Preparation of Monodisperse Chitosan Microcapsules with Hollow

Aug 18, 2010 - †Department of Chemical System Engineering, The University of Tokyo, ... Bunkyo-ku, Tokyo 113-0033, Japan, §Cooperative Support Team...
1 downloads 0 Views 3MB Size
pubs.acs.org/Langmuir © 2010 American Chemical Society

Preparation of Monodisperse Chitosan Microcapsules with Hollow Structures Using the SPG Membrane Emulsification Technique )

Kazuki Akamatsu,*,†,^ Wei Chen,† Yukimitsu Suzuki,‡ Taichi Ito,‡ Aiko Nakao,§ Takashi Sugawara,† Ryuji Kikuchi,† and Shin-ichi Nakao

)

† Department of Chemical System Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan, ‡Center for Disease Biology and Integrative Medicine, Faculty of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan, §Cooperative Support Team, RIKEN, ASI, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan, and Department of Environmental and Energy Chemistry, Faculty of Engineering, Kogakuin University, 2665-1 Nakano-machi, Hachioji-shi, Tokyo 192-0015, Japan. ^Present address: Department of Environmental and Energy Chemistry, Faculty of Engineering, Kogakuin University 2665-1 Nakano-machi, Hachioji-shi, Tokyo 192-0015, Japan.

Received May 17, 2010. Revised Manuscript Received July 15, 2010 We describe herein successful preparations of monodisperse chitosan microcapsules with hollow structures using the SPG membrane emulsification technique. Two preparation procedures were examined in this study. In the first method, monodisperse calcium alginate microspheres were prepared and then coated with unmodified chitosan. Subsequently, tripolyphosphate treatment was conducted to physically cross-link chitosan and solubilize the alginate core at the same time. In the second method, photo-cross-linkable chitosan was coated onto the monodisperse calcium alginate microspheres, followed by UV irradiation to chemically cross-link the chitosan shell and tripolyphosphate treatment to solubilize the core. For both methods, it was determined that the average diameters of the chitosan microcapsules depended on those of the calcium alginate microparticles and that the microcapsules have hollow structures. In addition, the first physical cross-linking method using tripolyphosphate was found to be preferable to obtain the hollow structure, compared with the second method using chemical cross-linking by UV irradiation. This was because of the difference in the resistance to permeation of the solubilized alginate through the chitosan shell layers.

Introduction Chitosan (poly(1,4-β-D-glucopyranosamine)) is one of the bestknown polysaccharides, formed by deacetylation of chitin (poly(1,4-β-N-acetyl-D-glucosamine)), the second most abundant polysaccharide in nature.1,2 Owing to its biodegradability and biocompatibility, various biomaterials using chitosan and its derivatives have been developed,3,4 especially in pharmaceutical applications. Among these, hollow spheres or containers of nanometer or micrometer scale have attracted increasing interest as drug delivery systems.5-8 This is because they are capable of encapsulating large-sized drugs or large quantities of guest molecules within the empty cores. To obtain such hollow capsules, several methods can be employed, including the interfacial polymerization method, phase *Corresponding author: Tel þ81-42-628-4584; Fax þ81-42-628-4542; e-mail [email protected].

(1) Muzzarelli, R. A. A.; Muzzarelli, C. Adv. Polym. Sci. 2005, 186, 151. (2) Rinaudo, M. Prog. Polym. Sci. 2006, 31, 603. (3) Gong, X.; Peng, S.; Wen, W.; Sheng, P.; Li, W. Adv. Funct. Mater. 2009, 19, 292. (4) Hsieh, W.-C.; Chang, C.-P.; Gao, Y.-L. Colloids Surf., B 2006, 53, 209. (5) Ubaidulla, U.; Sultana, Y.; Ahmed, F. J.; Khar, R. K.; Panda, A. K. Drug Delivery 2007, 14, 19. (6) van der Lubben, I. M.; Verhoef, J. C.; van Aelst, A. C.; Borchard, G.; Junginger, H. E. Biomaterials 2001, 22, 687. (7) Berthod, A.; Cremer, K.; Kreuter, J. J. Controlled Release 1996, 39, 17. (8) Agnihotri, S. A.; Mallikarjuna, N. N.; Aminabhavi, T. M. J. Controlled Release 2004, 100, 5. (9) Meier, W. Chem. Soc. Rev. 2000, 29, 295. (10) Minami, H.; Kanamori, H.; Hata, Y.; Okubo, M. Langmuir 2008, 24, 9254. (11) Kamio, E.; Yonemura, S.; Ono, T.; Yoshizawa, H. Langmuir 2008, 24, 13287. (12) Chen, L.-B.; Zhang, F.; Wang, C.-C. Small 2009, 5, 621. (13) Makuta, T.; Takada, S.; Daiguji, H.; Takemura, F. Mater. Lett. 2009, 63, 703.

14854 DOI: 10.1021/la101967u

separation method, and the drying-in-liquid method.9-16 These methods can be further integrated with the membrane emulsification method to achieve monodispersity in the obtained particles. The membrane emulsification method is a technique used to prepare emulsions utilizing membrane pores, where the pore size and pore size distribution are crucial factors affecting the average diameter of the emulsions and their monodispersity. The membrane emulsification technique using Shirasu porous glass (SPG) membranes is one of the best-known methods for obtaining monodisperse emulsions because of the sharpness of the pore size distribution.17-20 In fact, there are several reports of preparation of monodisperse particles using the SPG membrane emulsification method.21-24 We recently developed monodisperse chitosan microspheres using the SPG membrane emulsification method.25 The average (14) Yin, W.; Yates, M. Z. Langmuir 2008, 24, 701. (15) Chu, L.-Y.; Yamaguchi, T.; Nakao, S.-I. Adv. Mater. 2002, 14, 386. (16) Fujiwara, M.; Shiokawa, K.; Hayashi, K.; Morigaki, K.; Nakahara, Y. J. Biomed. Mater. Res., Part A 2006, 81A, 103. (17) Nakashima, T.; Shimizu, M.; Kukizaki, M. Key Eng. Mater. 1991, 61/62, 513. (18) Nakashima, T.; Shimizu, M.; Kukizaki, M. Adv. Drug Delivery Rev. 2000, 45, 47. (19) Vladisavlijevic, G. T.; Schubert, H. J. Membr. Sci. 2003, 225, 15. (20) Vladisavlijevic, G. T.; Shimizu, M.; Nakashima, T. J. Membr. Sci. 2004, 244, 97. (21) Hosoya, K.; Ohta, H.; Yoshizako, K.; Kimata, K.; Ikegami, T.; Tanaka, N. J. Chromatogr. A 1999, 853, 11. (22) Omi, S.; Senba, T.; Nagai, M.; Ma, G.-H. J. Appl. Polym. Sci. 2001, 79, 2200. (23) Chu, L.-Y.; Park, S.-H.; Yamaguchi, T.; Nakao, S.-I. Langmuir 2002, 18, 1856. (24) Kakazu, E.; Murakami, T.; Akamatsu, K.; Sugawara, T.; Kikuchi, R.; Nakao, S.-I. J. Membr. Sci. 2010, 354, 1. (25) Akamatsu, K.; Kaneko, D.; Sugawara, T.; Kikuchi, R.; Nakao, S.-I. Ind. Eng. Chem. Res. 2010, 49, 3236.

Published on Web 08/18/2010

Langmuir 2010, 26(18), 14854–14860

Akamatsu et al.

Article

Figure 1. Conceptual illustration of two preparation procedures for chitosan microcapsules with hollow structures using the SPG membrane emulsification technique. The first procedure utilizes tripolyphosphate to solubilize the core particles and, at the same time, physically crosslink the chitosan shell. The second procedure utilizes UV irradiation and photo-cross-linkable chitosan to chemically cross-link the chitosan shell and tripolyphosphate to solubilize the core particles.

diameters of the chitosan microspheres were successfully controlled within the submicrometer to 10 μm range, depending on the pore diameter of the SPG membrane, and the size distributions were narrow, with coefficient of variation of 9%. Both the controllability of average diameters and the monodispersity of the microspheres were due to the sharpness of the pore size distribution of the SPG membranes. However, the chitosan microspheres obtained in the previous study25 did not have hollow structures. Therefore, this study aims to prepare monodisperse chitosan microcapsules with hollow structures using the SPG membrane emulsification technique. As shown in Figure 1, we propose two novel preparation procedures, each consisting of three steps, to make hollow structures: (1) preparation of calcium alginate particles as core materials by mixing W/O emulsions containing alginate and calcium ions, (2) chitosan coating of the calcium alginate particles by mixing the calcium alginate particles and W/O emulsions containing chitosan, and (3) cross-linking of chitosan and solubilizing the core using a Ca2þ catcher. At step 3, the solubilized alginate is expected to permeate through the chitosan shell, resulting in formation of hollow structure. The first procedure is a “physical cross-linking method”, and unmodified chitosan is used at step 2, while at step 3, cross-linking and solubilization are performed at the same time by mixing the chitosan-coated particles and W/O emulsions containing tripolyphosphate. Tripolyphosphate can be a chitosan cross-linker because of electrostatic interactions and can be a capture agent of calcium ions. The second procedure is a “chemical cross-linking method”, using photo-cross-linkable chitosan at step 2, while at step 3, cross-linking is performed under UV irradiation and solubilization is achieved by tripolyphosphate treatment. Langmuir 2010, 26(18), 14854–14860

The photo-cross-linkable chitosan26-29 is a modified chitosan in which azido groups substitute for some of the amino groups. In this study, 4-azidobenzoic acid is used to introduce azido groups to chitosan. Using the second procedure involving photocross-linkable chitosan, the shell layer can be cross-linked with chemical bonds. Therefore, the shell layer of chitosan is reinforced, compared with the case utilizing physical cross-linking by tripolyphosphate. Comparisons are made to investigate the effect of the strength of chitosan cross-linking on the formation of hollow structures because the hollow structures can only be made after permeation of alginate through the chitosan shell layer. The difference of the cross-linking method will probably have an effect on the formation of hollow structure. In all steps, the SPG membrane emulsification technique was employed to prepare emulsions to form monodisperse particles. Thus, this method can provide biocompatible and biodegradable chitosan microcapsules that have monodispersity, controlled diameters, and hollow structures. Here, we report detailed results for each step and successful preparation of monodisperse, hollow chitosan microcapsules. In particular, controlling the average diameter of the microcapsules using these procedures and the effect of chitosan shell cross-linking on the formation of hollow structures are discussed in this paper. (26) Ono, K.; Saito, Y.; Yura, H.; Isikawa, K.; Kurita, A.; Akaike, T.; Ishihara, M. J. Biomed. Mater. Res. 2000, 49, 289. (27) Yeo, Y.; Burdick, J. A.; Highley, C. B.; Marini, R.; Langer, R.; Kohane, D. S. J. Biomed. Mater. Res. 2005, 78A, 668. (28) Zhu, A.; Zhang, M.; Wu, J.; Shen, J. Biomaterials 2002, 23, 4657. (29) Fukuda, J.; Khademhosseini, A.; Yeo, Y.; Yang, X.; Yeh, J.; Eng, G.; Blumling, J.; Wang, C.-F.; Kohane, D. S.; Langer, R. Biomaterials 2006, 27, 5259.

DOI: 10.1021/la101967u

14855

Article

Akamatsu et al.

Experimental Section Materials. Alginic acid sodium salt (viscosity of the 1% of aqueous solution at 20 C is 80-120 mPa s), calcium chloride, chitosan (Mw: 48 kDa; deacetylation rate: min 80.0 mol/mol %), hexane, acetone, Span 85, tripolyphosphate, hydrochloric acid, N,N,N0 ,N0 -tetramethylethylenediamine, 1-ethyl-3-(3-(dimethylamino)propyl)carbodiimide, 4-azidobenzoic acid, and dimethyl sulfoxide were purchased from Wako Pure Chemical Industries Ltd., Japan. All chemicals were used without further purification. Preparation of Calcium Alginate Particles as Core Materials. The apparatus used to perform the SPG membrane emulsification preparation of W/O emulsions was the same as that described elsewhere.25 Dispersed phase (2 wt % aqueous alginic acid solution) was prepared and poured into the dispersion tank. Continuous phase (hexane containing 3 wt % of Span 85) was prepared and poured into a beaker. While stirring, the dispersed phase was injected slowly into the continuous phase using pressurized nitrogen, and W/O emulsions containing alginic acid were prepared using the SPG membrane emulsification method. The SPG membrane pore sizes used were 4.9 and 2.6 μm. In the same way, W/O emulsions consisting of 4 wt % calcium chloride solution as the dispersed phase and hexane containing 3 wt % of Span 85 as continuous phase were also prepared. The pore size of the SPG membrane used was 12.0 μm. Subsequently, the W/O emulsions were mixed together, with stirring, for 12 h. After centrifugation and washing with hexane and acetone, the obtained particles were measured by DLS (Zetasizer Nano ZS90, Malvern Instruments Ltd., UK), and observed by FE-SEM (S-900, Hitachi, Japan). Chitosan Coating on the Calcium Alginate Particles. Using 1 wt % chitosan dissolved in 0.06 M hydrochloric acid as the dispersed phase, W/O emulsions containing chitosan were prepared using the SPG membrane emulsification method with 12.0 μm pore size. The obtained calcium alginate particles were mixed with the W/O emulsions containing chitosan, with stirring, for 2 days. After centrifugation and washing with hexane and acetone, the obtained particles were observed using FE-SEM and XPS (ESCALAB 250, Thermo Fisher Scientific K.K., Japan).

Figure 2. FE-SEM micrograph of calcium alginate microspheres prepared using the SPG membrane emulsification technique.

Cross-Linking of Chitosan and Solubilizing the Core by Tripolyphosphate Treatment. Using 3 wt % of tripolyphosphate aqueous solution as the dispersed phase, W/O emulsions were prepared using the SPG membrane emulsification method with 12.0 μm pore size. The obtained calcium alginate/chitosan particles were mixed with the W/O emulsion containing tripolyphosphate, with stirring, for 4 h. The obtained particles were observed by FE-SEM, TEM (JEM-2010F, JEOL Ltd., Japan), and XPS.

Preparation of Hollow Microcapsules with Photo-CrossLinkable Chitosan. The photo-cross-linkable chitosan (Az-

chitosan, chitosan-NHCO-C6H4-N3)26-29 was synthesized as follows. Chitosan (Mw: 48 kDa; deacetylation rate: min 80.0 mol/mol %) (200 mg) was dissolved in 15 mL of 0.06 M hydrochloric acid. N,N,N0 ,N0 -Tetramethylethylenediamine (120 mg) was dissolved in 1 mL of water and added to the chitosan solution. The pH of the solution was adjusted to 6 using aqueous HCl. 1-Ethyl3-(3-(dimethylamino)propyl)carbodiimide (70 mg) and 4-azidobenzoic acid (40 mg) were dissolved in 1 mL of water and 1 mL of DMSO, respectively, and mixed prior to addition to the chitosan solution. The reaction was conducted overnight at pH 5, and then the obtained Az-chitosan was filtered and freeze-dried. Using this Az-chitosan, hollow microcapsules were prepared according to the procedures described above. However, after the second step, UV irradiation (19 mW/cm2, irradiation range 290-400 nm) was conducted for 15 min to complete the cross-linking of chitosan at the shell parts of the microspheres. At the final step, tripolyphosphate treatment was conducted for 12 h instead of 4 h.

Results and Discussion Preparation of Calcium Alginate Particles as Core Materials. Figures 2 and 3 show FE-SEM micrographs of calcium 14856 DOI: 10.1021/la101967u

Figure 3. DLS measurements of calcium alginate microspheres prepared using SPG membranes with 4.9 or 2.6 μm pore sizes.

alginate microspheres prepared using the SPG membrane emulsification technique and their DLS measurement results. The FESEM observations and DLS measurements revealed that monodisperse, spherical calcium alginate microspheres with average diameters of 4.4 μm were obtained when a 4.9 μm pore size SPG membrane was used. It is well-known that an alginate molecule is a heteropolymer of D-mannuronic and L-guluronic acids and that gelling occurs when divalent cations such as calcium ions are captured by the L-guluronic acid-rich regions of the alginate molecule. Therefore, in this case, when the W/O emulsion containing alginic acid and the emulsion containing calcium ions were mixed, the calcium ions were captured by alginic acid, resulting in successful formation of calcium alginate microspheres in each emulsion. This mechanism of microsphere formation is reasonable because we have already succeeded in preparing monodisperse chitosan microspheres by mixing of W/O emulsions.25 In addition, it should be noted that the obtained calcium alginate microspheres were monodispersed in diameter. This monodispersity of the microspheres resulted from the monodispersity of the emulsions prepared by the SPG membrane emulsification technique, as shown elsewhere.25 Langmuir 2010, 26(18), 14854–14860

Akamatsu et al.

Article

Figure 4. FE-SEM micrograph of calcium alginate microspheres with chitosan coated onto their surfaces using the SPG membrane emulsification technique.

Figure 5. FE-SEM micrograph of chitosan microparticles prepared using the SPG membrane emulsification technique after tripolyphosphate treatment.

When a 2.6 μm pore size SPG membrane was used to prepare W/O emulsions containing alginic acid, monodisperse, spherical calcium alginate microspheres with average diameter of 2.1 μm were obtained, as shown in Figure 3. We can conclude that it is possible to control the average diameter of the calcium alginate microspheres by manipulating the pore size of the SPG membrane. In addition, the microspheres also acquired monodispersity during this preparation. This is because the SPG membrane emulsification technique was utilized in the preparation of monodisperse W/O emulsions, within which the microsphere formation reactions occurred. In this section, we successfully prepared monodisperse calcium alginate microspheres using the SPG membrane emulsification technique to provide a uniform reaction environment and controlled the average microsphere diameters by manipulating the pore sizes of the SPG membranes. This means that it would be possible to design the core space of the hollow microcapsules. Chitosan Coating of the Calcium Alginate Particles. Figure 4 shows FE-SEM micrographs of 4.4 μm average diameter calcium alginate microspheres with chitosan coated onto their surfaces. In this step, chitosan and alginate were assumed to interact with each other because they are a polycation and polyanion, respectively. As a result of this electrostatic interaction, chitosan coating would be realized. XPS measurement showed that the surface of the microspheres contained elemental nitrogen. Chitosan contains nitrogen but alginate does not; thus, the nitrogen was attributed to the presence of chitosan. Therefore, we can say that chitosan coating was successfully conducted in this step. However, compared with the calcium alginate microspheres obtained in the first step, these are similar in appearance and the average diameter of the alginate/chitosan core/shell microspheres was almost the same. Therefore, it was indicated that the chitosan coating to the alginate particles was certainly realized due to the electrostatic interaction and that the amount of chitosan adhering to the calcium alginate microsphere due to this electrostatic interaction was not be large to increase the diameter of the core/shell microspheres.

In this section, we successfully prepared monodisperse alginate/ chitosan core/shell microspheres by mixing of W/O emulsions containing chitosan and calcium alginate microspheres. This chitosan coating was due to the electrostatic interaction between chitosan and alginate. In addition, the average diameter of the core/shell microspheres proved to be almost the same as that of the calcium alginate microspheres. Cross-Linking of Chitosan and Solubilizing the Core by Tripolyphosphate. Figure 5 shows FE-SEM micrographs of chitosan microparticles after tripolyphosphate treatment. Compared with Figure 4, the particles have similar appearances before and after tripolyphosphate treatment. The average diameter of the chitosan microparticles after the tripolyphosphate treatment was almost the same as before the treatment. Figure 6 shows a TEM micrograph of chitosan microparticles after treatment. High contrast can be observed in the center of the microparticle, indicating that the calcium alginate particle in the core part was removed and the observed microparticle has a hollow structure. Other microparticles can also be seen to have hollow structures by TEM observation. In addition, XPS measurement was conducted to confirm that these hollow microcapsules consisted of chitosan and tripolyphosphate and determine that the surface of the microcapsules contained the elements N and P. Among alginate, chitosan, and tripolyphosphate, only chitosan contains nitrogen and only tripolyphosphate contains phosphorus. Thus, the chitosan microcapsules with hollow structures were found to include chitosan and tripolyphosphate. As mentioned above, tripolyphosphate forms a polyanion in aqueous media, so it interacts with polycations and works as a chitosan cross-linker because of electrostatic interactions. At the same time, it captures calcium ions. This strong force acts on the calcium alginate gel particles, resulting in hollow-structured particles. Although the chitosan shell was shown to be thin, described in the previous section, the hollow space in the microcapsule was rather small. This is probably because parts of the calcium alginate microsphere were not completely removed by the tripolyphosphate treatment. It is very likely that calcium ions permeated through the chitosan shell after solubilization

Langmuir 2010, 26(18), 14854–14860

DOI: 10.1021/la101967u

14857

Article

Akamatsu et al.

Figure 7. UV-vis spectra of synthesized Az-chitosan and unmodified chitosan.

Figure 6. TEM micrograph of chitosan microparticles prepared using the SPG membrane emulsification technique after the tripolyphosphate treatment.

by tripolyphosphate due to their small sizes; however, the diffusion resistance of the alginate molecules through the chitosan shell was high, and some of the alginate molecules could not permeate through the chitosan shell, even after solubilization by tripolyphosphate. Therefore, the permeability of the alginate as a core material would be a crucial factor controlling the hollow structure. The alginate that could not permeate through the shell would exist inside the cross-linked chitosan shell due to the electrostatic interaction. However, it is difficult to determine how much alginate was left in the microcapsules after this treatment, in this stage. In this section, it was determined that monodisperse chitosan microcapsules were successfully prepared by the first procedure shown in Figure 1 and that the average diameter of the hollow chitosan microcapsule can be determined from the average diameter of the alginate microspheres. Effect of Cross-Linking on the Formation of Hollow Structure. Successful synthesis of photo-cross-linkable chitosan (Az-chitosan) was confirmed using FT-IR, 1H NMR, and UV-vis measurements. In the FT-IR spectrum of synthesized Az-chitosan, a new absorption peak at 2131 cm-1 that did not appear in the case of unmodified chitosan was observed and attributed to the -N3 group. The UV-vis spectra of unmodified chitosan and synthesized Az-chitosan are shown in Figure 7. In the case of synthesized Azchitosan, UV-vis absorbance at 270 nm can be observed. On the other hand, in the case of unmodified chitosan, this absorbance does not appear. In addition, the azido content of the synthesized Azchitosan was calculated after calibrating the UV-vis absorbance using 4-azidobenzoic acid, and the results indicated that azido groups were introduced to 3% of the amino groups in chitosan molecules. This number of cross-linkable points is considered sufficient to form a rigid chitosan shell layer. In fact, we confirmed that the aqueous solution containing synthesized Az-chitosan turned into a gel after UV irradiation and, on the other hand, that unmodified chitosan did not change state after UV irradiation. Figure 8 shows a FE-SEM micrograph of calcium alginate microspheres with Az-chitosan coated onto their surfaces and 14858 DOI: 10.1021/la101967u

Figure 8. FE-SEM micrograph of calcium alginate microspheres prepared using the SPG membrane emulsification technique with Az-chitosan coated onto their surfaces and then cross-linked under UV irradiation.

then cross-linked by UV irradiation. XPS measurement confirmed that the Az-chitosan coating was successfully formed. Even after the Az-chitosan coating and UV irradiation, the diameters of the microspheres did not change and the core/shell microspheres maintained their spherical shapes, as was the case using unmodified chitosan. This means that Az-chitosan was successfully coated onto the surfaces of the alginate microparticles because of the electrostatic interaction in this case. However, as in the case of unmodified chitosan, the amount of Az-chitosan adhering to the calcium alginate microsphere would not be great. Figure 9 shows FE-SEM micrographs of Az-chitosan microparticles after tripolyphosphate treatment to solubilize and remove the core. The connection of the microcapsules are observed. This interparticle bridging probably resulted from the crosslinking reactions among the particles. Each core(alginate)/shell(photo-cross-linkable chitosan) microparticle before the crosslinking has azido group on their surface, so the undesirable interparticle cross-linking reaction may occur as well as the desirable intraparticle cross-linking reactions occur. To obtain microcapsules with discrete state, some treatment seems to be needed. Langmuir 2010, 26(18), 14854–14860

Akamatsu et al.

Figure 9. FE-SEM micrograph of Az-chitosan microparticles prepared using the SPG membrane emulsification technique after tripolyphosphate treatment to solubilize and remove the core.

However, in the photo-cross-linkable chitosan, azido groups were introduced to no less than 3% of the amino groups in chitosan molecules. Such a large amount of cross-linkable points are sufficient to form rigid cross-linking in the shell layer, even though some of the azido groups were used for undesirable interparticle cross-linking. Compared with Figure 8, the particles were similar before and after tripolyphosphate treatment, as was the case when microcapsules were prepared by the first procedure using unmodified chitosan. The average diameter of the Az-chitosan microparticles hardly changed after tripolyphosphate treatment, again showing the same trend. Figure 10 shows a TEM micrograph of the Az-chitosan microparticles after treatment. Even though high contrast was observed in the chitosan microparticle cross-linked with tripolyphosphate, as shown in Figure 6, in the case of Az-chitosan microparticles, contrast was not clearly observed in all microparticles. This indicates that the structures obtained were not completely hollow, compared with the case of using unmodified chitosan. However, some of the microspheres were observed to have hollow structures because contrast can be seen in Figure 10. In addition, SEM observation revealed that the microparticles deformed after vacuum drying treatment. Therefore, we can conclude that hollow structures were built using both preparation procedures but that the first procedure using unmodified chitosan is preferable for building hollow structures. This difference is attributed to the strength of cross-linking. In the case of the photo-cross-linking method, the chitosan shell was cross-linked strongly, increasing the permeation resistance of the chitosan shell and inhibiting solubilized alginate passing through. This is why the hollow structure was not formed completely, even though the solubilization time was increased from 4 to 12 h. The alginate that could not permeate out would also exist inside the cross-linked chitosan shell owing to the electrostatic interaction. It is clear that cross-linking of chitosan is one of the crucial factors in forming hollow structures in chitosan microcapsules using methods involving the SPG membrane emulsification technique. Langmuir 2010, 26(18), 14854–14860

Article

Figure 10. TEM micrograph of Az-chitosan microparticles prepared using the SPG membrane emulsification technique after tripolyphosphate treatment.

In this section, it was shown that the second procedure shown in Figure 1 can also create hollow chitosan microcapsules and that the average diameter of the microcapsules was determined by that of the calcium alginate microspheres prepared at step 1 as precursors. In addition, it is clear that the effect of cross-linking strength on the chitosan is a crucial factor in forming hollow structures.

Conclusion In this study, we proposed and demonstrated two novel preparation methods for hollow chitosan microcapsules using the SPG membrane emulsification technique and successfully prepared monodisperse chitosan microcapsules. In the first method, we first prepared calcium alginate microspheres as core particles and then coated them with unmodified chitosan. Finally, using tripolyphosphate, cross-linking of the chitosan shell layer and solubilization and removal of the alginate core were achieved at the same time. In the second method, we also first prepared calcium alginate microspheres as core particles and then coated them with Az-chitosan. Finally, UV irradiation was employed to crosslink the chitosan shell layer, and tripolyphosphate treatment was used to solubilize and remove the core particles. At each step, the SPG membrane emulsification technique was employed to provide a uniform reaction environment and obtain monodisperse microparticles. For both methods, it was determined that the average diameters of the chitosan microcapsules depended on those of the calcium alginate microparticles and that they had hollow structures. However, in both methods, a part of the alginate would be left inside the microcapsules according to the TEM observation. Therefore, the shell part was probably made of cross-linked chitosan and some alginate. The first procedure using tripolyphosphate treatment to obtain physical cross-linking was preferable to obtain a hollow structure, compared with the second procedure using UV irradiation to obtain chemical cross-linking in photocross-linkable chitosan. A hollow structure can only be constructed after solubilizing the calcium alginate microparticles and having them penetrate through the chitosan shell. In this study, the permeation resistance of alginate through the cross-linked chitosan DOI: 10.1021/la101967u

14859

Article

shell was assumed to be larger in the second procedure because chemical cross-linking is more rigid. And this assumption coincided with our obtained results. Therefore, the difference in final hollow structure between the two preparation procedures was attributed to the difference in the permeation resistances inhibiting alginate passing through the chitosan shell layers.

14860 DOI: 10.1021/la101967u

Akamatsu et al.

Acknowledgment. TEM observation was conducted in the Center for Nano Lithography & Analysis, The University of Tokyo, supported by the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan. A part of this research was supported by Grant-in-Aid for Young Scientists (B) (No. 21760605) from MEXT.

Langmuir 2010, 26(18), 14854–14860