 g of water. After the mixture was stirred for 1 h, miniemulsification was achieved by ultrasonicating the mixture for 120 s with a Branson sonifier W450 Digital at 90% amplitude. To avoid polymerization due to heating, the mixture was cooled in an ice-bath during homogenization. The polymerization was started by heating to 68 °C. If the water-soluble initiators (KPS and PEGA) were used for the synthesis, the initiator was added after miniemulsification. Analytical Methods. The particle size analyses were conducted using a Nicomp particle sizer (model 370, PSS, Santa Barbara, CA) at a fixed scattering angle of 90°. The polymer molecular weights were determined by gel permeation chromatography (GPC) analysis performed on a P1000 pump with UV1000 detector (λ ) 260 nm) (both from Thermo Separation Products) with 5 µm 8 × 3000 mm SDV columns with 106, 105, and 103 Å from Polymer Standard Service in THF with a flow rate of 1 mL min-1 at 30 °C. The molecular weights were calculated with a calibration relative to PS standards. The samples were dried and redissolved in THF before analysis. Transmission electron microscopy was performed with a Zeiss 912 Omega electron microscope operating at 100 kV. The diluted samples were mounted on 400-mesh carbon-coated copper grids and left to dry. The MMA samples were then “coated” with C/Pt in order to increase the stability in the electron beam. No further contrasting was applied. All measurement regarding the surface tension were performed with the K10 processor-tensiometer from Kru¨ss employing the DuNo¨uy-Ring method. The radius of the Pt-Ir ring RI12 was 9.45 mm, and the wire had a radius of 0.185 mm. Scanning force microscopy (SFM) was performed with a NanoScope IIIa microscope (Digital Instruments, Santa Barbara, CA) operating in tapping mode. The instrument was equipped (34) Tauer, K. Polym. Adv. Technol. 1995, 6, 435.
with a 10 × 10 micrometer E-Scanner and commercial silicon tips (model TSEP, the force constant was 50 N/m, the resonance frequency was 300 kHz, and the tip radius was smaller than 20 nm). The samples were prepared by allowing droplets of diluted aqueous solution (0.05 wt %) to dry on freshly cleaved muscovite mica surfaces at room temperature. The morphologies were also examined using scanning electron microscopy (SEM, Zeiss DSM 940). Samples were prepared by allowing a droplet of diluted aqueous solution (0.05 wt %) dry on a glass support and were sputter-coated with gold. The solid contents were determined gravimetrically by drying a specified amount of latex.
Results and Discussion (1) The System Methyl Methacrylate/Hexadecane. (1a) Variation of the MMA/HD Ratio. In a first set of experiments, miniemulsion polymerization was performed using MMA as a monomer and hexadecane as a hydrophobic oil where the relative ratio is varied. PMMA is regarded as rather polar (but is not fully water soluble), whereas hexadecane is very unpolar so that the spreading coefficients are of the right order to stabilize a structure in which a hexadecane droplet core is encapsulated by a PMMA shell surrounded by water. The miniemulsions were obtained by mixing MMA and hexadecane together with the hydrophobic, oil-soluble initiator AIBN and miniemulsifying the mixture in an aqueous solution of SDS. The polymerization leads to polymer particles with a narrow particle size distribution (less than 15%) and with the usual low amount of coagulum. In the state of miniemulsion the monomer and the hexadecane are miscible, but phase separation occurs during the polymerization process due to the immiscibility of hexadecane and PMMA. Since MMA is rather water soluble (150 mM‚L-1), the use of the hydrophobic initiator AIBN is advised, otherwise secondary nucleation of pure PMMA particles from the water phase may occur. The characteristics of the particles are summarized in Table 1. JEME 2 is a standard miniemulsion recipe used as a reference where the hexadecane is only used as osmotic stabilizing agent in order to keep the droplets stable. The particle diameter is 73 nm, the molecular weight Mw of the polymer forming the particle is as high as 106 g‚mol-1. The relatively small quantity of hexadecane is distributed homogeneously in the particles as previously shown by small-angle neutron scattering (SANS) measurements.35 With increasing amount of hexadecane covering ratios between MMA:HD of 24:1 to 1:5 nanocapsules are formed. The particle diameter increases to 160 nm, and the molecular weight of the constituting polymer chains decreases by 1 order of magnitude to 85 000. The solid (35) Landfester, K.; Bechthold, N.; Fo¨rster, S.; Antonietti, M. Macromol. Rapid Commun. 1999, 20, 81.
912
Langmuir, Vol. 17, No. 3, 2001
Tiarks et al.
Figure 3. TEM photographs of nanocapsules with MMA to HD ratios of 5:1 and 3:3 (JEME11 and JEME 1). Table 2. Latexes with a Constant Ratio MMA/HD of 3:3, but at Different Amounts of the Surfactants SDS and Lutensol AT 50a latex JEME 3 b JEME 1 b JEME 3 c FTME 284 b JEME 5 b JEME 4 a
monomer (g) MMA MMA MMA MMA MMA MMA
3.0 3.0 3.0 3.1 3.0 3.0
hydrophobe (g) HD HD HD HD HD HD
3.0 3.0 3.0 3.0 3.0 3.0
surfactant (g) SDS SDS SDS SDS Lut AT 50 Lut AT 50
0.49 0.25 0.10 0.04 0.50 0.25
solid content (%)
dl (nm)
12.9 12.8 11.2 10.4 7.23 4.19
139 142 151 150 223 231
σ (mN/m)
Mw (g/mol)
40.9
5.7 × 105 5.5 × 105 5.9 × 105
55.9 37.4
9.2 × 104 4.5 × 105
0.07 g of AIBN and 24 g of water are used.
contents determined after polymerization match well with the weighted MMA contents; for ratios of 5:1 to 4:2 (JEME 11, 12, and 13) the solid contents are higher than expected due to nonevaporated hexadecane entrapped in the particles. The transmission electron microscopy (TEM) pictures in Figure 3 depict the morphologies of the products with ratios of MMA to HD of 5:1 and 1:1. Please note that PMMA usually degrades in the electron beam. Therefore, the particles were protected by a carbon film, which suppresses degradation, but cannot fully avoid it. In the case of JEME 11 (5:1 ratio), nanocapsules are detected. Since the particle size of the single particles corresponds very well with the size obtained by light scattering, the multiblobs are probably a preparation artifact. In latex JEME 1, nanocapsules are also detected. Comparing the particle size of the single particles with the size obtained by dynamic light scattering (DLS), the particle size obtained by DLS is larger than obtained by TEM. For the preparation of the TEM sample, the hexadecane is evaporated. This can lead to a visual shrinkage of the nanocapsules. With increasing hexadecane content, a decrease of the shell thickness is detected. Nanocapsules with a higher shell stability can be obtained by using up to 10 wt % of EGDMA as cross-linking agent. The particle size of the hollow latexes is not affected by the cross-linker. (1b) Variation of Type and Amount of Surfactant at Constant MMA/HD Ratio. In a next set of experiments, the ratio of MMA to HD was kept constant, but the amount of the anionic surfactant SDS or the nonionic surfactant Lutensol AT 50 (a hexadecyl-modified poly(ethylene glycol)) was varied. The characteristics of the latexes are summarized in Table 2. With a decrease in the amount of SDS from 0.5 to 0.04 g, the size of the particles
stays at about 140-150 nm and does not increase as would be expected for a miniemulsion composed of a purely hydrophobic oil. This turned out to be a fortunate experimental situation: with decreasing SDS content, the interfacial tension at the droplet/water interface increases due to a lower occupation by surfactant molecules. That way, the influence of one of the interfacial tensions on particle morphology can be examined. TEM shows that engulfed structures are obtained in all cases. As depicted in Figure 4a, a low concentration of SDS and the resulting higher interfacial energies between PMMA and water lead to the formation of capped particles, which sometimes look like partial moons and sometimes like disintegrated hulls. As illustrated in Figure 5, the sizes of the PMMA fragments depend on the original morphology: Assuming that all fragments have the same weight, half moon fragments are compact and appear therefore smaller than the fragment of nanocapsules. With Lutensol AT 50, the particle size is about 230 nm, independent of the amount of emulsifier, and the same systematic variation can be performed. With less than 0.25 g of Lutensol AT 50, the particle size increases, and aggregates are formed. As seen in Figure 4b at 0.25 g of Lutensol AT 50, there are some nanocapsules but also a significant number of capped particles. Within the surfactant variation, we also tried to use the cationic surfactant CTMA-Cl and another polymeric surfactant, SE3030. Both systems are not suitable for the formation of nanocapsules from PMMA. (2) The System Styrene/Hexadecane. Due to the pronounced difference of polarity of PMMA and hexadecane, the system was obviously very well suited for the preparation of nanocapsules. PMMA however has some drawbacks, as it usually does not go too well with a clean
Polymeric Nanocapsules
Langmuir, Vol. 17, No. 3, 2001 913
Figure 4. Coexistence of nanocapsules and capped particles (a) in the case of low SDS concentration and (b) in the case of Lutensol AT 50. Table 3. Characteristics of Nanocapsules Using Styrene as Shell Material and a Low Concentration of SDS (30 mg SDS)a latex FTME 292 JEME 21 JEME 22 FTME 293 a
monomer (g) styrene styrene styrene styrene
5.8 3.0 2.0 1.5
hydrophobe (g) HD HD HD HD
0.15 3.0 4.0 4.5
surfactant (g) SDS SDS SDS SDS
0.03 0.03 0.03 0.03
solid content (%)
dl (nm)
σ (mN/m)
Mw (g/mol)
16.6 11.7 10.5 4.5
101 149 170 184
65 66 65 61
2.3 × 105 8.1 × 104
22 mg of AIBN/1 g of MMA and 30 g of H2O are used for the synthesis.
Figure 5. Morphology of particles consisting of a polymer (dark) and a liquid (light) before and after evaporation of the liquid.
heterophase polymerization (too polar), and it causes problems in structural analysis with TEM. With more hydrophobic monomers, such as butyl methacrylate or styrene, it is more difficult to create nanocapsules. Changes in the parameters in the Sty/HD system are very critical and will be discussed in detail: ratio of styrene to hexadecane; type and amount of the surfactant; type of initiator; addition of comonomers. (2a) Variation of the Sty/HD Ratio. The characteristics of the latexes resulting from a systematic variation of the ratio styrene/hexadecane are summarized in Table 3. Latex FTME 292 represents a standard latex where hexadecane is only used as an osmotically stabilizing agent. Particles with a size of 100 nm are obtained where due to the small amount of HD no phase separation can be detected by TEM (Figure 6a). Keeping the amount of SDS constant at 30 mg (0.5 wt % related to the organic phase) and changing only the styrene-to-hexadecane ratio,
nanocapsules coexisting with shell fragments as well as with compact solid polystyrene latexes are obtained. With increasing HD ratio, the particle size increases similar to that seen in the MMA/HD system, while molecular weights of the polymers forming the particle from JEME 37 b to i decrease. On comparison of JEME 21 and JEME 22 with a Sty to HD ratio of 3:3 and 4:2, respectively, especially the size of the nanocapsules increases with decreasing Sty ratio (see parts b and c of Figure 6). At a ratio of 1.5:4.5, the PS shells become rather transient and fragile (see Figure 6d). The structure of the nanoparticles depends much stronger on the amount of the emulsifier than in the case of MMA. A variation of the relative styrene/hexadecane composition at 4 wt % surfactant (with respect to the dispersed phase) results in completely different latexes, the characteristics of which are summarized in Table 4. In the TEM, a capsule morphology is not detected (data not shown). From the TEM picture, it cannot be concluded where the hexadecane is situated, but due to the similarity of the particle sizes obtained by light scattering and TEM, it is still located within the particles, presumably in a less well structured spongelike or scrambled-egg morphology. At higher HD concentrations (2:4), stronger particle deformations are found which underlines the high deformability of the mixed structure under evaporation of the hexadecane (data not shown). (2b) Polymeric Nonionic Surfactants. The high sensitivity of the resulting nanoparticle morphology on the amount of SDS makes the application of nonionic polymeric surfactants very promising. When sufficiently long, these molecules exhibit due to mutual steric repulsion a rather high surface area demand per molecule and therefore a rather high interfacial energy being felt by the monomer molecules. We employed both a hexadecyl
914
Langmuir, Vol. 17, No. 3, 2001
Tiarks et al.
Figure 6. Nanocapsules coexisting with solid particles in the system Sty/HD with various Sty to HD ratios. Table 4. Characteristics of Sty/HD Latexes with Varying Ratio at High SDS Concentration (250 mg of SDS)a latex JEME 37 a JEME 37 b JEME 37 c JEME 37 d JEME 37 e JEME 37 f JEME 37 g JEME 37 h JEME 37 i a
monomer (g) styrene styrene styrene styrene styrene styrene styrene styrene styrene
5.9 5.0 4.5 4.0 3.5 2.5 2.0 1.5 1.0
hydrophobe (g) HD HD HD HD HD HD HD HD HD
0.100 1.001 1.506 1.999 2.501 3.505 4.003 4.564 5.001
solid content (%)
dl (nm)
σ (mN/m)
Mw (g/mol)
16.94 16.27 14.62 13.72 12.07 7.67 6.17 4.48 2.59
75 78 104 110 125 164 163 164 167
53.4 52.4 55.0 54.2 52.9 50.8 50.1 50.1 46.7
3.0 × 105 5.2 × 105 5.0 × 105 4.7 × 105 4.2 × 105 3.1 × 105 1.3 × 105 6.6 × 104 6.1 × 104
The concentration of the initiator AIBN in Sty was kept constant at 22 mg of AIBN/g of Sty. 30 mL of water was used.
modified poly(ethylene glycol) (Lutensol AT 50) as well as a block copolymer of polystyrene and poly(ethylene oxide) (SE3030). The latter should besdue to its polystyrene anchorsideally suited for polystyrene capsules. The resulting data are summarized in Table 5. The use of Lutensol as surfactant leads to some capped particles in the case of a higher concentration of surfactant (0.25 g), whereas a lower concentration leads (as already in the MMA/HD system) to instability, aggregation, and large particles. By use of the block copolymer surfactant SE3030, a minimum amount of 0.1 g is required to form capsules. The capsules are rather uniform, but the hexadecane droplet seems not to be always centered in the middle of the particle. With increasing SE3030 concentration, the nanoparticle size decreases. Apart from
closed nanocapsules, other opened morphologies are also found as shown in Figure 7. This speaks for a break up of the hulls at smaller overall droplet sizes. From all these variations, it is concluded that SE3030 is the most appropriate surfactant of our series to synthesize polystyrene capsules. (2c) Variation of the Initiator. Another important parameter is the type of initiator that has a large influence on the interfacial properties of the system but also dictates the dynamics of polymerization and phase separation. All data shown above were obtained with the oil-soluble initiator AIBN. The use of a standard water-soluble ionic initiator, KPS, to start the polymerization however leads for polystyrene in no case to the formation of nanocapsules,
Polymeric Nanocapsules
Langmuir, Vol. 17, No. 3, 2001 915
Table 5. Characteristics of the Particles Containing Styrene and Hexadecane in a 3:3 Ratio Synthesized with Varying Amounts of Lutensol and SE3030a latex FTME 285 FTME 278 FTME 286 FTME 323 FTME 282b a
monomer (g) Sty Sty Sty Sty Sty
3.0 3.0 3.0 3.0 3.1
surfactant (g) Lutensol AT 50 Lutensol AT 50 SE 3030 SE 3030 SE 3030
hydrophobe (g) 0.03 0.25 0.03 0.1 0.25
HD HD HD HD HD
3.0 3.0 3.0 3.0 3.0
solid content (%)
dl (nm)
σ (mN/m)
3.6 8.9 1.7 1.0 8.8
757 214 444 479 296
44.7 15.5 36.8 50.4 52.8
70 mg of AIBN and 30 g of water are used for the synthesis.
Figure 7. Morphology of particles consisting of styrene and hexadecane in a 3:3 ratio synthesized with Lutensol AT 50 and different amounts of SE3030. Table 6. Characteristics of the Latexes Using KPS or PEGA200 as Initiatorsa latex FTME 287 FTME 274 FTME 324 FTME 325 a
monomer (g) styrene styrene styrene styrene
3.1 3.0 3.0 3.0
surfactant (g) SDS SDS SDS SE 3030
0.03 0.25 0.03 0.25
hydrophobe (g) HD HD HD HD
3.0 3.1 3.0 3.0
initiator (g) KPS KPS PEGA 200 PEGA 200
0.07 0.08 0.07 0.07
solid content (%)
dl (nm)
σ (mN/m)
4.8 7.0 7.4 6.6
215 142 184 407
33.9 51.9 49.6 52.3
30 g of water is used for the synthesis.
but mostly spherical homo-PS particles are obtained. For the characteristics of the latexes see Table 6. Starting the polymerization by the nonionic initiator PEGA200 on the other hand improves the structural
perfection of the polystyrene capsule morphology. For SDS, nanocapsules with a size of about 180 nm can be made. The combination of the polymeric SE3030 as surfactant and PEGA200 as initiator leads to rather perfect poly-
916
Langmuir, Vol. 17, No. 3, 2001
Tiarks et al.
Figure 8. TEM, SEM, and AFM (amplitude) photographs of sample FTME325 (Sty/HD, SE3030 as surfactant, and PEGA200 as initiator).
styrene capsules larger than 400 nm as shown in Figure 8 by different microscopic methods (TEM, SEM, and atomic force microscopy (AFM)). In the TEM picture, the nanocapsule structure becomes easily visible. In the SEM, one can see that most of the nanocapsules are collapsed and regular bowls with a continuous uniform shell are obtained. This leads to the deduction that the hexadecane was incorporated in the center of the particle. Some of the hollow spheres are still intact. In the AFM, some collapsed nanocapsules are detected as bowls and some as pancakes. There are also some particles with “explosion holes”. Summarizing, the particles of the set of experiments show a high perfection of morphology which might be due to the fact that PEGA200 probably acts at both sides of the interface. (2c) Addition of a Comonomer. Another very promising way to improve the perfection of the capsules is the addition of a comonomer to the oil phase. Depending on the polarity of the monomer, it will enter one of the two interfaces (polymer/water) or (polymer/oil) and reduce the corresponding interfacial tensions and spreading coefficients.
In a first experiment, the hydrophobic lauryl methacrylate was chosen to minimize the interfacial tension between styrene and the hexadecane phase. It turns out that this change does not have a significant effect on the resulting morphology of the particles, meaning that this interfacial energy is of minor importance since it is already quite low. On the other hand, the slightly more hydrophilic MMA and the very hydrophilic AA affect the interfacial tension of styrene to water, and here, a pronounced influence on the morphology is found. MMA as Comonomer. Varying amounts of MMA were added to the styrene at two different SDS concentrations. The characteristics of the latexes are listed in Table 7. In the case of 0.25 g of SDS, particles with a size of about 130-150 nm are obtained; in the case of less surfactant (0.03 g), the particle size is about 160-190 nm. As described above, no formation of capsules was found for pure styrene at higher amounts of surfactant (0.25 g). Adding MMA, the formation of nanocapsules becomes, more favorable, and from a ratio of Sty to MMA of 1.5:1.5, nanocapsules are seen with the TEM (data not shown). There are also some half-moon structures found, which
Polymeric Nanocapsules
Langmuir, Vol. 17, No. 3, 2001 917
Table 7. Addition of MMA as Comonomer to the System Styrene/Hexadecanea latex FTME 275 FTME 281a FTME 282a FTME 288 FTME 290 FTME 302 FTME 289 a
monomer (g) styrene MMA styrene MMA styrene MMA styrene MMA styrene MMA styrene MMA styrene MMA
2.7 0.3 1.5 1.5 1.0 2.0 2.7 0.3 2.0 1.0 1.5 1.5 0.3 2.7
surfactant (g)
hydrophobe (g)
solid content (%)
dl (nm)
σ (mN/m)
SDS
0.25
HD
3.0
11.4
131
48.9
SDS
0.25
HD
3.0
15.5
139
47.4
SDS
0.25
HD
3.0
15.2
149
45.3
SDS
0.03
HD
3.0
9.1
187
63.8
SDS
0.03
HD
3.0
10.1
163
61.3
SDS
0.03
HD
3.0
13.0
169
62.8
SDS
0.03
HD
3.0
12.8
161
57.7
solid content (%)
70 mg of AIBN and 30 g of water are used for the synthesis.
Figure 9. Copolymer capsules made of MMA/styrene with hexadecane as an oil phase. Table 8. Addition of AA as Comonomer to the System Styrene/Hexadecanea latex FTME 328 FTME 371 FTME 329 FTME 372 FTME 373 a
monomer (g) styrene AA styrene AA styrene AA styrene AA styrene AA
2.7 0.3 2.7 0.3 2.7 0.3 2.97 0.03 2.85 0.15
surfactant (g)
hydrophobe (g)
dl (nm)
σ (mN/m)
SDS
0.25
HD
3.0
15.3
128
41.0
SDS
0.1
HD
3.0
12.9
136
48.4
SDS
0.03
HD
3.0
16.1
155
51.7
SDS
0.03
HD
3.0
12.4
158
62.6
SDS
0.03
HD
3.0
12.3
163
56.8
70 mg of AIBN and 30 g of water are used for the synthesis.
become less favorable with still increasing MMA content. The data for pure MMA were discussed above. At a lower surfactant content of 0.03 g, capsules were already found for pure polystyrene, but always in coexistence with homogeneous PS spheres. With increasing MMA content the number of capsules also increases, and at a Sty to MMA ratio of 1:1, only nanocapsules are found (Figure 9a). The addition of minor amounts of styrene to MMA stabilizes the polymer capsules against electron beam degradation (Figure 9b) and shows that the multiloop morphology found for pure PMMA (see Figure 3) was indeed an artifact due to beam damage. Acrylic Acid as Comonomer. Acrylic acid (AA) has a very high water solubility. Preparing a miniemulsion
containing styrene and AA, there is a partitioning of AA between the droplets and the water phase. During the polymerization, the AA of the water phase is consumed. It is checked by NMR that after reaction no free AA is left in the aqueous phase. For the experiments, relatively low amounts of acrylic acid below 10 wt % are used. Keeping the amount of AA constant at 10 wt % and changing the surfactant amount leads to an increase of particle size with decreasing SDS amount (see Table 8). Keeping the SDS concentration constant at 0.03 g and changing the AA amount does not influence the particle size. This means that already 1 wt % of AA is sufficient to saturate the capsule surface with carboxylic groups, in good agreement with surface area calculations.
918
Langmuir, Vol. 17, No. 3, 2001
Tiarks et al.
discuss the amount of side products in dependence of the reaction parameters and the kinetics.
Figure 10. Addition of 10 wt % AA as comonomer leads to an increase of the number of nanocapsules.
In all cases, TEM shows that the capsules exhibit a significantly improved morphology with a constant capsule thickness; practically no torn-up capsules are found (Figure 10). There is however a minor fraction of small homogeneous polymer latexes, which we attribute to secondary nucleation due to the very high content of watersoluble acrylic acid. Currently, density matching experiments with the analytical ultracentrifuge are performed to determine the structural purity of the samples and to
Conclusion In this paper the synthesis of hollow polymer nanocapsules as a convenient one-step process using the miniemulsion polymerization is described. The nanocapsules are formed by a variety of monomers in the presence of larger amounts of a hydrophobe. The hydrophobe and monomer form a common miniemulsion before polymerization, whereas the polymer is immiscible with the hydrophobe and phase separates throughout polymerization to form particles with a morphology consisting of a hollow polymer structure surrounding the hydrophobe. In this work the effect of different monomers and monomer mixtures, of the type and amount of surfactant and of the hydrophobe on the morphological characteristics of the polymer/oil composite particle, is studied by dynamic light scattering, scanning and transmission electron microscopies, and atomic force microscopy. The differences in the hydrophilicity of the oil and the polymer turned out to be the driving force for the formation of nanocapsules. In the case of PMMA and hexadecane, the pronounced differences in hydrophilicity are suitable for direct nanocapsule formation. In the case of styrene as monomer, the hydrophilicity of the polymer phase has to be adjusted in order to favor the nanocapsule structure, which is done by addition of either an appropriate comonomer or an initiator. Acknowledgment. Financial support by the Max Planck Society and the Fonds der Chemischen Industrie is gratefully acknowledged. LA001276N
