Langmuir 2006, 22, 4397-4402
4397
Polymeric Nano/Microcapsules of Liquid Perfluorocarbons for Ultrasonic Imaging: Physical Characterization E. Pisani,†,‡ N. Tsapis,*,† J. Paris,† V. Nicolas,§ L. Cattel,‡ and E. Fattal† UMR CNRS 8612, School of Pharmacy, UniV. Paris Sud, France, Facolta` di Farmacia, UniVersita` degli Studi di Torino, Torino, Italy, and IFR141-ITFM, Plateau technique-Imagerie cellulaire, UniV. Paris Sud, France ReceiVed January 16, 2006. In Final Form: February 13, 2006 Ultrasonic imaging is a widely available, noninvasive, and cost-effective diagnostic modality, but vessels smaller than 200 µm in diameter are impossible to visualize. Commercial ultrasound contrast agents (UCAs), consisting of encapsulated gas microbubbles injected intravenously, enable only a qualitative visualization of the microvascularization for a short period of time since they are rather unstable. In a strategy to develop more stable UCAs, we designed a process to obtain nano/microcapsules with a single core of liquid perfluorocarbons within a biodegradable polymeric shell of homogeneous thickness. The polymer shell should improve the stability of the capsules as compared to UCAs stabilized by a monomolecular layer, while the acoustic impedance of the perfluorocarbons should ensure their echogenicity. These capsules have been optimized to encapsulate several liquid perfluorocarbons: perfluorohexane, perfluorodecalin, and perfluorooctyl bromide. The system is rather versatile: the mean size of the capsules can be adjusted between 70 nm and 25 µm and the thickness-to-radius ratio (T/R) can be easily modulated by simply modifying the polymer-to-perfluorocarbon ratio. T/R does not depend on the size of the capsules and is between 0.2 and 0.6. The dependence of the echogenic properties of the capsules with their size and their T/R has yet to be studied experimentally before this system can be evaluated in vivo.
Introduction Ultrasonic imaging is a widely available, noninvasive, and cost-effective diagnostic modality. However, differentiation of tissues of diagnostic importance is often hampered by similar levels of echogenicity. Furthermore, clinical Doppler is not able to image vessels smaller than 200 µm in diameter, thus preventing the mapping of the capillary network of an organ or a tumor. Commercial ultrasound contrast agents (UCAs), consisting of encapsulated gas microbubbles injected intravenously, enable the visualization of the microvascularization, but only a qualitative or, at the best, semiquantitative evaluation of blood flow can be achieved based on this visualization. The technical key necessary to reach this goal is the development of UCAs with greater stability. Elaboration of more stable UCAs would lead to a more widespread clinical use and take full benefit of harmonic imaging, which considerably enhances the signal-to-noise ratio.1,2 To increase the stability of UCAs, many authors have shown interest in polymers since polymeric shells are more resistant to ultrasonic waves than monomolecular layers of lipids or surfactants usually stabilizing commercial UCAs such as Levovist or Optison.3-6 On the other hand, perfluorocarbons, because of their difference of density with air and their poor solubility in * Corresponding author: Nicolas Tsapis, E-mail: nicolas.tsapis@ cep.u-psud.fr. † School of Pharmacy, Univ. Paris Sud. ‡ Universita ` degli Studi di Torino. § Plateau technique-Imagerie cellulaire, Univ. Paris Sud. (1) Gramiak, R.; Shah, P. M. Echocardiography of the aortic root. InVest. Radiol. 1968, 3 (5), 356-66. (2) Correas, J. M.; et al. Ultrasound contrast agents: properties, principles of action, tolerance, and artifacts. Eur Radiol. 2001, 11 (8), 1316-28. (3) Straub, J. A.; et al. Porous PLGA microparticles: AI-700, an intravenously administered ultrasound contrast agent for use in echocardiography. J. Controlled Release 2005, 108 (1), 21-32. (4) Cui, W.; et al. Preparation and evaluation of poly(L-lactide-co-glycolide) (PLGA) microbubbles as a contrast agent for myocardial contrast echocardiography. J. Biomed. Mater. Res. B: Appl. Biomater. 2005, 73 (1), 171-8. (5) El-Sherif, D. M.; et al. Ultrasound degradation of novel polymer contrast agents. J. Biomed. Mater. Res. A 2004, 68 (1), 71-8.
water, have been shown to increase both the stability and the echogenicity of UCAs.7 In addition to gaseous perfluorocarbons, Wickline et al.8 have shown that nanoemulsions of liquid perfluorocarbons in water also have interesting echogenic properties. We therefore chose to encapsulate perfluorocarbons in the liquid state within polymeric particles because of its easier feasibility. The challenge to overcome for perfluorocarbons encapsulation is their very low miscibility with hydrogenated organic solvents. We developed innovative UCAs consisting of a biodegradable polymeric shell encapsulating liquid perfluorocarbons based on a modified solvent emulsion/evaporation method. The biodegradable polymer shell should improve the stability of the capsules as compared to UCAs stabilized by a monomolecular layer, while the acoustic impedance of the perfluorocarbons should ensure the echogenicity.9 Once the objective of more stable UCAs is reached, one would be able to consider using ultrasounds to trigger the drug release from hybrid UCAs encapsulating a drug. Indeed, ultrasonic energy has been shown to have a role in increasing polymer degradation rates and therefore drug delivery. The ability to externally control degradation and drug release rates opens many doors to noninvasive, targeted drug-delivery systems. Additionally, ultrasound is a relatively safe triggering mechanism. These drugloaded UCAs would have a great potential for targeted treatment of diseases such as cancer, for which current systemic treatments have severe toxic side effects.10,11 (6) El-Sherif, D. M.; Wheatley, M. A. Development of a novel method for synthesis of a polymeric ultrasound contrast agent. J. Biomed. Mater. Res. A 2003, 66 (2), 347-55. (7) Quay, S. Microbubble-based ultrasound contrast agents: the role of gas selection in microbubble persistence. J. Ultrasound Med. 1994, 13, S9. (8) Wickline, S. A.; et al. Blood contrast enhancement with a novel, nongaseous nanoparticle contrast agent. Acad Radiol. 2002, 9 (Suppl. 2), S290-3. (9) Andre, M.; Nelson, T.; Mattrey, R. Physical and acoustical properties of perfluorooctylbromide, an ultrasound contrast agent. InVest. Radiol. 1990, 25 (9), 983-7. (10) Kost, J.; Leong, K.; Langer, R. Ultrasonically controlled polymeric drug delivery. Makromol. Chem. Macromol. Symp. 1988, 19, 275-285.
10.1021/la0601455 CCC: $33.50 © 2006 American Chemical Society Published on Web 03/09/2006
4398 Langmuir, Vol. 22, No. 9, 2006
We present here the details of the preparation method as well as the physical characterization of the polymeric nano/microcapsules encapsulating perfluorocarbons. Capsules have been characterized in terms of size, polydispersity, morphology, and encapsulation efficacy. The ability to modify particle thickness was also evaluated as this parameter is of utmost importance for the echogenic properties of the capsules. Materials and Methods Materials. Poly(lactide-co-glycolide) Resomer RG502 was provided by Boehringer-Ingelheim (Germany). Poly(vinyl alcohol) (PVA) (MW 30 000-70 000, 89% hydrolyzed), sodium cholate (SC), and Nile Red were provided by Sigma-Aldrich. Perfluorooctyl bromide (PFOB) was a gift from Alliance Pharmaceuticals. Perfluoropentane, perfluorohexane, perfluorooctane, and perfluorodecalin were provided by Fluorochem (United Kingdom). These perfluorocarbons will be denoted PFCs in the remainder of the paper. Water was purified using a Milli-Q system from Millipore (France). Methylene chloride RPE-ACS 99.5% was provided by Carlo Erba Reactifs (France). Sample Preparation. PLGA was dissolved into 4 mL of methylene chloride along with the desired amount of liquid PFC and placed in a thermostated bath maintained at 20 °C to ensure full miscibility of the PFC. The organic solution was then emulsified into 20 mL of 1.5% sodium cholate (w/v) aqueous solution using an Ultraturrax T25 (IKA) operating with a SN25-10G dispersing tool at a velocity between 8000 and 24 500 rpm. Emulsification was performed in a 50 mL beaker placed over ice for 2 min. Methylene chloride was then evaporated by magnetic stirring for about 3 h at 300 rpm in a thermostated bath (20 °C). For fluorescent or confocal microscopy, Nile Red was added to the organic solution prior to emulsification. Typically, about 100 µL of a concentrated Nile Red solution (0.057 mg/mL in methylene chloride) was added to the organic solution prior to emulsification. To further decrease the capsule size and obtain nanocapsules, a pre-emulsion was prepared by mixing the organic and aqueous phases by Ultra-turrax at 8000 rpm for 30 s. The pre-emulsion was then sonicated with a vibrating metallic tip, IBP7677 supplied by Ultrasons Annemasse (France), for 2 min over ice. The variation of sonication power is expressed as a function of the applied voltage. The voltage range is between 40 and 140 V. The rest of the process is the same as that described for microcapsules. Freeze Drying. Fresh capsules are frozen at -20 °C after addition of PVA (final concentration 0.2% (w/v)) as a cryoprotectant.12 Samples are then freeze dried for 24-48 h using a LYOVAC GT2. Optical and Fluorescence Microscopy. Samples in water were placed between glass slides and observed with a Leitz Diaplan microscope equipped with a Coolsnap ES camera (Roper Scientific). Fluorescent samples dyed with Nile Red were excited at 543 nm and observed at 560 nm (long-pass filter). Confocal Microscopy. Glass slides were examined with a Zeiss LSM-510 confocal scanning laser microscope equipped with a 1 mW helium neon laser, using a Plan Apochromat 63X objective (NA 1.40, oil immersion). Red fluorescence was observed with a long-pass 560 nm emission filter under 543 nm laser illumination. The pinhole diameter was set at 71 µm. Stacks of images were collected every 0.42 µm along the z axis. Particle size and shell thickness were measured directly on the confocal images using either the Zeiss software or ImageJ (Scion Corporation). Measurements were carried out in the equatorial plane of each capsule to minimize the error due to the position of the slice. The size of a pixel was 70 nm. Granulometry. Size measurements on microcapsules were performed using a LS230 Coulter-Beckmann granulometer based (11) Kost, J.; Leong, K.; Langer, R. Ultrasound-enhanced polymer degradation and release of incorporated substances. Proc. Natl. Acad. Sci. U.S.A. 1989, 86 (20), 7663-6. (12) Abdelwahed, W.; Degobert, G.; Fessi, H. A pilot study of freeze-drying of poly (epsilon-caprolactone) nanocapsules stabilized by poly(vinyl alcohol): formulation and process optimization. Int. J. Pharm. 2006, 309 (1-2), 178-188.
Pisani et al. on laser diffraction. Drops of the capsules suspension were added to water in the measurement cuvette. Measurements were performed in triplicate. Size distribution was analyzed using either the Fraunhoffer model or the Mie model, according to the size observed with microscopy. Quasi-Elastic Light Scattering. Size measurements of the nanocapsules were performed using a Malvern Zetasizer Nano ZS based on quasi-elastic light scattering. Measurements were performed in triplicate at an angle of 173° to avoid multiple scattering. Scanning Electron Microscopy. Scanning electron microscopy (SEM) was performed using a LEO 1530 (LEO Electron Microscopy Inc, Thornwood, NY) operating between 1 and 3 kV with a filament current of about 0.5mA. Liquid samples were deposited on carbon conductive double-sided tape (Euromedex, France) and dried at room temperature. They were coated with a palladium-platinum layer of ca. 4 nm using a Cressington sputter-coater 208HR with a rotaryplanetary-tilt stage, equipped with a MTM-20 thickness controller. Particles were washed before imaging by either centrifugation or dialysis to remove the excess amount of surfactant that reduces the quality of the images. Transmission Electron Microscopy (TEM). transmission electron microscopy was performed using a Philips EM208 operating at 80 kV. Suspensions of nanocapsules were deposited on copper grids covered with a formwar film (400 mesh) for 2 min. The excess solution was blotted off using filter paper, and grids were air dried before observation. Images were acquired using a high-resolution camera, Advantage HR3/12GO4 (AMT-Hamamatsu). Freeze-Fracture Electron Microscopy (FFEM). Electron microscopy observations were preceded by freeze-fracture of samples. Samples were placed on a copper holder and then snapfrozen into liquid propane. Frozen samples were fractured under vacuum (10-7 Torr) with a single edge scalpel maintained at 77 K. Freeze-fracture and replication were successively performed using a Balzers BAF 400T apparatus (BAL-TEC, Liechtenstein). Replication of the surface exposed to freeze-fracture was achieved in two steps: 2 nm of platinum was evaporated from an oblique angle (45°) to provide contrast enhancement of the surface topology; a thicker continuous layer (20 nm) of an electron transparent material (carbon) was then deposited at an angle of 90°. The carbon layer allowed strengthening of the replica. Layer thickness was controlled by a quartz crystal gauge. Replicas were then washed using acetone and then water to eliminate the underlying sample. Replicas were examined using a LEO 912 electron microscope equipped with an omega filter working at 120 kV. Gas Chromatography Coupled with Mass Spectroscopy (GCMS). Perfluorooctyl bromide analysis was achieved by slightly modifying the method of Audran et al.13 Briefly, headspace analysis was performed with an HP 5989A mass spectrometer coupled with an HP5890 gas chromatography. The chromatographic separation was performed by injection in the split mode (split, 40 mL/min) of 1 µL of the headspace gas of the sample in a Varian WCOT fused silica capillary column (length, 60 m; internal diameter, 0.32 mm; film thickness, 1.8 µm). The injector temperature was 200 °C, and the initial oven temperature was 40 °C, maintained for 1 min. The temperature was then programmed as follows: 15 °C/min up to 250 °C, maintained 1 min. The transfer line temperature was set to 200 °C. Analysis was performed by electronic impact ionization with an electron energy of 70 eV. The ion source temperature was set to 200 °C and the quadrupole to 100 °C. Samples were heated at 60 °C for 30 min to ascertain that the thermodynamical balance was reached. The gas was then taken with a gas syringe. PFOB was analyzed in the SIM mode with the following selected ions due to their abundance and specificity: m/z 69, 131, 169, 219, 331, and 419. Headspace of PFOB in methylene chloride was injected. One peak was observed in chromatography around 12 min corresponding to PFOB specific ions. 13C Nuclear Magnetic Resonance (NMR) Spectra. The 13C nuclear magnetic resonance (NMR) spectra were recorded on a Bruker (13) Audran, M.; et al. Assay method for the perfluorooctyl bromide (perflubron) in rat blood by gas chromatography-mass spectrometry. J. Chromatogr. B: Biomed. Sci. Appl. 1999, 734 (2), 267-76.
Polymeric Nano/Microcapsules
Langmuir, Vol. 22, No. 9, 2006 4399
AC 200 instrument (Karlsruhe, Germany) in CDCl3 solution at room temperature with SiMe4 as an internal standard.
Results and Discussion We developed an echographic contrast agent candidate based on the encapsulation of liquid perfluorocarbon within a biodegradable14 and biocompatible15 polymeric shell. The method used to obtain nano/microcapsules composed of a solid polymeric shell encapsulating a liquid perfluorocarbon core is derived from the work of Loxley et al.16 It is a modification of the commonly used emulsion-evaporation process. The organic phase is a mixture of PLGA, methylene chloride (a low-boiling good solvent for PLGA), and liquid perfluorocarbon (a high-boiling very poor solvent for PLGA). Enough methylene chloride is present to ensure that PLGA is completely dissolved and that the liquid perfluorocarbon is fully miscible. The organic phase is then emulsified in an aqueous solution of surfactant, and the lowboiling solvent is evaporated. This process allows one to gradually remove the low-boiling solvent from the emulsion droplets. Since perfluorocarbons are very poorly miscible in methylene chloride, the droplet composition quickly reaches the binodal boundary and the polymer phase separates as small droplets of liquid which are rich in solvent and polymer within the emulsion droplets. These polymer-rich droplets are mobile and migrate to the oil/ water interface where they fuse and spread to engulf the original perfluorocarbon droplet if the wetting conditions are correct. Further solvent removal causes the polymer to precipitate at the interface, forming the shell.16 If the wetting conditions are not correct, other morphologies than nice spherical core-shell structure can be observed. These morphologies have been theoretically investigated by Torza and Mason.17 Briefly, if droplets of immiscible liquids (phases 1 and 3) are brought in contact in a third mutually immiscible liquid (phase 2), the final equilibrium morphology can be rationalized by analyzing the various interfacial tensions between the phases (γ12, γ23, and γ13). By defining the spreading coefficients Si for each phase as
Si ) γjk - (γij + γik)
Figure 1. Expected morphologies when eq 2 is satisfied (left), eq 3 is satisfied (middle), and eq 4 is satisfied (right)
(1)
and designating phase 1 to be that for which γ12 > γ 23, then S1 < 0. It then follows that there are only three possible combinations of Si
S1 < 0; S2 < 0; S3 > 0
(2)
S1 < 0; S2 < 0; S3 < 0
(3)
S1 < 0; S2 > 0; S3 < 0
(4)
When the conditions in eq 2 are satisfied, the particles adopt a core-shell morphology with phase 1 appearing as the core within a shell of phase 3. When eq 3 is satisfied “acorn”-shaped particles are formed, and when eq 4 is satisfied two separate droplets are preserved (Figure 1). In our study it is essential to avoid both the “acorn”-shaped morphology and separate droplets to ensure total encapsulation (14) Reed, A. M.; Gilding, D. K. Biodegradable polymers for use in surgerypoly(glycolic)/polymer (lactic acid) homo and copolymers. In vitro degradation. Polymer 1981, 22, 494-498. (15) Yamaguchi, K.; Anderson, J. M. In vivo biocompatibility studies of Medisorb 65/35 D,L-lactide/glycolide copolymer microspheres. J. Controlled Release 1993, 24, 81-93. (16) Loxley, A.; Vincent, B. Preparation of Poly(methylmethacrylate) Microcapsules with Liquid Cores. J. Colloid Interface Sci. 1998, 208 (1), 49-62. (17) Torza, S.; Mason, S. G. Three-phase interactions in shear and electrical fields. J. Colloid Interface Sci. 1970, 33 (1), 67-83.
Figure 2. Microscopy images of a suspension of microcapsules. Bright field is presented on the top, whereas fluorescence is on the bottom (the polymer appears bright, and the liquid perfluorocarbon appears dark). Scale bars represent 20 µm in both images.
of the PFCs. However, since our system is not at equilibrium due to the constant evaporation of the organic solvent, it is impossible to measure all interfacial tensions. We therefore reviewed a variety of surfactants and finally used sodium cholate. This surfactant allows us to reach the right conditions and obtain only particles with core-shell structure. This result is already obvious observing the suspension of microcapsules by optical microscopy (Figure 2, top). To underline the structure of the capsules, we added a fluorescent marker, Nile Red, to the organic solution prior to emulsification. This marker colors the hydrophobic polymer red but does not color PFCs. Fluorescent microscopy images show spherical particles with a nice red shell of homogeneous thickness and a darker core (Figure 2, bottom). Moreover, the different slices obtained by confocal microscopy prove that the cavities are well centered within the particles and that the shell thickness is homogeneous for each particle (Figure 3, top). Several PFCs were evaluated18 for encapsulation within polymeric shells; they were chosen with diverse boiling points ranging from 36 °C for perfluoropentane to 140 °C for
4400 Langmuir, Vol. 22, No. 9, 2006
Pisani et al. Table 1. Physicochemical Characteristics of the Different Perfluorocarbons Used in This Study PFC
MW
density (g/mL)
boiling point
perfluoropentane perfluorohexane perfluorooctyl bromide perfluorodecalin
288.03 338.04 498.9 462.08
1.62 1.669 1.9 1.92
26-36 °C 58-60 °C 140.5 °C 144 °C
Table 2. Size Distribution of Microcapsules as a Function of the Emulsification Speed and Sodium Cholate (SC) Concentration emulsification speed
[SC]
D10 (µm)
D50 (µm)
D90 (µm)
1000 rpm 8000 rpm 9500 rpm 13 500 rpm 20 500 rpm 24 500 rpm 24 500 rpm
3% 1.5%
10.4 4.9 2.6 0.9 0.8 0.8 0.7
25.0 6.7 3.5 2.1 1.8 1.7 1.3
54.3 8.7 4.7 3.5 2.9 3.0 2.5
3%
Table 3. Size Distribution of Nanocapsules as a Function of the Variation of Sonication Power (expressed as applied voltage)
Figure 3. Confocal microscopy image (scale bar ) 10 µm). PLGA is dyed in red, whereas PFC appears dark (top). SEM images of microcapsules; they present a smooth surface. The insert is a zoom on a collapsed capsule, proving the existence of cavities within the spheres (bottom).
perfluorooctyl bromide. Capsules could be obtained with perfluorooctyl bromide (PFOB), perfluorodecalin (PFD), and perfluorohexane (PFH) but not with perfluoropentane. The failure of perfluoropentane encapsulation probably arises from the lowboiling point of this chemical, close to the boiling point of methylene chloride. We decided to pursue experiments with PFOB since no toxicity has been reported for this chemical19,20 and because of its high boiling point. The morphology and porosity of the capsules have been studied by SEM. The majority of capsules are spherical with smooth surfaces. A few collapsed capsules can be observed. This finding confirms the existence of cavities within the spherical particles (Figure 3, bottom). Optical microscopy images show a single spherical core, and no meniscus can be distinguished within the core. This absence of meniscus indeed proves that the entire core of the capsule is either liquid or gaseous. 13C NMR spectra of the capsules performed before and after freeze drying and after dissolution in deuterated methylene chloride clearly exhibit the characteristic peaks of perfluorooctyl bromide. Although qualitative, this result confirms that the inner core of the capsules consists of perfluorooctyl bromide. Moreover, when capsules are dissolved (18) Marsh, J. N.; et al. Improvements in the ultrasonic contrast of targeted perfluorocarbon nanoparticles using an acoustic transmission line model. IEEE Trans. Ultrason. Ferroelectr. Freq. Control 2002, 49 (1), 29-38. (19) Leese, P. T.; et al. Randomized safety studies of intravenous perflubron emulsion. I. Effects on coagulation function in healthy volunteers. Anesth. Analg. 2000, 91 (4), 804-11. (20) Noveck, R. J.; et al. Randomized safety studies of intravenous perflubron emulsion. II. Effects on immune function in healthy volunteers. Anesth. Analg. 2000, 91 (4), 812-22.
voltage
D50 (nm)
distribution width (nm)
140 V 80 V 40 V
72 139 209
36 105 147
into methylene chloride and these samples analyzed by GC-MS (headspace mode), we observe the peak corresponding to PFOB according to the retention time and specific ions. Taken together, these results further confirm the presence of liquid PFOB within the capsules. For microbubbles contrast agents echogenicity depends strongly on the size. To the best of our knowledge, neither theoretical nor experimental works have been performed on the effect of size on echogenicity of solid capsules with a liquid core. This type of experiment will be feasible in the future since our preparation method allows for adjustment of capsule size. A simple way to tune capsule size is to vary the emulsification speed as presented in Table 2. The size range can be widened by also changing the concentration of sodium cholate, and finally, one can obtain capsules with a mean size between 1 and 25 µm. Size can even be further reduced, and one can obtain nanocapsules using the same process only by performing emulsification with an ultrasound tip. Adjustment of the sonication power is used to vary nanocapsule size, and one can obtain a mean size ranging from 70 to 200 nm (Table 3). For capsules down to 1 µm in diameter, fluorescent microscopy allows verification of the preservation of the core-shell morphology: even for the smaller capsules one can observe a bright shell around a darker core. To ascertain that the coreshell morphology was preserved even for nanocapsules, FFEM and TEM were performed. Images present spherical capsules with a core-shell structure. Although the shell is difficult to distinguish on the FFEM picture, there is an obvious difference of electronic density visible on the TEM picture between the core and the shell (Figure 4). The mechanical properties of our capsules are very important for their use as UCAs since ultrasounds are mechanical waves. In classical mechanics of shells, mechanical properties usually depend on the shell thickness-to-radius ratio.21 This dimensionless parameter is probably important for our system. We therefore characterized our microcapsules by measuring their thickness (T) and radius (R). These measurements have been carried out directly on confocal microscopy slices. In the first set of (21) Landau, L. D.; Lifshitz, E. M. Mechanics, 3rd ed.; Pergamon Press: New York, 1976; Vol. 1.
Polymeric Nano/Microcapsules
Langmuir, Vol. 22, No. 9, 2006 4401
Figure 6. Confocal images of microcapsules prepared with the same volume of PFOB and various masses of PLGA: 0.5 (left), 0.1 (middle), and 0.04 g (right). The scale bar represents 5 µm. Figure 4. FFEM image of a typical nanocapsule: the core-shell structure is more obvious on the left side of the picture; the scale bar represents 200 nm (left). TEM image of a typical nanocapsule: because of the difference of electronic density, the PFOB liquid core appears in gray, whereas the polymeric shell seems darker; the scale bar represents 100 nm (right).
Figure 5. Variation of the shell thickness as a function of the radius for microcapsules ranging between 2 and 9 µm and for nanocapsules with diameters between 140 and 340 nm (insert). Straight lines represent the best linear fit. Both fits have a slope of 0.35 ( 0.04.
experiments the perfluorooctyl bromide volume is fixed as well as the PLGA mass, and we vary only the speed of the emulsion blender. The thickness is presented as a function of the radius. The thickness increases as the radius increases. Experimental points are well fitted by a straight line: the thickness varies linearly with the capsule radius. In the case of 60 µL of PFOB and 100 mg of PLGA, the slope of the fit gives 0.35 ( 0.01 (Figure 5). For capsule diameters below 2 µm, the confocal microscopy resolution becomes poor and the error of the measurements becomes too important. Crude measurements performed on TEM/FFEM images of nanocapsules obtained with the same PFOB/PLGA proportions also present a linear variation of the shell thickness with the radius. The slope of the linear fit is the same as that for microcapsules: 0.35 ( 0.04 (Figure 5). As already observed with the morphology, physical characteristics of nanocapsules do not differ from those of microcapsules. We can therefore conclude that the shell thickness-to-radius ratio depends only on the PFOB/PLGA proportions. The effect of freeze drying on the physical characteristics of microcapsules has been evaluated. Size and thickness measurements performed on the same sample before and after freeze drying do not exhibit significant differences. Our system can therefore withstand freeze drying without its physical properties being modified, which is a key advantage for storage. To evaluate the range of thickness-to-radius ratio possible to obtain with our process, the PLGA to PFOB proportions are varied. In one set of experiments the volume of PFOB is fixed to 60 µL and the PLGA mass is varied from 10 to 500 mg, keeping everything else constant. For each PFOB to PLGA proportion the thickness and radius are measured on confocal
Figure 7. Variation of the value of T/R as a function of the mass of PLGA (top) and the volume of PFOB volume (bottom). The mass of PLGA is between 0.04 and 0.5 g, and the volume of PFOB is between 0.02 and 0.06 mL.
slices. Confocal images show that the higher the mass of PLGA, the thicker the shells (Figure 6). The variation of T as function of R is plotted for a given proportion. As shown before, the variation is well fitted by a straight line (not shown here). The slope corresponds to the thickness-to-radius ratio (T/R). Values of the slope (T/R) for a given PFOB to PLGA proportion are then reported as a function of the PLGA mass (Figure 7, top). As already observed on the confocal images, T/R increases as the PLGA mass increases. For very low masses of polymer, 0.02 and 0.01 g, we are able to observe microcapsules with fluorescence microscopy but the shell thickness is smaller than the resolution of the microscope. These data points are not reported in Figure 7. The thickness-to-radius ratio we obtain from the linear fits is between 0.2 and 0.6. Additional experiments, such as cryoultramicrotomy, should be performed to determine accurately the shell thickness for the very low masses of PLGA (mPLGA e 0.02 g). In another set of experiments, to check how much PFOB could be encapsulated into polymeric shell, the amount of polymer in the organic solvent is fixed (0.1 g) and the volume of PFOB varies between 0.02 and 0.15 mL. For PFOB volumes larger
4402 Langmuir, Vol. 22, No. 9, 2006
Pisani et al.
than 0.06 mL, we observe a coexistence of PFOB droplets along with nice microcapsules of PFOB covered with polymeric shells. The nice shells are visible with fluorescence microscopy, whereas PFOB droplets do not fluoresce and are visible only with bright field microscopy. This coexistence probably can be explained by the very low miscibility of perfluorocarbons with the hydrogenated organic solvent. Miscibility is further reduced when samples are placed at 4 °C for emulsification and PFC droplets appear within the organic phase. These droplets are therefore free in the aqueous phase after emulsification. For PFOB volumes between 0.02 and 0.06 mL it seems that all the PFOB is encapsulated. For these volumes measurements on confocal images have been carried out and T has been plotted as function of R. Once again, T varies linearly with R and the slope of the fit gives T/R (not shown here). The slope is reported as a function of the initial PFOB volume. T/R decreases as the volume of PFOB increases (Figure 7, bottom). This result is in agreement with confocal microscopy observations: the smaller the PFOB volume, the thicker the shells (not shown). As T/R can be determined experimentally, one can therefore deduce from it an experimental value of the volume fraction of PFOB, φPFOB since
4 π(R - T)3 Vcore 3 T3 φPFOB ) ) ) 1- Z Vcore + Vshell 4 3 R πR 3
(
)
(5)
where Vcore and Vshell are, respectively, the volume of the core and shell of the capsule. The theoretical PFOB volume fraction value can also be calculated from the amounts of PFOB and PLGA used for the preparation, assuming that the density F of PLGA is around 1 g/cm3
theoretical φPFOB )
VPFOB VPFOB + F -1mPLGA
(6)
where VPFOB is the volume of PFOB and mPLGA the mass of PLGA used to prepare the emulsion. When the experimental value is plotted as a function of the theoretical value for the two sets of experiments where the volume of PFOB is varied or the mass of PLGA is varied, experimental points can be fitted by a straight line (Figure 8). In both cases, the slope of the best linear fit gives 0.7 ( 0.1. The fact that the slope differs from 1 proves that about 30% of the initial PFOB is lost during sample preparation. PFOB probably evaporates partially along with methylene chloride.
Conclusion We developed a process to obtain nano/microcapsules with a single core of liquid perfluorocarbons within a polymeric shell
Figure 8. Experimental PFOB volume fraction presented as (1 - T/R)3 as a function of theoretical PFOB volume fraction calculated from the initial amount of chemicals used for sample preparation. Results for a fixed mass of PLGA (100 mg, b) and for a fixed volume of PFOB (60 µL, ]) are well fitted by a straight line going through zero. The slope is the same in both cases: 0.7 ( 0.1.
(PLGA) of homogeneous thickness. These capsules have been characterized thoroughly. Our method of preparation based on a modified solvent emulsion evaporation process allows one to obtain a rather versatile system. The mean size of the capsules can be adjusted between 70 nm and 25 µm; several liquid perfluorocarbons can be encapsulated within the polymeric shell: perfluorohexane, perfluorodecalin, and perfluorooctyl bromide. The thickness-to-radius ratio of the capsules can be easily varied by simply modifying the polymer-to-perfluorocarbon ratio in the organic phase prior to emulsification. The range of thickness-to-radius ratio is between 0.2 and 0.6 for PFOB. Thinner shells can possibly be obtained, but the thickness has yet to be measured experimentally. Preliminary echogenicity results with both fresh and freeze-dried capsules are very promising and will be detailed in a forthcoming article. This versatile process could be adapted to other biodegradable polymers. Acknowledgment. Authors acknowledge financial support from ANR. E.P. acknowledges the financial support from the GALENOS-Network. The authors thank D. Jaillard and J. Degrouard (CCME, Orsay) for TEM experiments, G. Fre´bourg and J.-P. Lechaire (UMR CNRS 7622, Universite´ Pierre et Marie Curie) for FFEM experiments, A. Valette for access to the SEM facility, M. Besnard for her help with granulometry, A. Solgadi, D. Libong, and P. Chaminade for their time and advice for GC-MS, and S. Arpicco for 13C NMR. LA0601455
