Colloid Polym Sci (2003) 281: 157–163 DOI 10.1007/s00396-002-0763-3
Jin-Woong Kim Ji-Young Ko Jung-Bae Jun Ih-Seop Chang Hak-Hee Kang Kyung-Do Suh
Received: 10 April 2002 Accepted: 3 June 2002 Published online: 6 September 2002 Springer-Verlag 2002
J.-W. Kim Æ I.-S. Chang Æ H.-H. Kang Pacific Corporation R&D Center, 314-1, Bora-ri, Kiheung-eup, Yongin-si, Gyeonggi-do, 449–729, Korea J.-Y. Ko Æ J.-B. Jun Æ K.-D. Suh (&) Division of Chemical Engineering, College of Engineering, Hanyang University, Seoul, 133–791, Korea E-mail:
[email protected] Tel.: +82-2-22900526 Fax: +82-2-22952102
ORIGINAL CONTRIBUTION
Multihollow polymer microcapsules by water-in-oil-in-water emulsion polymerization: morphological study and entrapment characteristics
Abstract Multihollow cross-linked poly(methyl methacrylate) (PMMA) microcapsules were produced by water-in-oil-in-water emulsion polymerization and their applicability to encapsulate water-soluble ingredients was evaluated. In the microscopic analysis, all the PMMA microcapsules showed multihollow structures. In order to evaluate the entrapment efficacy, continuously, a water-soluble ingredient (monosodium phosphate, MSP) was incorporated into the inner voids of the microcapsules, and then its releasing profiles were measured with the storage conditions in pure water. In the releasing test, it was observed
Introduction Water-in-oil-in-water (W/O/W) multiple emulsions are three-phase systems in which small internal aqueous droplets, surrounded by a primary surfactant-stabilizing layer, are dispersed in the oil phase, which, in turn, is dispersed in the external aqueous phase and is also surrounded by a secondary surfactant layer [1, 2, 3]. In spite of their attractive emulsion structure, the application of W/O/W multiple emulsions has been limited by the lack of stability. Therefore, there have been many attempts to enhance their stability by employing several approaches, including the incorporation of systematic surfactant compositions, especially, polymeric surfactants [4, 5, 6], the interfacial complexation of macromolecules [7], the treatment of phase properties [8], the control of interdiffusion behavior [9, 10], the addition of water-soluble ingredients [11, 12, 13], and so on.
that the degree of cross-linking of PMMA had a significant influence on the migration of MSP through the polymer phase. At a sufficient degree of cross-linking of the polymer phase, the leakage of MSP out of the microcapsules was stopped successfully. It is believed that the extremely small network size and the mechanically strong network structure hindered effectively the water flow caused by the concentration gradient between water-soluble ingredients. Keywords Multihollow Æ Waterin-oil-in-water Æ Water-soluble ingredient Æ Releasing profiles Æ Network structure
Most studies of W/O/W emulsions have dealt with the stabilization of active ingredients and their sustained-release properties [14, 15, 16, 17]. Even though W/O/W emulsions often appear to be superior in the aspect of a prolonged-release system, a serious problem can be found in that there have been few works that satisfy the requirements for wide application. The reasons for this are their intrinsic thermodynamic instability and low active ingredient entrapment efficacy [3]. To overcome all those failures of W/O/W emulsions, we designed a useful product via W/O/W multiple emulsion polymerization [18, 19]. After preparing a stable W/O/W emulsion composed of water1, monomers, and water2, a radical polymerization was carried out selectively in the monomer oil phase. The morphology after polymerization showed multihollow structures where many voids are located in the rigid polymer particles, resulting in multihollow polymer microcapsules.
158
In this contribution, we evaluate the usefulness of the multihollow polymer microcapsules for the entrapment of water-soluble ingredients and their releasing properties in water. The releasing properties of an electrolyte, monosodium phosphate (MSP), were controlled considering the network density of the polymer wall in the microcapsules. Finally, the importance of the network density of the polymer wall was emphasized for the preparation of the hollow-structured polymer microcapsules showing effective entrapment of water-soluble ingredients and their prolonged release.
Table 1 A standard water-in-oil-in-water (W/O/W) emulsion recipe for multihollow polymer microcapsules. Ingredients: methyl methacrylate (MMA), ethylene glycol dimethacrylate (EGDMA), Arlacel P135, 2,2¢-azobis(2,4-dimethylvaleronitrile) (ADVN), monosodium phosphate (MSP), distilled, deionized water (DDI), poly(vinyl alcohol) (PVA) Emulsifying step
Ingredient
Weight (g)
W/O emulsiona
MMA EGDMAc Arlacel P135 ADVN MSP DDI water W/O emulsion PVA NaNO2 DDI water
9.35 9.35 (variable) 0.3 0.187 1 (variable) 9.813 30 3 0.01 266.99
W/O/W emulsionb
Experimental Materials a
The oil phase chosen was methyl methacrylate monomer (MMA, Junsei Chemical Co.) and ethylene glycol dimethacrylate (EGDMA, Aldrich Chemical Co.). The lipophilic primary surfactant was Arlacel P135, a poly(ethylene glycol)(30) dipolyhydroxystearate (Uniquema Americas), that was introduced into the internal oil phase. The hydrophilic secondary surfactant, poly(vinyl alcohol) (PVA, Mw=8.8·104–9.2·104 gmol–1, 87–89% degree of saponification) was purchased from Kuraray Co. 2,2¢-Azobis(2,4-dimethylvaleronitrile) (ADVN, Wako Pure Chemicals) was recrystallized from methanol before use. As a model water-soluble ingredient, MSP (Aldrich) was selected. Throughout the process, distilled, deionized water was used. W/O/W emulsion polymerization W/O/W emulsion polymerization was carried out by modifying the two-step emulsification procedure that was developed by Matsumoto [1]. First, a W/O emulsion was prepared as follows. Aqueous solution (10.8 g) containing MSP was dropped into an oil phase (19.2 g) composed of MMA, EGDMA, Arlacel P135, and ADVN. Then, the water/oil mixture was homogenized with an MX-5 homogenizer (Nihonseiki Co., Japan) at 1.0·104 rpm for 5 min at room temperature. The W/O emulsion prepared was a milky viscous phase (about 1.0·104 cP). Second, the MSP-containing W/O emulsion was redispersed by homogenizing it mildly again at 5.0·102 rpm for 5 min in 1 wt% PVA aqueous solution, obtaining finally a stable W/O/W emulsion. Right after the preparation of the W/O/W emulsion, the emulsion was transferred into a double-walled glass reactor equipped with a mechanical stirrer, a reflux condenser, thermocouples, and a nitrogen gas inlet system. The polymerization in the aqueous phase was inhibited by adding a small amount of sodium nitrite (0.01 g) [20, 21]. After deaerating with nitrogen gas, the polymerization was carried out at 60 C for 10 h while agitating it at a speed of 2.5·102 rpm. The microcapsules produced were washed repeatedly by decantation in water and ethanol (50/50, w/w) and dried under vacuum at ambient temperature. A typical recipe is summarized in Table 1. Measurement of entrapment efficacy Conductivity analysis was employed to assess the release of MSP from microcapsules to the outer aqueous phase. After dispersing the MSP-containing microcapsules in 0.1 wt% Tween 60 [poly(oxyethylene)(20) sorbitan monostearate, Aldrich] aqueous solution, the conductivity was measured with a conductimeter (Wissenschaftlich Technische Werksta¨ten). From a log–log linear relationship of the conductivity with the concentration of MSP, we
1.0·104 rpm/5 min/25 C 5.0·102 rpm/5 min/25 C c The degree of cross-linking of the polymer phase was controlled by varying the concentration of EGDMA b
determined the release fraction, Q(t), of MSP, Q(t)=m(t)/m0, where, m0 is the calibrated mass and m(t) is the mass determined by conductivity measurements for a given time t [22, 23]. From the Q(t) obtained, an entrapment efficacy, EMSP, of MSP in the microcapsules was defined as follows: EMSP=[1–Q(t)]·100. Determination of conversion The conversions were determined with a gravimetric method. An aliquot (10 g) containing all the ingredients was extracted from the reactor at different time intervals. Promptly, after adding a drop of 1% hydroquinone solution, the extract was quenched in ice–water. Each sample was then dried under a vacuum at ambient temperature until the weight change was less than 0.001 g. Finally, the conversion was calculated as follows [24, 25]: Conversionð%Þ ¼
M2 M1 ðw2 þ w3 þ w4 Þ 100; M1 w1
where M1 is the weight of the sample before drying, M2 is the weight of the sample after drying, w1 is the weight fraction of the monomers, w2 is the weight fraction of the surfactant and stabilizer, w3 is the weight fraction of the initiator, and w4 is the weight fraction of the polymerization inhibitor in the aqueous phase. Characterizations The images of the surface and the inner part of the microcapsules were observed with a field-emission scanning electron microscope (JSM-6330F, JEOL). In order to verify the hollow structure, the microcapsules were placed onto a slide glass and diluted with a drop of toluene (n=1.4967), which has a similar refractive index to that of PMMA (n=1.4893). Then, the contrast of the image was observed with an optical microscope (OM, Nikon Microphot FXA). The average diameter of the microcapsules was measured with a laser scattering size analyzer (LS32, Beckman Coulter).
Results and discussion There has been intensive research to produce polymer microcapsules loading a variety of active materials by
159
employing many different techniques [18, 19, 26, 27, 28]. In this work, we propose hollow-structured polymer microparticles as a useful microcapsule capable of encapsulating water-soluble ingredients. We emphasize the importance of a dense network of the polymer barrier in the microcapsules in order to entrap effectively the water-soluble ingredients. Formation of stable W/O/W multiple emulsions For a successful polymerization, the W/O/W emulsions must retain their stability during polymerization. In our previous studies [18, 19], we produced hollow-structured polymer microparticles via W/O/W emulsion polymerization. In those studies, good emulsion stability could be achieved by manipulating the characteristics of the monomer phase, i.e., viscoelastic control. To enlarge the scope of this study, we constructed a surfactant system for W/O/W emulsion polymerization, composed of macromolecular surfactants: Arlacel P135 as a lipophilic primary surfactant and PVA as a hydrophilic secondary surfactant. An OM image of W/O/W multiple emulsions prepared with the surfactant system just described is shown in Fig. 1. All the emulsions contained many small internal water droplets. Right after the preparation of the W/O/W multiple emulsions, the entrapment efficacy of MSP and the emulsion droplet sizes were measured and the results are summarized in Table 2. One finds that the surfactant system Arlacel P135/PVA showed good ability to entrap MSP in the inner water phase. Especially, the concentration of Arlacel P135 did not have much influence on the loading efficacy of MSP. JagerLezer et al. [22] reported a similar result. It appears that with only a small amount, the primary surfactant was
Fig. 1 Optical microscope (OM) image of a water-in-oil-in-water (W/O/W) multiple emulsion right after preparation. In this sample, the emulsion system had the standard composition shown in Table 1
Table 2 Characteristics of a W/O/W multiple emulsion. Conditions: 3.3 wt% MSP in the emulsion composition, 10 wt% EGDMA against total monomer weights [Arlacel P135] (wt%)a
[PVA] (wt%)b
Davg (lm)
Loading yield (%)c
0.05 0.1 0.1 0.1 0.25
1.0 0.5 1.0 3.0 1.0
136 254 131 95 140
92.6 94.8 93.2 95.0 94.1
a The concentration of Arlacel P135 was controlled against total composition b The concentration of PVA was controlled against total composition c The loading yield of MSP in the W/O/W multiple emulsions determined by conductivity analysis (loading yield is equal to EMSP)
arrayed sufficiently in the W/O interface. Judging from the high loading yield of MSP, it is reasonable to say that the W/O/W emulsions prepared in this study had a lower breakdown rate, meaning a high emulsion stability. In our study, the size of the final W/O/W emulsions was rather affected by the concentration of PVA, resulting from the hydrophobic association of unsaponificated parts in PVA molecules in the interface between monomer and continuous water. From these results, it was established that the Arlacel P135/PVA surfactant system was useful for the application of W/O/W multiple emulsions and their polymerization. Characteristics of multihollow PMMA microcapsules The W/O/W multiple emulsions containing MSP were polymerized carefully by changing the concentration of cross-linker, EGDMA, in the polymer phase at a high temperature of 60 C. Apparently, the morphology of the initial multiple emulsions shown in Fig. 1 was retained throughout the polymerization process. Scanning electron microscopy photographs of the multihollow PMMA microcapsules obtained after the polymerization are shown in Fig. 2. As shown, clean microcapsules were produced in the size range of 100 lm. From the observation of the fracture surface shown in Fig. 2b, it was verified that the microcapsules contained a large number of small voids in their inner phase. The presence of the voids must be attributed to the internal aqueous droplets before the polymerization of W/O/W multiple emulsions. In order to confirm the inner phase, the dried microcapsules were redispersed in toluene and their images were captured with an OM. OM images at different cross-linking conditions of the PMMA phase are shown in Fig. 3. Because the refractive index of PMMA is similar to that of toluene, only the inner voids filled with air display contrast different from the PMMA polymer
160
Fig. 2 Scanning electron microscope photographs of multihollow poly(methyl methacrylate) (PMMA) microcapsules cross-linked with 50 wt% ethylene glycol dimethacrylate (EGDMA): a surface image and b inner phase image. The calculated loading yield of monosodium phosphate (MSP) in the microcapsules was 5%
phase [18, 19]. In all the OM photographs, high contrast regions were observed in the form of small domains, meaning there are many air voids in the microcapsules. In our observation, the presence of MSP in the microcapsules did not have any influence on the polymerization process and the capsule morphology. When the microcapsules were not cross-linked or only slightly cross-linked, the dark regions started to disappear within a few minutes. This happens because the toluene penetrates through the PMMA walls from the continuous phase to inner voids and fills the voids. However, as the degree of cross-linking increased, the dark regions in the microcapsules maintained their contrast for a long time. As an intermediate state, Fig. 3a shows a good example. Partially toluene-filled regions could be clearly observed in the OM image. Eventually, the dark regions also disappeared. For high cross-linking, however, there was no change in the darkness of the inner voids, even after
Fig. 3 OM image of multihollow PMMA microcapsules after 10 min in toluene: a 10 wt% EGDMA and b 50 wt% EGDMA. The calculated loading yield of MSP in the microcapsules was 5%
several hours. This indicates that the toluene molecules were not capable of penetrating the polymer wall, owing to the dense network structure. The kinetics of the W/O/W emulsion polymerization was observed in the presence of MSP in the internal water phase and is shown in Fig. 4. The overall behavior of the polymerization conversion was similar to that of suspension polymerization, because the polymerization proceeded within the monomer droplets. However, in the case of cross-linking, a gel effect could be observed within 1 h, meaning a high polymerization rate by the formation of a polymer network. This high polymerization rate was expected to stop the easy elution of MSP out of the microcapsules during polymerization. Effective entrapment of water-soluble ingredients After polymerization, the loading yield of MSP in the microcapsules was measured by conductivity analysis.
161
Fig. 4 Conversion of multihollow PMMA microcapsules with polymerization time in the presence of MSP (5%): 0 wt% (open circles) and 30 wt% (closed circles) EGDMA in the PMMA phase
The MSP loading trends after the polymerization are shown in Fig. 5. In the case of no cross-linking, the loading of MSP in the microcapsules was impossible. Only in the case of cross-linking was MSP loaded suc-
Fig. 5 Calculated MSP loading yield in multihollow PMMA microcapsules, [MSP]c, with the measured concentration of MSP, [MSP]m, at different degrees of cross-linking: 0 wt% (squares), 10wt% (circles), 30 wt% (up triangles), and 50wt% (down triangles) EGDMA in the PMMA phase. The dotted line is the theoretical loading yield
cessfully in the microcapsules. Compared with the loading amount calculated from the weight fraction in the recipe, the loading amount measured by the conductivity measurement displayed a relatively low yield at the level of 20–50%. After the preparation of the W/O/ W multiple emulsions, the loading yield reached about 95%. The low loading yield means that a portion of MSP in the microcapsules was eluted to the continuous aqueous phase during polymerization. At such a high polymerization temperature of 60 C, it seems that the W/O/W emulsions could not maintain their emulsion stability, leading to partial emulsion breakdown. In Fig. 5, it is observed that the loading yield can be recovered by increasing the degree of cross-linking of the polymer phase. The reason can be found in the fact that as already mentioned, the polymer network that formed quickly in the initial stage of the polymerization prevented the elution of MSP from the microcapsules. When the microcapsules are redispersed in pure water, MSP located in the inner voids of the microcapsules has a tendency to be released. The concentration gradient between all the water-soluble ingredients causes a water flow from the continuous phase to the inner void phase [22]. The water flow makes the capsule structure unstable. On the basis of our experimental results, the cross-linking of the polymer wall is expected to provide a possibility to control the water flow. The release profiles of MSP measured with the storage time by conductivity analysis at room temperature are shown in Fig. 6. At a
Fig. 6 Entrapment efficacy of MSP in multihollow PMMA microcapsules with the storage time at different degrees of crosslinking: 10 wt% (squares), 30 wt% (circles), and 50 wt% (triangles) EGDMA in the PMMA phase. The storage temperature was 25 C. The calculated loading yield of MSP in the microcapsules was 5%
162
(Fig. 7). These results indicate that the formation of a dense polymer network did not allow easy penetration of water molecules. The generation of water flow could be suppressed by controlling the network structure of the microcapsules. In the usual W/O/W emulsion system containing electrolytes, the water flow induces the emulsion breakdown. That is because the wall of the W/ O/W emulsions is composed of soft oils or waxes. However, in our multihollow polymer microcapsules, one should note that the wall is a mechanically strong polymer network. Owing to the strong and fine network characteristics of the polymer wall, therefore, the release rate of MSP showed a negligible difference over the temperature range (Figs. 6, 7), contributing consequently to the improvement in the entrapment efficacy.
Conclusions
Fig. 7 Entrapment efficacy of MSP in multihollow PMMA microcapsules with the storage time at different degrees of crosslinking: 10 wt% (squares), 30 wt% (circles), and 50 wt% (triangles) EGDMA in the PMMA phase. The storage temperature was 45 C. The calculated loading yield of MSP in the microcapsules was 5%
low degree of cross-linking, the MSP in the microcapsules was released readily. In contrast, in the case of sufficient cross-linking, MSP could be entrapped effectively in the microcapsules for a long time. However, a typical burst effect could not be avoided in the release profiles of MSP. It appears that the refilling of the dried inner voids with water was responsible for the initial quick leakage of MSP out of the microcapsules. At equilibrium, one can find that although the amount of MSP released slightly increased, the release of MSP could be prevented favorably even at high temperature
In this study, we introduced a useful approach that can encapsulate water-soluble ingredients in multihollow polymer microcapsules. Multihollow PMMA microcapsules were produced via a W/O/W multiple emulsion polymerization method. Before the polymerization, the stable W/O/W multiple emulsions composed of internal water, monomer mixture, and external water were prepared by employing a surfactant system of Arlacel P135 and PVA. After the polymerization, the multihollow structure of the microcapsules was confirmed with optical microscopy and scanning electron microscopy. In the release test carried out in water, it was found that the cross-linking of the polymer phase provided the microcapsules with the ability to entrap the water-soluble ingredients effectively, which is responsible for the suppression of the water flow in the system. Acknowledgement This work was supported in part by the National Research Laboratory program (project no. 2000-N-NL-01C-270) by the Ministry of Science and Technology, South Korea.
References 1. Matsumoto S (1985) Formulation and stability of water-in-oil-in-water emulsions, ACS symposium series 274. American Chemical Society, Washington, DC, pp 415–436 2. Florence AT, Whitehill D (1982) Int J Pharm 11:277 3. Okochi H, Nakano M (2000) Adv Drug Delivery Rev 45:5 4. Rosano HL, Gandolfo FG, Hidrot JDP (1998) Colloids Surf A 138:109 5. Sela Y, Magdassi S, Garti N (1994) Colloid Polym Sci 272:684 6. Magdassi S, Frenkel M, Garti R, Kasan R (1984) J Colloid Interface Sci 374:97
7. Law TK, Whateley TL, Florence AT (1986) J Controlled Release 3:279 8. Kawashima Y, Hino T, Takeuchu H, Niwa T (1992) Chem Pharm Bull 40:1240 9. Davis SS, Round HP, Purewal TS (1981) J Colloid Interface Sci 80:508 10. Aronson MP, Petko MF (1993) J Colloid Interface Sci 159:314 11. Florence AT, Whitehill D (1981) J Colloid Interface Sci 79:243 12. Matsumoto S, Koda M (1980) J Colloid Interface Sci 73:13 13. Yazan Y, Aralp U, Seiller M, Grossiord JL (1996) Cosmet Toilet 111:1153
14. Magdassi S, Garti N (1986) J Controlled Release 3:2273 15. Fukushima S, Nishida M, Nakano M (1987) Chem Pharm Bull 35:3375 16. Ohwaki T, Nakamura M, Ozawa H, Kawashima Y, Hino T, Takeuchi H (1993) Chem Pharm Bull 46:741 17. Mishra B, Pandit JK (1990) J Controlled Release 14:53 18. Kim JW, Joe YG, Suh KD (1999) Colloid Polym Sci 277:252 19. Kim BS, Kim JW, Suh KD (2000) J Appl Polym Sci 76:38 20. Kim JW, Suh KD (2000) Polymer 41:6181
163
21. Okubo M, Yamashita T, Suzuki K, Shimizu T (1997) Colloid Polym Sci 175:288 22. Jager-Lezer N, Terrisse I, Bruneau, F, Tokgoz S, Ferreira L, Clausse D, Seiller M, Grossiord JL (1997) J Controlled Release 45:1
23. Ozer O, Baloglu E, Ertan G, Muguet V, Yazan Y (2000) Int J Cosmet Sci 22:459 24. Ho CH. Chen SA, Amiridis MD, VanZee JW (1997) J Polym Sci Part A Polym Chem 35:2907 25. Kim JW, Lee CH, Jun JB, Suh KD (2001) Colloids Surf A 194:57
26. Arshady R (1999) Microspheres, microcapsules and liposomes. Citus, London 27. Alexander K, Vogel M, Blankenship RH (1984) US Patent 4,468,498 28. Okubo M, Minami H (1997) Colloid Polym Sci 275:992
