Multishell Encapsulation Using a Triple Coaxial Electrospray System

May 11, 2010 - electrospray method as a drug delivery system, an elec- trospray system using a triple coaxial nozzle was devel- oped to generate multi...
0 downloads 0 Views 4MB Size
Download PDF
Anal. Chem. 2010, 82, 4644–4647

Multishell Encapsulation Using a Triple Coaxial Electrospray System Woojin Kim and Sang Soo Kim* Aerosol and Particle Technology Laboratory, School of Mechanical, Aerospace & Systems Engineering, Korea Advanced Institute of Science and Technology, Guseong-dong, Yuseong-gu, Daejeon 305-701, Republic of Korea To overcome the limitations of the conventional encapsulation methods and improve the potential use of the electrospray method as a drug delivery system, an electrospray system using a triple coaxial nozzle was developed to generate multishell capsules. Two conducting fluids, ethylene glycol and 4-hydroxybutyl acrylate, and one nonconducting fluid, olive oil, were chosen to manufacture the multishell capsules. The capsules were solidified by a photopolymerization device. We investigated the size distributions and visualized the capsules changing fluid flow rates. Dispersive Raman spectra were also monitored to determine the chemical composition of the capsules. The multishell capsules were generated in the overlapped cone-jet mode regime of the conducting fluids, and the sizes and shell thicknesses were controlled by the flow rates and applied voltages. Microencapsulation techniques have been widely studied because they can produce synthetic polymers, control targeted drug delivery, and protect a core from harsh environments. These techniques have especially focused on food, clinical, and biological applications. Among the various encapsulation techniques, the emulsion method, which includes microsphere hardening, solvent evaporation, and phase separation, has been used primarily for the synthesis of polymer capsules or controlled release devices.1,2 However, the process required for this method is time-consuming and requires additives, such as surfactants. Furthermore, the generated particles have wide size distributions, and it can be difficult to produce uniformly sized particles. To resolve these limitations, electrospray techniques have recently been considered because they are simple, cost-effective, and have a fast production time. These techniques can also generate monodisperse particles, allowing them to be used to generate bioaerosols, to synthesize particles and threads using polyester polymers (e.g., poly(lactic acid) (PLA) or poly(lactic-co-glycolic acid) (PLGA)), and to encapsulate using a coaxial nozzle. Kim et al.3 have shown the viability of creating bacteria aerosols using an electrospray system, and Xie et al.4 and Hong et al.5 have electrosprayed PLGA particles with paclitaxel and rifampicin (RIF). Especially, since Loscertales’s * To whom correspondence should be addressed: (phone) Tel: +82-42-3502009; (fax) +82-42-350-3210; (e-mail) [email protected]. (1) Mu, L.; Feng, S. S. J. Controlled Release 2003, 86, 33–48. (2) Hinds, K. D.; Campbell, K. M.; Holland, K. M.; Lewis, D. H.; Piche, C. A.; Schmidt, P. G. J. Controlled Release 2005, 104, 447–460. (3) Kim, K.; Kim, W.; Yun, S. H.; Lee, J. H.; Kim, S.; Lee, B. U. J. Aerosol Sci. 2008, 39, 365–372.

4644

Analytical Chemistry, Vol. 82, No. 11, June 1, 2010

pioneering studies of electrospraying with a coaxial nozzle,6 the technique has been used in many research fields. Chen et al.7 experimentally studied the influence of viscosity and flow rates on various coaxial jet electrospraying modes, and Herrera et al.8 introduced the driving liquid concept and demonstrated that the current scaling law of compound-jet electrospraying was in good agreement with that of single-jet electrospraying.9 Mei and Chen10 proposed criteria for particle encapsulation using the charge relaxation length and inertial length. The electrospraying of protein-based drugs11 and electrospinning of living cells12 using a coaxial nozzle have demonstrated the applicability of this technique in drug delivery and controlled release systems. However, encapsulation by electrospraying still has limitations for operating fluids because of properties of the fluids such as electrical conductivity, viscosity, and surface tension, and the capsules which are able to contain various drugs in the shell are helpful for potential use as a drug delivery system because most diseases need complex prescriptions and the capsule containing various medicines in the shells can reduce the dosing frequency. The triple coaxial nozzle system was introduced by Lallave et al.13 These authors obtained the Alcell lignin hollow fibers using electrospinning method and the ethanol which was sprayed from the outermost nozzle of the triple coaxial nozzle was used to prevent solidification of the Taylor cone and compensate fast evaporation of the jet. However, multishell encapsulation using the triple coaxial electrospray system has not yet been studied. In this work, we used for the first time the electrospray system mounted with a triple coaxial nozzle to generate multishell capsules that were composed of three immiscible materials. Multishell encapsulation can be used to enhance the diversity and commercial applicability of the technique for targeted drug (4) Xie, J.; Lim, L. K.; Phua, Y.; Hua, J.; Wang, C. H. J. Colloid Interface Sci. 2006, 302, 103–112. (5) Hong, Y.; Li, Y.; Yin, Y.; Li, D.; Zou, G. J. Aerosol Sci. 2008, 39, 525–536. (6) Loscertales, I. G.; Barrero, A.; Guerrero, I.; Cortijo, R.; Marquez, M.; Gan ˜a`nCalvo, A. M. Science 2002, 295, 1695–1698. (7) Chen, X.; Jia, L.; Yin, X.; Cheng, J. Phys. Fluids 2005, 17, 032101. (8) Lo´pez-Herrera, J. M.; Barrero, A.; Lo´pez, A.; Loscertales, I. G.; Ma´rquez, M. J. Aerosol Sci. 2003, 34, 535–552. (9) Gan ˜a´n-Calvo, A. M.; Da´vila, J.; Barrero, A. J. Aerosol Sci. 1997, 28, 249– 275. (10) Mei, F.; Chen, D. R. Phys. Fluids 2007, 19, 103303. (11) Xie, J.; Ng, W. J.; Lee, L. Y.; Wang, C. H. J. Colloid Interface Sci. 2008, 317, 469–476. (12) Townsend-Nicholson, A.; Jayasinghe, S. N. Biomacromolecules 2006, 7, 3364–3369. (13) Lallave, M.; Bedia, J.; Ruiz-Rosas, R.; Rodrı´guez-Mirasol, J.; Cordero, T.; Otero, J. C.; Marquez, M.; Barrero, A.; Loscertales, I. G. Adv. Mater. 2007, 19, 4292–4296. 10.1021/ac100278c  2010 American Chemical Society Published on Web 05/11/2010

Table 1. Physical Properties of the Test Fluids

fluid Ethylene glycol Olive oil 4-HBA + PI (10 wt %)b a

Figure 1. Schematic diagram of the experimental setup.

delivery. The generated multishell capsules were examined throughout the hardening process. We also investigated the chemical components of the capsules and controlled the shell thickness by changing the flow rates of the fluids and the applied voltages to analyze the morphologies of the multishell capsules. EXPERIMENTAL SECTION To generate multishell capsules, we mounted the electrospray system using a triple coaxial nozzle, high-voltage (H.V.) power supply, and photopolymerization device (Figure 1). The triple coaxial nozzle consisted of three stainless steel capillaries; the inner and outer diameters (i.d. and o.d.) of the capillaries were 0.4 mm (i.d.)/0.785 mm (o.d.) for capillary A, 1.168 mm (i.d.)/ 1.573 mm (o.d.) for capillary B, and 2 mm (i.d.)/3.17 mm (o.d.) for capillary C. The innermost capillary was connected to a power supply (∼15 kV, ∼10 mA, Korea Switching, Inc.) that was able to supply a high voltage through the conducting fluid in the innermost capillary. The photopolymerization device consisted of four ultraviolet (UV) lamps (lamp length: 120 mm, λ ) 360 nm, SMT, Inc., Korea), a power supply, and an aluminum reflector that focused UV light on the generated capsules, reflecting the light in the opposite direction. Two conducting fluids, ethylene glycol (EG; Junsei Chemical Co., Ltd., Japan) and 4-hydroxybutyl acrylate (4-HBA; Tokyo Chemical Industry Co., Ltd., Japan), and one nonconducting fluid, olive oil (Junsei Chemical Co., Ltd., Japan), were used as the working fluids. The conducting fluids were selected considering their viscosities, because a conducting fluid with a large viscosity has a wide cone-jet mode regime.14 Furthermore, 2,2-dimethoxy-2-phenylacetophenone (Tokyo Chemical Industry Co., Ltd., Japan), as a photoinitiator (PI), was added to the 4-HBA which has photopolymerization characteristics when exposed to UV. The physical properties of the working fluids are listed in Table 1. The viscosities and surface tensions of the fluids were measured with a vibro-viscometer (SV-10, A&D, Japan) and a surface tensiometer (Sigma702, KSV Instruments, Finland), respectively. A conductivity meter (CM-21PW, DKK TOA, Japan) was also used to measure the electrical conductivities of the fluids. EG, olive oil, and 4-HBA were injected into capillaries A, B, and C, respectively, by syringe pumps (KDS-100, KD-Scientific), and then EG, electrified by the high-voltage power supply, formed a Taylor cone with the olive oil and 4-HBA of capillary B and C at the nozzle tip. An electrometer (Model 6514, Keithley Instruments) was used to measure the current, which was collected at (14) Hwang, Y. K.; Jeong, U.; Cho, E. C. Langmuir 2008, 24, 2446–2451.

surface viscosity tension conductivity (cP ) mPa · s)a (dyn/cm ) mN/m)a (µS/m)a 20.4 82.2 11.1

47.83 32.36 37.94

10.6 0.1 22.4

Values measured at 20 °C. b PI: Photo initiator.

the ground by the compound jet, to determine the cone-jet mode. We also visualized the electrospray mode at the capillary tip using a charge-coupled device (CCD) camera (Marlin F-145C2, AVT, Germany) with a zoom lens (70XL, Optem, Korea). We measured the size distributions of the multishell capsules that were generated from the cone-jet using a phase-Doppler particle analyzer (PDPA, TSI, Inc.). The PDPA which is composed of a fiber optic probe, signal processor, Receiver probe, and an Ar-ion laser (AirCooled 543, Melles Griot) can measure the particle size using the periodic variation of the scattering generated by interference fringe of the two intersected lasers. The measuring point was 40 mm below the nozzle tip and the size distributions were based on 1000 capsules. The capsules were also collected onto a glass slide that was 40 mm distant from the nozzle tip and covered with an oil film, and the 4-HBA layer, which formed the outer shell of collected capsules, was polymerized by UV lamps in the photopolymerization device. We generated the multishell capsules at various working fluid flow rates. The flow rate of 4-HBA was varied from 2 to 4 mL/h at fixed flow rates of EG and olive oil, and the flow rate of olive oil was increased from 0.2 to 2.6 mL/h at fixed flow rates of 4-HBA and EG. The capsules were observed by an optical microscope (×1000, Eclipse ME600L, Nikon, Japan) and analyzed the changes in capsule size and shell thickness based on 100 capsules using the image processing technique. A high-resolution dispersive Raman microscope (LabRAM HR UV/vis/NIR, Horiba Jobin Yvon) was used to determine the chemical components of the generated multishell capsules. The dispersive Raman microscope consisted of an Ar-ion laser, a sample chamber, a monochromator with a Rayleigh filter, and a detection device. When the laser irradiation generated Rayleigh scattering at selected multishell capsules, the Raman signals of the capsules were detected by the monochromator. RESULTS AND DISCUSSION Because two conducting fluids (EG and 4-HBA) were used for encapsulation, we examined the cone-jet mode regime of each fluid for stability. Because each fluid had different cone-jet mode ranges, the flow rates of the working fluids were determined experimentally so that they fell in the overlapping cone-jet mode regimes of EG and 4-HBA. The flow rates of EG and olive oil were fixed at 1 mL/h, while the flow rate of 4-HBA was increased from 2 to 3 to 4 mL/h. Figure 2 shows the stable cone-jet mode regimes of the compound jet by the three immiscible fluids, namely, the lower curve and the upper curve are the onset voltage and the collapse voltage of the cone-jet mode, respectively. The dripping and spindle mode were observed in the region below the onset voltage curve, and the spray became unstable and observed the Analytical Chemistry, Vol. 82, No. 11, June 1, 2010

4645

Figure 2. Stable region for the cone-jet mode of the electrospray changing the 4-HBA flow rate. The flow rates of the olive oil and EG are fixed at 1 mL/h, respectively. Figure 4. Raman spectra of the fluids and the generated multishell capsule.

Figure 3. Size distributions of the multishell capsules at various flow rates. For the Q4-HBA ) 2 mL/h, the mean diameter and geometric standard deviation (GSD) are 16.9 µm and 1.29. For the Q4-HBA ) 3 mL/h, the mean diameter and GSD are 19.0 µm and 1.26. For the Q4-HBA ) 4 mL/h, the mean diameter and GSD are 21.4 µm and 1.28, respectively.

multijet mode in the region above the collapse voltage curve. The stable cone-jet modes were investigated at 6.8-8.9, 7.1-9.0, and 7.3-9.2 kV increasing the flow rate of the 4-HBA, and the onset voltage of the cone-jet mode slightly increased with increasing the total flow rate. The size distributions of the capsules for the different flow rates are shown in Figure 3. The sprayed capsules had monodisperse distributions of which the geometric standard deviations (GSD) are 1.29, 1.26, and 1.28, respectively. Furthermore, as the flow rate of the 4-HBA was increased, from 2 to 4 mL/h, the maximum diameters and the mean sizes of the capsules increased linearly from 16.5 to 22.5 µm and from 16.9 to 21.4 µm, respectively. These tendencies were similar to those observed in a previous study using two immiscible fluids;8 namely, they are different than the Q1/2 and Q1/3 laws of single-nozzle electrospray9,15 because the olive oil, with its low electrical conductivity and high viscosity, influenced the breakup process. With the (15) Ferna´ndez de la Mora, J.; Loscertales, I. G. J. Fluid Mech. 1994, 260, 155– 184.

4646

Analytical Chemistry, Vol. 82, No. 11, June 1, 2010

electrospray using three immiscible fluids, the viscous stresses of the fluids interfaces should be also considered. Figure 4 shows the chemical components of the sprayed capsules measured by the high-resolution dispersive Raman microscope. The Raman spectra of the olive oil and EG were in good agreement with earlier studies,16,17 and the spectra of capsules clearly contained the bands from 4-HBA, olive oil, and EG. Consequently, we can demonstrate the sprayed capsules were composed of a core made of EG and two shells, made from olive oil and 4-HBA. We also visualized the sprayed capsules using the optical microscope and investigated that not only the capsule size, but also the shell thickness was affected by the fluid flow rates. Figure 5a-c correspond to the pictures of the multishell capsules upon increasing the flow rates of the 4-HBA from 2 to 4 mL/h. For the case (a), (b), and (c), the mean thicknesses of the outer shell are 1.9, 2.4, and 3.1 µm; namely, when the flow rate of 4-HBA increased, the compound capsule size, especially the outer shell thickness of the 4-HBA, increased. The pictures of the Figure 6 are the capsules upon increasing the flow rates of the olive oil from 0.2 to 2.6 mL/h and the mean diameters and GSD of the compound capsules in these cases are shown in Figure 7. The inner olive-oil-shell thickness increased with the nonconducting olive oil flow rate; the mean diameter of the olive oil being 11.4, 14.2, and 16.7 µm for the cases shown in Figure 6a-c. Moreover, the mean diameter of the compound capsule was also slightly elevated, although the increasing degree was insignificant compared with the case of the conducting 4-HBA fluid (see Figure 7). However, when the flow rates of olive oil was of 2.6 mL/h, as shown in Figure 6d, the capsules which had not core or had more than two cores in the inner shell were observed and GSD of the compound capsules was remarkably increased. This result can be explained by the necking timing difference of the fluids in the jet. When the flow rates of the sprayed fluids are changed, the breakup wavelengths of the fluids become also changed in the jet, and the necking timing difference is generated among fluids by alteration of the breakup process. (16) Krishnan, K.; Krishnan, R. S. Proc. Indian Acad. Sci. Math. Sci. 1966, 64, 111–122. (17) Yang, H.; Irudayaraj, J. J. Am. Oil Chem. Soc. 2001, 78, 889–895.

Figure 5. Generated multishell microcapsules under different flow rates of the 4-HBA. 4-HBA/Olive oil/EG [mL/h] (a) 2:1:1, (b) 3:1:1, (c) 4:1:1, Case (a): the mean diameter and the mean thickness of the outer shell are 16.9 and 1.9 µm. Case (b): the mean diameter and the mean thickness of the outer shell are 19.0 and 2.4 µm. Case (c): the mean diameter and the mean thickness of the outer shell are 21.4 and 3.1 µm, respectively.

Figure 6. Generated multishell microcapsules under different flow rates of the olive oil. 4-HBA/Olive oil/EG [mL/h] (a) 3:0.2:1, (b) 3:1:1, (c) 3:1.8:1, and (d) 3:2.6:1. (a) The mean diameter of the olive oil: 11.4 µm. (b) The mean diameter of the olive oil: 14.2 µm. (c) The mean diameter of the olive oil: 16.7 µm. (d) The mean diameter of the olive oil: 14.4 µm.

Consequently, the necking timing difference causes less uniform encapsulation. In practice, for the compound capsule at 2.6 mL/ h, the broader size distribution was measured and the mean diameter was not uniform as other cases because of the irregular encapsulation (Figure 7). CONCLUSIONS In conclusion, we developed an electrospray system that can produce multishell capsules using three immiscible fluids. This

Figure 7. Variation in the mean diameter and geometric standard deviation (GSD) of the capsules under different flow rates of the olive oil. For the Qolive oi) 0.2 mL/h, the mean diameter is 18.4 µm and GSD is 1.24. For the Qolive oi) 1.0 mL/h, the mean diameter is 19.0 µm, and GSD is 1.26. For the Qolive oi) 1.8 mL/h, the mean diameter is 19.7 µm and GSD is 1.27. For the Qolive oi) 2.6 mL/h, the mean diameter is 19.6 µm and GSD is 1.35, respectively.

system can generate a cone-jet mode, similar to single and dual coaxial nozzle electrospray systems, and the produced capsules have monodisperse size distributions. Furthermore, the size of the capsule and the thickness of each layer can be controlled by the fluid flow rates. We also investigated that the necking timing difference which is affected by breakup wavelengths of the fluids plays an important role in uniform encapsulation. However, the present study should be expanded to determine the stability criteria for the three fluids and to identify factors that influence the multishell encapsulations. In particular, because we focused on possible solidified multishell capsules that can be produced using the electrospray method, the working fluids used were not biocompatible materials. Accordingly, a synthetic polymer and biocompatible materials, such as PLGA, proteins, and biosuspensions, should be considered in future work. When biocompatible materials are applied to this system, the present results can be extended to targeted drug delivery and controlled release systems, and the synthesis of multilayered composite threads.

ACKNOWLEDGMENT This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Education, Science, and Technology (KRF-2008-313-D00114). This work was also partially supported at BK21 program of the South Korea Ministry of Education, Science, and Technology.

Received for review February 1, 2010. Accepted April 28, 2010. AC100278C

Analytical Chemistry, Vol. 82, No. 11, June 1, 2010

4647