Assembly of Polymeric Micelles into Hollow Microcapsules with

7810

Langmuir 2008, 24, 7810-7816

Assembly of Polymeric Micelles into Hollow Microcapsules with Extraordinary Stability against Extreme pH Conditions Yi Zhu,†,‡ Weijun Tong,† Changyou Gao,*,† and Helmuth Mo¨hwald‡ Key Laboratory of Macromolecular Synthesis and Functionalization, Ministry of Education, and Department of Polymer Science and Engineering, Zhejiang UniVersity, Hangzhou 310027, China and Max-Planck-Institute of Colloids and Interfaces, 14424 Potsdam, Germany ReceiVed April 9, 2008. ReVised Manuscript ReceiVed May 14, 2008 Hollow microcapsules containing polymeric micelles in their walls were fabricated by alternating assembly of poly(allylamine hydrochloride) (PAH) and poly(styrene-b-acrylic acid) (PS-b-PAA) micelles on MnCO3 microparticles followed by sacrificing the templates in acid solution. The successful formation of PAH/micelle multilayers on both planar and curved substrates was confirmed by UV-vis spectroscopy, ellipsometry, and ξ-potential measurements. The PS-b-PAA micelles retained their structure during the whole assembly process. The as-prepared microcapsules showed extraordinary stability against concentrated HCl (37%) and 0.1 M NaOH solutions. No variation in capsule size or shape was observed in acidic solution, while slight swelling and distortion of the capsules took place in alkaline solution. However, these capsules completely recovered their original size and morphology after being incubated in acidic solution again. The microcapsules, in which large voids exist between the micelle grains on the walls, were totally permeable to fluorescein-tagged dextran with an Mw of 2000 kDa. Assembly of additional PAH/poly(sodium 4-styrenesulfonate) multilayers could substantially reduce the permeation of the same molecules. These multicompartmental capsules combine polymeric micelles with multilayer polyelectrolyte microcapsules and could possibly be imparted with multifunctions, thus possibly finding diverse applications in the fields of drug delivery, biosensing, and nanobiotechnology.

Introduction As a versatile way for fabricating ultrathin multilayer films, the layer-by-layer (LbL) assembly technique has gained tremendous attention in the past decades. This technique, first introduced by Iler1 in 1966 and reestablished and refined by Decher2 in the 1990s, employs different interactions3 such as electrostatic forces,4,5 covalent linkages,6,7 hydrogen bonds,8–10 and hydrophobic forces11,12 to deposit various components onto different substrate surfaces, producing ultrathin films with tailored structure, composition, and properties. Recently, this technique has been successfully extended to colloidal particles.13 Hollow microcapsules with defined structure and composition are obtained after removal of the sacrificial colloids.14,15 Different kinds of functional components such as biomacromolecules, nanoparticles, lipids, and photoactive dyes have been successfully incorporated * To whom correspondence should be addressed. Phone: +86-57187951108. E-mail: [email protected]. † Zhejiang University. ‡ Max-Planck Institute of Colloids and Interfaces.

(1) Iler, R. K. J. Colloid Interface Sci. 1966, 21, 569. (2) Decher, G. Science 1997, 277, 1232. (3) Zhang, X.; Chen, H.; Zhang, H. Y. Chem. Commun. 2007, 14, 1395. (4) Fery, A.; Scho¨ler, B.; Cassagneau, T.; Caruso, F. Langmuir 2001, 17, 3779. (5) Yoo, P. J.; Nam, K. T.; Qi, J. F.; Lee, S. K.; Park, J.; Belcher, A. M.; Hammond, P. T. Nat. Mater. 2006, 5, 234. (6) Sun, J. Q.; Wu, T.; Sun, Y. P.; Wang, Z. Q.; Zhang, X.; Shen, J. C.; Cao, W. X. Chem. Commun. 1998, 17, 1853. (7) Such, G. K.; Quinn, J. F.; Quinn, A.; Tjipto, E.; Caruso, F. J. Am. Chem. Soc. 2006, 128, 9318. (8) Stockton, W. B.; Rubner, M. F. Macromolecules 1997, 30, 2717. (9) Wang, L. Y.; Wang, Z. Q.; Zhang, X.; Shen, J. C. Macromol. Rapid Commun. 1997, 18, 509. (10) Sukhishvili, S. A.; Granick, S. J. Am. Chem. Soc. 2000, 122, 9550. (11) Kotov, N. A. Nanostruct. Mater. 1999, 12, 789. (12) Lojou, E.; Bianco, P. Langmuir 2004, 20, 748. (13) Sukhorukov, G. B.; Donath, E.; Davis, S. A.; Lichtenfeld, H.; Caruso, F.; Popov, V. I.; Mo¨hwald, H. Polym. AdV. Tech. 1998, 9, 759. (14) Caruso, F.; Caruso, R. A.; Mo¨hwald, H. Science 1998, 282, 1111. (15) Donath, E.; Sukhorukov, G. B.; Caruso, F.; Davis, S. A.; Mo¨hwald, H. Angew. Chem., Int. Ed. 1998, 37, 2201.

into the microcapsules either on their shells or in their interiors.16–19 Properties and functions, particularly the permeability of these capsules, can be easily adjusted by factors such as pH value, ionic strength, solvent quality, and layer number as well as annealing and resealing after core removal.20–24 All these features have made the LbL microcapsules promising candidates as drug delivery vehicles, microreactors, and biosensors.25 Micelles are thermodynamically stable molecular assemblies of either amphiphilic small molecules or copolymers in selective solvents26 and possess a unique core/shell structure, which can be tailored by molecular design and preparation conditions. Thus, they also may find potential applications in the fields of drug delivery,27 nanofabrication,28 as well as biosensing.29 In most cases, the hydrophobic cores of the micelles in water are employed to load and sustained-release hydrophobic substances. (16) Berth, G.; Voigt, A.; Dautzenberg, H.; Donath, E.; Mo¨hwald, H. Biomacromolecules 2002, 3, 579. (17) Radt, B.; Smith, T. A.; Caruso, F. AdV. Mater. 2004, 16, 2184. (18) Katagiri, K.; Caruso, F. AdV. Mater. 2005, 17, 738. (19) Skirtach, A. G.; Antipov, A. A.; Shchukin, D. G.; Sukhorukov, G. B. Langmuir 2004, 20, 6988. (20) Sukhorukov, G. B.; Antipov, A. A.; Voigt, A.; Donath, E.; Mo¨hwald, H. Macromol. Rapid Commun. 2001, 22, 44. (21) Ibarz, G.; Da¨hne, L.; Donath, E.; Mo¨hwald, H. AdV. Mater. 2001, 13, 1324. (22) Lvov, Y.; Antipov, A. A.; Mamedov, A.; Mo¨hwald, H.; Sukhorukov, G. B. Nano Lett. 2001, 1, 125. (23) Ibarz, G.; Da¨hne, L.; Donath, E.; Mo¨hwald, H. Chem. Mater. 2002, 14, 4059. (24) Georgieva, R.; Moya, S.; Donath, E.; Baumler, H. Langmuir 2004, 20, 1895. (25) Peyratout, C. S.; Da¨hne, L. Angew. Chem., Int. Ed. 2004, 43, 3762. (26) Gohy, J. F. AdV. Polym. Sci. 2005, 190, 65. (27) Ro¨sler, A.; Vandermeulen, G. W. M.; Klok, H. A. AdV. Drug DeliVery ReV. 2001, 53, 95. (28) Selvan, S. T.; Hayakawa, T.; Nogami, M.; Mo¨ller, M. J. Phys. Chem. B 1999, 103, 7441. (29) Dubertret, B.; Skourides, P.; Norris, D. J.; Noireaux, V.; Brivanlou, A. H.; Libchaber, A. Science 2002, 298, 1759.

10.1021/la801119z CCC: $40.75  2008 American Chemical Society Published on Web 07/01/2008

Assembly of Polymeric Micelles

So far, it is still difficult to incorporate hydrophobic substances into the LbL multilayer assemblies in a general and conventional way. Considering their capacity of loading and releasing hydrophobic substances, micelles have been successfully used as building blocks to obtain planar LbL multilayer films. For example, Kataoka and co-workers successfully fabricated multilayer films comprised of covalently bonded polyallylamine and poly(ethylene glycol)-poly(D,L-lactide) (PEG-PLA) micelles.30 Zhang and co-workers employed electrostatic forces between poly(diallyl dimethylammonium chloride) (PDADMAC) and poly(styrene-b-acrylic acid) (PS-b-PAA) micelles to prepare multilayer films for incorporation and controlled release of waterinsoluble substances.31–33 Caruso and co-workers obtained nanoporous multilayer films comprised of only micelles of poly(styrene-b-4-vinylpyridine) (PS-b-PVP) and PS-b-PAA, which showed dual optical properties.34 Although the LbL micelle multilayers have been assembled onto colloidal particle surfaces as smart coatings,35 to the best of our knowledge, hollow microcapsules with micelles as wall components have not been reported yet. Compared to the multilayer films and core/shell particles, the hollow capsules have the advantages of enhanced loading capacity and good mobility, which are critical for certain applications such as drug delivery, thus expanding the potential applications of these hybrid assemblies. Herein, we report the successful fabrication of multicompartmental microcapsules with micelles as one component of the capsule wall. Poly(allylamine hydrochloride) (PAH) and PSb-PAA are chosen as the building blocks because they are both commercially available and extensively studied. Nile red, a hydrophobic fluorescent dye, is used as a model for hydrophobic substances such as anticancer drugs. A model of a multicompartmental encapsulation vehicle for both hydrophilic and hydrophobic substances has been constructed. The extreme stability of these capsules against harsh conditions in aqueous environment was demonstrated, which can broaden their potential applications. It could be anticipated that, in such a system, the micelles could serve as universal hydrophobic reservoirs with high loading capacity in the shells while the LbL shells could provide the micelles with additional support and protection as well as multifunction such as targeting. Besides their potential use as a drug delivery vehicle, considering the tremendous potential of micelles in nanofabrication and biosensing fields, these capsules may also provide an alternative option to serve as bioreactors and biosensors.

Experimental Section Materials. Poly(styrene-b-acrylic acid) (Mn PS(16.5 kDa)-bPAA(4.5 kDa), Mw/Mn ) 1.05) was purchased from Polymer Source Inc. Poly(allylamine hydrochloride) (PAH, Mw 70 kDa), poly(sodium 4-styrenesulfonate) (PSS, Mw 70 kDa), fluorescein isothiocyanate labeled dextran (FITC-dextran, Mw 2000 kDa), Nile red (NR), manganese sulfate hydrate (MnSO4 · H2O), ammoniumhydrogencarbonate (NH4HCO3), and N,N-dimethylformamide (DMF) were obtained from Sigma-Aldrich. Concentrated hydrochloric acid (HCl, fuming 37%, extra pure) was obtained from Merck. All chemicals were used as received. FITC-PAH was prepared by labeling PAH (30) Emoto, K.; Iijima, M.; Nagasaki, Y.; Kataoka, K. J. Am. Chem. Soc. 2000, 122, 2653. (31) Ma, N.; Zhang, H. Y.; Song, B.; Wang, Z. Q.; Zhang, X. Chem. Mater. 2005, 17, 5065. (32) Ma, N.; Wang, Y. P.; Wang, Z. Q.; Zhang, X. Langmuir 2006, 22, 3906. (33) Ma, N.; Wang, Y. P.; Wang, B. Y.; Wang, Z. Q.; Zhang, X.; Wang, G.; Zhao, Y. Langmuir 2007, 23, 2874. (34) Cho, J.; Hong, J.; Char, K.; Caruso, F. J. Am. Chem. Soc. 2006, 128, 9935. (35) Biggs, S.; Sakai, K.; Addison, T.; Schmid, A.; Armes, S. P.; Vamvakaki, M.; Bu¨tu¨n, V.; Webber, G. AdV. Mater. 2007, 19, 247.

Langmuir, Vol. 24, No. 15, 2008 7811 with FITC according to the literature.36 Spherical MnCO3 microparticles with a mean diameter of approximate 5 µm were synthesized by mixing MnSO4 and NH4HCO3 solutions.37 Preparation of PS-b-PAA Micelles. The PS-b-PAA micelles were prepared according to the literature.38 Briefly, a suitable amount of water was added dropwise into 4 mL of 5 mg/mL PS-b-PAA/ DMF solution under vigorous agitation, followed by addition of a large amount of water to “quench” the micelles formed. The final concentration of PS-b-PAA was 0.2 mg/mL, and only 4% (v/v) DMF remained in the solution. To obtain micelles loaded with Nile red, typically 80 µL of 0.5 mg/mL Nile red/acetone solution was added into 10 mL of micelle solution. After sonication for 30 min, the solution was let open overnight to evaporate the acetone.31 The pH of the micelle solution was adjusted to 7 using 0.1 M NaOH or 0.1 M HCl before assembly. Layer-by-Layer Assembly on Planar Substrates. Quartz slides and silicon wafers were used as planar substrates for the LbL assembly. Both substrates were treated with “piranha” solution for 2 h and thoroughly rinsed with water (Caution: piranha solution containing 7:3 (V/V) H2SO4/H2O2 is extremely corrosiVe and should be handled with great care). Then the hydroxyl-tailored substrates were immersed sequentially into PAH solution (2 mg/mL, 0.2 M NaCl) for 15 min and micelle solution (0.2 mg/mL, pH 7) for 30 min with thorough water rinses and N2 drying at each interval. After each micelle layer assembly, the absorbance of the multilayers on quartz slides was monitored with UV-vis spectroscopy and the thickness was measured by ellipsometry. The assembly steps were repeated until six bilayers were assembled on the slides. Layer-by-Layer Assembly on MnCO3 Particles and Hollow Capsule Fabrication. Multilayer build up onto the MnCO3 microparticles was conducted under the same conditions as that on the planar substrates. In brief, a certain amount of MnCO3 microparticles were dispersed sequentially in PAH solution (2 mg/ mL, 0.2 M NaCl) for 15 min and micelle solution (0.2 mg/mL, pH 7) for 30 min. The assembly started with PAH. After adsorption of each layer, the excess assembly components were removed by centrifugation and three washings with water. The adsorption steps were repeated until 11 layers were assembled (PAH being the outmost layer). For permeability control, another 3 or 5 bilayers of PAH/PSS were further adsorbed onto the preassembled (PAH/micelle)5 multilayers on the MnCO3 particles by sequentially dispersing the particles in PAH (2 mg/mL, 0.2 M NaCl) solution and PSS (2 mg/ mL, 0.2 M NaCl) solution for 15 min, followed by centrifugation and three washings. For visualization, one layer of NR-loaded micelles or FITC-PAH was used occasionally. The MnCO3 cores were decomposed by 0.1 M HCl with a membrane filtration apparatus equipped with a cellulose filter having a pore size of 0.45 µm. The capsules obtained were washed until a constant pH value was reached. Stability of Microcapsules. For stability investigation, an equal volume of (PAH/micelle)5/PAH microcapsule suspension and concentrated HCl (fuming, 37%) or 0.1 M NaOH were mixed. The whole process was monitored under CLSM for 30 and 20 min for HCl and 0.1 M NaOH treatments, respectively. The mean diameter of the capsules was determined. Permeability of Microcapsules. To investigate the permeability of different kinds of microcapsules, an equal volume of capsule suspension and FITC-dextran (Mw 2000 kDa, 2 mg/mL) solution were mixed and observation was performed 20 min later under CLSM. Confocal Laser Scanning Microscopy (CLSM). Confocal images were obtained by a Leica confocal scanning system mounted to a Leica Aristoplan and equipped with a 100× oil immersion objective with a numerical aperture (NA) of 1.4. Scanning Electron Microscopy (SEM). The samples were prepared by applying a drop of the particle or capsule suspension onto clean glass slides. After drying overnight, the samples were sputtered with gold and measured by a Gemini Leo 1550 instrument at an operation voltage of 3 keV. (36) Klitzing, R.; Mo¨hwald, H. Langmuir 1995, 11, 3554. (37) Tong, W. J.; Gao, C. Y. Colloids Surf. A 2007, 295, 233. (38) Zhang, L. F.; Eisenberg, A. J. Am. Chem. Soc. 1996, 118, 3168.

7812 Langmuir, Vol. 24, No. 15, 2008

Zhu et al.

Figure 1. (a) UV-vis spectra of PAH/micelle multilayers assembled on quartz slides. (Inset) Absorbance at 218 nm as a function of bilayer number. (b) Thickness of PAH/micelle multilayers on silicon slides as a function of bilayer number, as measured by ellipsometry.

Scanning Force Microscopy (SFM). A drop of capsule suspension was applied to freshly cleaved mica and dried overnight in air. The images were taken with a Digital Instruments nanoscope IIIa Multimode SFM (Digital Instruments Inc., Santa Barbara, CA) in air using a tapping mode at room temperature. UV-vis Spectroscopy. The UV-vis absorption spectra of the multilayers on quartz slides were recorded with a Cary 50 UV-vis spectrophotometer. ξ-Potential Measurements. The ξ-potentials of polystyrene (PS) microparticles (φ 1.37 µm) alternatingly coated with PAH and the micelles were measured with a Zetasizer Nanoinstrument Nano Z in water. Each value was averaged from five parallel measurements.

Results and Discussion Layer-by-Layer Assembly on Planar Substrates. The multilayers of PAH and PS-b-PAA micelles were first assembled on quartz slides to monitor their growing regime by UV-vis spectroscopy (Figure 1a). The absorbance at ca. 218 nm is assigned to the π-π* transition of the benzene ring in the PS block of the micelles.31 The results show that the absorbance increased along with the increase of the bilayer number (inset of Figure 1a), indicating successful formation of PAH/micelle multilayers. The growth regime of larger increment with increasing layer number was further confirmed by the film thickness measured by ellipsometry (Figure 1b). Indeed, similar growth regimes of the multilayers, namely, exponential growth, have been observed before for other systems. Generally, two explanations are given, i.e., the increase in surface roughness of the multilayer film during the assembly39,40 and the diffusing “in” and “out” of the polyelectrolytes at each assembly step.41,42 In the present case, it is reasonable to assume that the main reason should be the increase in the film surface roughness. Micelles are molecular assemblies rather than single molecules; thus, each adsorption cycle of the micelles would make the film surface rougher, which, in turn, could result in a larger amount of micelle and/or PAH adsorption during the next cycle. Hollow Capsule Fabrication. To confirm the alternating adsorption of the building blocks on curved surfaces, microelectrophoresis experiments were conducted using polystyrene (PS) microparticles as templates instead of MnCO3 microparticles to avoid quick sedimentation. As shown in Figure 2, the ξ-potential value for the bare PS particles is -38.7 mV, whereas the values for the particles with PAH as the outmost layer are ca. 40 mV, which are quite normal in comparison with other systems.43 However, the values are -6.5, -2.0, -1.7, 0.4, and 2.0 mV at (39) DeLongchamp, D. M.; Kastantin, M.; Hammond, P. T. Chem. Mater. 2003, 15, 1575. (40) Schoeler, B.; Poptoshev, E.; Caruso, F. Macromolecules 2003, 36, 5258. (41) Picart, C.; Mutterer, J.; Richert, L.; Luo, Y.; Prestwich, G. D.; Schaaf, P.; Voegel, J. C.; Lavalle, P. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 12531. (42) Lavalle, P.; Vivet, V.; Jessel, N.; Decher, G.; Voegel, J.; Mesini, P. J.; Schaaf, P. Macromolecules 2004, 37, 1159.

Figure 2. ξ-Potential of PS particles (φ 1.37 µm) alternatingly coated with PAH (odd) and PS-b-PAA micelles (even) as a function of layer number. The line is to guide the eye and has no physical meaning.

2, 4, 6, 8, and 10 layers, respectively, with the micelles being the outermost layer. This is considerably lower than that of the micelles (-30 mV). One most possible explanation could be the incomplete coverage by the micelles on the preassembled layers. Because the micelles are spherical and there is electrostatic repulsion between the negatively charged PAA chains of neighboring micelles, after adsorption only part of the outermost surface is covered by the micelles, while the rest of the surface, most probably the area between neighboring micelles, is still covered by the positively charged PAH, as can be seen in Figure 3b. This could result in a relatively low overall potential. However, an alternating trend in the ξ-potential values can be observed during the assembly, indicating successful formation of multilayers on the PS microparticles. The assembly of PAH and PS-b-PAA micelles was then conducted on MnCO3 microparticles, whose surface morphology change was monitored by SEM. Figure 3a shows a typical surface morphology of the MnCO3 particles, which is very rough. After assembly of 5 bilayers of PAH/micelles, the particle surface is still rather rough (Figure 3b). Nevertheless, it is visible that the particles are partly covered with grains of several tens of nanometers. The size of the micelles used for assembly was determined by TEM (Figure 3c) as well as DLS (data not shown), giving a mean diameter of ca. 30 nm and ca. 60-70 nm with narrow size distribution, respectively. Considering the possible interference of water to the micelle size, the micelle size obtained by TEM is more reliable in the dry state and consistent with the (43) Radtchenko, I. L.; Sukhorukov, G. B.; Leporatti, S.; Khomutov, G. B.; Donath, E.; Mo¨hwald, H. J. Colloid Interface Sci. 2000, 230, 272.

Assembly of Polymeric Micelles

Langmuir, Vol. 24, No. 15, 2008 7813

Figure 3. SEM images of (a) bare MnCO3 particles, (b) surface of MnCO3/(PAH/micelle)5 particles, and (d) surface of MnCO3/(PAH/micelle)5/PAH particles. All scale bars are 600 nm. (c) TEM image of PS-b-PAA micelles used for LbL assembly.

grain size on the MnCO3 particle surface. Therefore, the grains are most possibly dried micelles or micelle aggregates. Of course the intensified lateral segregation of the micelles caused by the drying process cannot be ruled out at present. Between the grains, large void areas can be observed, which supports the explanation of incomplete coverage above. After further assembly of one layer of PAH, the surface became smooth to some extent with grains sticking to each other or “bridged” together, although the void areas still exist (Figure 3d). During the whole process, PSb-PAA copolymers retained the micelle structure, which was confirmed by the fluorescence from Nile red, a hydrophobic dye preloaded in the micelle cores before multilayer assembly (see Figure 6). After core removal with 0.1 M HCl, hollow and intact microcapsules containing micelles in their shells were obtained. With PAH as the outermost layer, apparent capsule aggregation could be effectively minimized. As shown in Figure 4a, the folds and creases are not as apparent as that of the traditional LbL multilayer microcapsules. Nevertheless, the microcapsules still collapsed due to their thin walls and low mechanical strength. Closer observation reveals that the capsule surface is comprised of grains bound to each other and large pores (Figure 4b), resembling the surface morphology of the core/shell particles before core removal (Figure 3d). SFM measurements (Figure 5) further confirm the above observation. The grains with a size of several tens to hundreds of nanometers are observable, which

are consistent with the results shown in Figures 3d and 4b. A control experiment conducted similarly on silicon wafers showed that the grains of similar size also could be observed (data not shown); thus, the grains should not be intrinsically incurred by the rough MnCO3 particles. As such one can conclude that the grains are most possibly the micelle particles, micelle aggregates, or overlapping micelles in adjacent layers. The wall thickness of these capsules is 57.7 nm by averaging over five measurements. Considering the original size of the micelles before assembly, they should become “flat” to some extent after adsorption onto the substrates and drying. Since the assembly possibly proceeds in a special way in which some of the micelles could adsorb onto the void areas after the previous adsorption cycle, it is difficult and meaningless to calculate the approximate mean thickness of one bilayer in the current case. Stability of Microcapsules. The stability of the microcapsules is a critical issue concerning their potential applications in various circumstances. Here the stability of the (PAH/micelle)5/PAH microcapsules against harsh acidic and alkaline conditions was investigated. As shown in Figure 6, these capsules show extraordinary stability. Incubation of the capsule suspension with an equal volume of concentrated HCl (37%, 12 M) for 30 min did not cause any observable change in either capsule size or appearance (Figure 6a). Since the capsules were obtained using excessive 0.1 M HCl, a similar stability could also be guaranteed under a less acidic condition, which has been adopted more

7814 Langmuir, Vol. 24, No. 15, 2008

Zhu et al.

Figure 4. SEM images of (a) (PAH/micelle)5/PAH microcapsules, (b) magnification of the (PAH/micelle)5/PAH capsule wall, (c) MnCO3/(PAH/ micelle)5/(PAH/PSS)5 particles, and (d) magnified (PAH/micelle)5/(PAH/PSS)5 capsule wall. The scale bar in a is 3 µm while in b-d is 400 nm.

Figure 5. SFM images of (a) a hollow (PAH/micelle)5/PAH microcapsule and (b) magnification of a flat region of the capsule wall in a.

widely. However, incubation with an equal volume of 0.1 M NaOH (the final solution pH 12.6) could cause instantaneously capsule swelling from 5.3 to 6.1 µm within 2-3 min. Meanwhile, the capsule walls became distorted (Figure 6b and 6c). Nevertheless, these capsules retained their integrity instead of quick disassembly, a phenomenon observed for PAA/PAH and

PSS/PAH microcapsules as a result of full decharging of the polyelectrolytes. For instance, the PAA/PAH microcapsules44 and the PAH/PSS capsules45 are reported to be readily dissolved (44) Gao, C. Y.; Mo¨hwald, H.; Shen, J. C. AdV. Mater. 2003, 15, 930. (45) Dejugnat, C.; Sukhorukov, G. B. Langmuir 2004, 20, 7265.

Assembly of Polymeric Micelles

Figure 6. CLSM images of (PAH/micelle)5/PAH microcapsules incubated in (a) concentrated HCl for 30 min, (b) 0.1 M NaOH for 20 min, and (d) 0.1 M NaOH for 20 min followed by addition of concentrated HCl. (c) Transmission image corresponding to b.

in 0.1 M HCl within 10 min and in 0.1 M NaOH (pH 12.5) within seconds, respectively. Moreover, the swelling as well as distortion of the capsules are reversible. Neutralization of the NaOH with excessive HCl caused immediate shrinkage of the capsules within 30 s to recover their original size. The capsule walls regained their smooth morphology under CLSM observation as well (Figure 6d). Also, no apparent change in capsule number was identified in all the cases. These results reveal that although the electrostatic force between the PAH and PAA chains of the micelles may play an important role in assembling the components into capsules, it is not the only force that stabilizes the walls. Additional forces such as hydrophobic force and van der Waals forces should also have a considerable contribution. According to the fact that the layer thickness increment (Figure 1b and SFM results in Figure 5) is at least 5 times smaller than that of the micelle size, it is reasonable to assume that the micelles become more or less “flattened” upon adsorption onto the solid substrates. In this case, some of the successively deposited micelles may interact and even partly fuse with the previously adsorbed ones by hydrophobic interaction, since the complexation of micelle corona chains (PAA) with PAH chains can draw the micelles closer and thereby the attraction between the hydrophobic cores can take place. Also, there is a high chance that the PAA chains in the adjacent micelle layers entangle with each other in acidic condition due to physical entanglement and hydrogen bonding. Therefore, when the capsules are in a highly acidic environment, although the electrostatic force between the PAH and PAA is suppressed because almost all the PAA chains are decharged, the microcapsules can still keep their shell structure intact and the rearrangement of PAH chains can be possibly retarded. When the capsules are incubated in 0.1 M NaOH, the PAH molecules are largely deprotonated while almost all the PAA chains are charged. In this case, the hydrophobic interaction between the micelles in adjacent layers is somehow interrupted by the hydrophilic regions dominated by the charged PAA chains since the size of the micelle corona is relatively large and can no longer be “ignored”. As a result of the interplay between the

Langmuir, Vol. 24, No. 15, 2008 7815

Figure 7. CLSM images of (a) (PAH/micelle)5/PAH, (b) (PAH/micelle)5/ (PAH/PSS)3, and (d) (PAH/micelle)5/(PAH/PSS)5 microcapsules after mixing with FITC-dextran (Mw 2000 kDa) for 20 min. (c) CLSM image of (PAH/micelle)5/(PAH/PSS)5 microcapsules in which one layer of FITC-PAH is used to replace the normal PAH for visualization.

charge repulsion of PAA chains (outgoing force) and the hydrophobic interaction between the micelle cores as well as the hydrogen bonding between the PAH chains (holding force), the capsules can still keep their structure intact but swell to some extent to release the charge repulsion force. It is conceivable that the swelling might be inhomogeneous for different parts of the capsule walls; thus, distortion takes place. The reversible shrinkage of the capsules after neutralization indicates that the swelling degree should be in an “elastic” range, namely, a range within which the PAA chains can reinteract with the PAH chains. Permeability of Microcapsules. The permeability of the (PAH/micelle)5/PAH microcapsules was further investigated by CLSM using FITC-dextran (Mw 2000 kDa) as a probe. After incubation of the capsule suspension with an equal volume of probe solution for 20 min, almost all microcapsules were permeable to the probe molecules judging by the strong fluorescence from the capsule interiors (Figure 7a). This is consistent with the results in Figure 4b in which large pores could be observed in the capsule walls. However, semipermeability, implying the ability to load and hold substances, is of great importance if the microcapsules serve as potential drug delivery vehicles. Therefore, these pores need to be “sealed”, at least partially. For this context, 3 or 5 additional PAH/PSS bilayers were assembled onto the MnCO3 particles coated with (PAH/ micelle)5 multilayers. Successful adsorption of the additional PAH/PSS multilayers is confirmed by the fluorescence from FITC-PAH used in the second bilayers for visualization (Figure 7c). As shown in Figure 4c, after assembly of 5 additional bilayers the particle surfaces became denser and fewer pores could be found, in contrast to their precursors (Figure 3b and 3d). After core removal, hollow and intact capsules with fewer “defects” were obtained. The permeability investigation showed that about 30-40% of the capsules were impermeable to the FITC-dextran if three additional bilayers were assembled (Figure 7b), while more than 90% of the (PAH/micelle)5/(PAH/PSS)5 capsules were impermeable (Figure 7d) to the probe molecules. Thus, assembly of additional PAH/PSS multilayers is effective in imparting the

7816 Langmuir, Vol. 24, No. 15, 2008

capsules semipermeability. It should be noted here that some pores can still be seen on either the (PAH/micelle)5/(PAH/PSS)5 core/shell particles (Figure 4c) or the capsules (Figure 4d). Nevertheless, from permeability investigation one can conclude that the defects in the capsule walls have been effectively “sealed”. The remaining pores probably only indicate that the capsule surfaces are still rough considering the difference in the sizes of the micelles and the polyelectrolytes used to “seal” the capsules. Also, it cannot be ruled out that some polyelectrolytes may exist between the pores, which have an impact on reducing the size of permeable molecules like dextran with an Mw of 2000 kDa and also possibly create charge repulsion between the negative capsule surface (PSS as the outmost layer) and the slight negative FITC-dextran.46,47

Conclusion We demonstrate in this work the successful fabrication of multilayer microcapsules using PAH and PS-b-PAA micelles as the building blocks and MnCO3 microparticles as the sacrificial template. Multilayer growth is confirmed using UV-vis spectroscopy and ellipsometry by depositing the multilayers on quartz slides and silicon wafers, respectively. Zeta-potential measurement confirms the assembly process on curved surfaces. Hollow and intact microcapsules with intact micelles as one of the wall components are then obtained after core removal. These capsules (46) Tong, W. J.; Dong, W. F.; Gao, C. Y.; Mo¨hwald, H. J. Phys. Chem. B 2005, 109, 13159. (47) Tong, W. J.; Song, H. Q.; Gao, C. Y.; Mo¨hwald, H. J. Phys. Chem. B 2006, 110, 12905.

Zhu et al.

show extraordinary stability against highly acidic or alkaline conditions possibly because of the hydrophobic interaction between the PS cores of the micelles as well as hydrogen bonding of the PAA chains in adjacent layers and the PAH chains. The capsules are permeable to FITC-dextran with an Mw of 2000 kDa, but assembly of additional PAH/PSS multilayers can reduce the permeation of the same molecules. The capsules obtained could serve as models for multicompartmental containers which may encapsulate both hydrophobic and hydrophilic substances. The extreme stability of the capsules against harsh conditions in an aqueous environment could broaden their potential applications, while their responsiveness to external stimuli could still be realized at high pH. It might also be possible to tailor the release of substances in each compartment by proper molecular design of the micelles and incorporation of various functional components, which would make them promising candidates as drug delivery vehicles, microreactors, and microsensors. Acknowledgment. We thank Professor J. C. Shen for his continuous support and stimulating discussions. A. Heilig, R. Pitschke, L. Hu, and H. Zastrow are greatly acknowledged for their assistance in SFM, SEM, TEM, and ξ-potential measurements. Y.Z. thanks the Max-Planck Society for a visiting scientist grant. This study was financially supported by the Natural Science Foundation of China (20434030, 20774084), the Major State Basic Research Program of China (2005CB623902), and the National Science Fund for Distinguished Young Scholars of China (50425311). LA801119Z