488
Langmuir 2007, 23, 488-495
pH-Controlled, Polymer-Mediated Assembly of Polymer Micelle Nanoparticles Sang Cheon Lee* and Hong Jae Lee Nanomaterials Application DiVision, Korea Institute of Ceramic Engineering and Technology, Seoul 153-801, Korea ReceiVed August 9, 2006. In Final Form: October 17, 2006 We describe pH-controlled, polymer-mediated assembly of polymer micelles in aqueous media based on reversible complexation between the micelles of pyrene-labeled poly(-caprolactone)-b-poly(carboxylic acid) copolymers and proton-accepting water-soluble polymers such as poly(ethylene glycol) (PEG), poly(2-ethyl-2-oxazoline) (PEtOz), and poly(1-vinylpyrrolidone) (PVP). The key factor determining assembly phenomena was identified as the modulation of hydrogen-bonding interaction between ionizable anionic micellar shells and the proton-accepting polymers by the pH control. As pH decreased from 7.4 to 2.0, the mixture of the polymer micelles and polymers underwent assembly and formed solid hybrids at specific pH values. The micelles assembled in the hybrid could be reversibly dispersed as micelles above specific pH ranges. The assembly/disassembly behavior as well as phase transitions of the micelle/ proton-accepting polymer could be precisely controlled by adjusting pH. This assembling behavior depended on the rationally designed parameters such as the chemical structure and length of micellar shell-forming poly(carboxylic acid)s and the class of proton-accepting polymers.
Introduction A diverse class of polymer/nanoparticle hybrids constructed by polymer-mediated assembly of nanoparticles has attracted expanding interest due to its usefulness in generating novel assembly systems with tunable properties for various applications such as biotechnology and microelectronics.1-5 To date, for creating nanohybrids featuring programmed structures and properties, many classes of polymers and nanoparticles have been designed for specific interspecies interaction such as hydrogen bonding and oppositely charged ionic interactions.1,3,6-8 The candidates for polymer systems were dendrimers and homopolymers or block copolymers with specific ligands capable of interacting with nanoparticles.1,3,6-8 For nanoparticles, surfacefunctionalized metal and semiconductor nanoparticles such as gold, iron oxide, and palladium were the main components.2,9,10 As representative examples for polymer-mediated nanoparticle assembly, Rotello et al. demonstrated that polymers functionalized with recognition elements could be used for the ordering of gold or iron oxide nanoparticles into structured assemblies through hydrogen-bonding interaction.1,2 Recently, polymer-mediated assembly of nanoparticles that can sense external signals and response in the form of the change in phase, shape, and structure has been an issue in the growing field of intelligent nanohybrids. Particularly, the precise control of phase behavior or photophysical properties of assemblies in solutions is one of the key issues.11,12 * To whom correspondence should be addressed: Sang Cheon Lee, Ph.D., Nanomaterials Application Division, Korea Institute of Ceramic Engineering and Technology, 233-5 Gasan-Dong, Guemcheon-Gu, Seoul 153-801, Korea, Tel 82-2-3282-2469, Fax 82-2-3282-7811, E-mail:
[email protected]. (1) Boal, A. K.; Ilhan, F.; DeRouchey, J. E.; Thurn-Albrecht, T.; Russell, T. P.; Rotello, V. M. Nature (London) 2000, 404, 746-748. (2) Boal, A. K.; Frankamp, B. L.; Uzun, O; Tuominen, M. T.; Rotello, V. M. Chem. Mater. 2004, 16, 3252-3256. (3) Frankamp, B. L.; Uzun, O.; Ilhan, F.; Boal, A. K.; Rotello, A. M. J. Am. Chem. Soc. 2002, 124, 892-893. (4) Shenhar, R.; Norsten, T. B.; Rotello, V. M. AdV. Mater. 2005, 17, 657669. (5) Shenhar, R.; Rotello, V. M. Acc. Chem. Res. 2003, 36, 549-561. (6) Frankamp, B. L.; Boal, A. K.; Rotello, V. M. J. Am. Chem. Soc. 2002, 124, 15146-15147.
Polymeric micelles assembled from amphiphilic block copolymers have found a wide range of applications including drug delivery, encapsulation, and nanoreactors.13-20 The unique characteristics of polymer micelles as nanoparticles lie in nanosize, core-shell structure, and easily controllable surface functionality. Thus, polymer micelles may act as a good building block for constructing new kinds of polymer/nanoparticle assembly. In particular, polymer/micelle nanohybrids of which the assembly/ disassembly state can be controlled in response to external stimuli may open the new application field. To precisely control the pH-dependent polymer-mediated assembly of polymer micelles, it is important to investigate the influence of the design parameters such as the type of the protonaccepting polymers and the chemical structure of micellar shellforming polymers. The deep understanding of the relationship between the design parameters and the stimuli-sensitive assembly behavior may allow us to generate the novel polymer/micelle nanohybrids of which the assembly/disassembly state can be precisely controlled in response to external stimuli. Our aim in this work is to describe the precise control of polymer-mediated reversible assembly of polymer micelles by considering the design parameter of systems. The main variables are the class of the proton-accepting polymers and the chemical structure of the poly(carboxylic acid) micellar outer shell. We synthesized a series of block copolymers to design the polymer micelles with the poly(-caprolactone) (PCL) core and the shell of ionizable anionic poly(carboxylic acid)s, such as poly(acrylic acid) (PAA) or poly(methacrylic acid) (PMAA). PEG, PEtOz, and PVP that have the different capability for hydrogen bonding with the micellar outer shells were utilized as polymer components. In aqueous media, the dependency of pH-controlled reversible polymer-mediated assembly on systematic design (7) Krasteva, N.; Besnard, I.; Guse, B.; Bauer, R. E.; Mu¨llen, K.; Yasuda, A.; Vossmeyer, T. Nano Lett. 2002, 2, 551-555. (8) Vossmeyer, T.; Guse, B.; Besnard, I.; Bauer, R. E.; Mu¨llen, K.; Yasuda, A. AdV. Mater. 2002, 14, 238-242. (9) Galow, T. H.; Drechsler, U.; Hanson, J. A.; Rotello, V. M. Chem. Commun. 2002, 1076-1077. (10) Srivastava, S.; Verma, A.; Frankamp, B. L.; Rotello, V. M. AdV. Mater. 2005, 17, 617-621.
10.1021/la0623580 CCC: $37.00 © 2007 American Chemical Society Published on Web 11/23/2006
Assembly of Polymer Micelle Nanoparticles
parameters was examined by fluorescence spectroscopy, dynamic light scattering, and transmission electron microscopy (TEM). Experimental Section Materials and Equipment. Ethylene glycol (EG), tert-butyl acrylate (tBA), tert-butyl methacrylate (tBMA), poly(ethylene glycol) (PEG) (Mn ) 3400, GPC), poly(1-vinylpyrrolidone) (PVP) (Mn ) 10 000, GPC), trifluoroacetic acid (TFA), 1-pyrenebutyric acid, 1,3dicyclohexylcarbodiimide (DCC), 4-dimethylaminopyridine (DMAP), copper (I) bromide (Cu(I)Br, 99.999%), N,N,N′,N′,N′′-pentamethyldiethylenetriamine (PMDETA), stannous octoate (Sn(Oct)2), and anhydrous magnesium sulfate were purchased from Aldrich Co. (Milwaukee, WI) and used without further purification. 2-Bromopropionyl bromide (BPB) was purchased from Aldrich Co. (Milwaukee, WI) and freshly distilled under vacuum. -Caprolactone (CL), 2-ethyl-2-oxazoline (EtOz), and methyl p-toluenesulfonate were purchased from Aldrich Co. (Milwaukee, WI), dried over calcium hydride, and vacuum-distilled. Tetrahydrofuran (THF) and toluene were distilled from Na/benzophenone under N2 prior to use. Triethylamine (TEA), acetonitrile, and methylene chloride were dried and distilled over calcium hydride. Chloroform, n-hexane, 35% hydrogen chloride (HCl), and sodium hydroxide (NaOH) were of reagent grade. The 1H NMR spectra were recorded at 400 MHz on a Varian INOVA400 NMR spectrometer with a sample spinning rate of 5 kHz at 25 °C. Molecular weight distributions were determined using a GPC equipped with a Waters 2414 refractive index detector, 515 HPLC pump, and three consecutive Styragel columns (HR1, HR2, and HR4). The eluent was THF with a flow rate of 1 mL/min. The molecular weights were calibrated with polystyrene standards. Synthesis of 2-Hydroxyethyl 2′-Bromopropionate (HEBP). A heterobifunctional initiator HEBP was synthesized by following a literature procedure.13 In brief, a solution of EG (43 g, 0.69 mol) and TEA (8.4 g, 0.08 mol) in dry THF (250 mL) was cooled to 0 °C, and BPB (15 g, 0.069 mol) in dry THF (50 mL) was then added dropwise to the stirred solution. The pure HEBP was isolated by vacuum distillation. Elemental analysis showed that high purity of HEBP was obtained. Synthesis of Pyrene-Labeled PCL-b-Poly(acrylic acid) (PyPCL-b-PAA) and PCL-b-Poly(methacrylic acid) (Py-PCL-bPMAA). Py-PCL-b-PAA and Py-PCL-b-PMAA were prepared by the following three-step synthetic route: (i) Synthesis of R-bromopropionyl-ω-hydroxy PCL (HO-PCL-Br), (ii) pyrene-labeled PCL-b-poly(tert-butyl acrylate) (Py-PCL-b-PtBA) and PCL-b-poly(tert-butyl methacrylate) (Py-PCL-b-PtBMA), and (iii) deprotection of Py-PCL-b-PtBA and Py-PCL-b-PtBMA. For HO-PCL-Br, HEBP (3.44 g, 0.0175 mol) in distilled toluene (90 mL) was dried azeotropically, and -caprolactone (30 g, 0.26 mol) was added in the HEBP solution. The polymerization was initiated by the addition of 0.14 g (0.35 mmol) of Sn(Oct)2 at 120 °C, and the reaction mixture was stirred under nitrogen for 24 h. The HO-PCL-Br was isolated by precipitation from toluene into n-hexane (1 L). Yield 93%. To prepare pyrene-labeled PCL-b-poly(tert-butyl acrylate) (Py-PCLb-PtBA) and PCL-b-poly(tert-butyl methacrylate) (Py-PCL-bPtBMA), the HO-PCL-Br macroinitiator (3 g, 1.8 mmol) and Cu(I)Br (0.49 g, 7.1 mmol) were added to a flame-dried round-bottom flask. Toluene (10 mL) and tBA or tBMA were degassed by N2 and added to the flask, and then PMDETA (1.48 g, 7.1 mmol) was introduced. The reaction was maintained at 85 °C for 15 h. The block copolymers were purified by passing through the silica gel column and precipitation from THF into cold n-hexane. Pyrene was labeled at the terminal of the PCL block by DCC chemistry. The block copolymers were denoted as Py-PCL-b-PtBA28, Py-PCL-b-PtBA45, and Py-PCL-b-PtMBA28, respectively. Finally, the deprotection of (11) Mori, H.; Lanzendo¨rfer, M. G.; Mu¨ller, A. H. E. Langmuir 2004, 20, 1934-1944. (12) Mori, H.; Mu¨ller, A. H. E.; Klee, J. E. J. Am. Chem. Soc. 2003, 125, 3712-3713. (13) Lee, S. C.; Kim, K. J.; Jeong, Y.-K., Chang, J. H.; Choi, J. Macromolecules 2005, 38, 9291-9297. (14) Lee, S. C.; Chang, Y.; Yoon, J.-S.; Kim, C.; Kwon, I. C.; Kim, Y.-H.; Jeong, S. Y. Macromolecules 1999, 32, 1847.
Langmuir, Vol. 23, No. 2, 2007 489 Py-PCL-b-PtBA and Py-PCL-b-PtBMA was performed by treating the block copolymers (3 g) with TFA (10 mL, 0.13 mol) in methylene chloride (10 mL). The aqueous solution was then dialyzed using a membrane (molecular weight cutoff (MWCO): 1000) for 24 h, followed by freeze-drying. The final block copolymers were denoted as Py-PCL-b-PAA28, Py-PCL-b-PAA45, and Py-PCL-b-PMAA28, respectively. Synthesis of Poly(2-ethyl-2-oxazoline) (PEtOz). A solution of 2-ethyl-2-oxazoline (10 g, 0.1 mol) and methyl p-toluenesulfonate (0.23 g, 1.2 mmol) in dry acetonitrile (100 mL) was stirred at reflux for 30 h under nitrogen. The product, PEtOz, was isolated following a reference procedure.14 Mn of PEtOz estimated by GPC was 8000. Fluorescence Measurements. Pyrene emission spectra were recorded on a JASCO FP-6500 spectrofluorometer. For the measurement of pyrene emission spectra, emission and excitation band widths were set at 1 and 5 nm, respectively. The excitation wavelength was 336 nm, and the pyrene emission was recorded in the wavelength range 350-600 nm. The spectra were accumulated with an integration of 3 s/nm. Light Scattering Measurements. Dynamic light scattering measurements were performed using a 90 Plus particle size analyzer (Brookhaven Instruments Corporation). All the measurements were carried out at 25 °C. The sample solutions were purified by passing through a Millipore 0.45 µm filter. The scattered light of a vertically polarized He-Ne laser (632.8 nm) was measured at an angle of 90° and was collected on an autocorrelator. The hydrodynamic diameters (d) of micelles were calculated by using the Stokes-Einstein equation d ) kBT/3πηD where kB is the Boltzmann constant, T is the absolute temperature, η is the solvent viscosity, and D is the diffusion coefficient. The polydispersity factor of micelles, represented as µ2/Γ2, where µ2 is the second cumulant of the decay function and Γ is the average characteristic line width, was calculated from the cumulant method.21 Constrained regularized continuous inversion (CONTIN) algorithms were used in the Laplace inversion of the intensity-intensity autocorrelation function from dynamic light scattering to obtain the micelle size distribution.22 Transmission Electron Microscopy. Transmission electron microscopy (TEM) was performed on a Philips CM 200, operating at an acceleration voltage of 200 kV. For the observation of size and distribution of micellar particles, a drop of sample solution (concentration ) 1 g/L) was placed onto a 200-mesh copper grid coated with carbon. About 2 min after deposition, the grid was tapped with a filter paper to remove surface water, followed by air-drying. Negative staining was performed by using a droplet of a 5 wt % uranyl acetate solution. The samples were air-dried before measurement. pH-Controlled Polymer-Mediated Assembly of Polymer Micelles and Reversible Micelle Release from the Hybrids. To prepare aqueous micellar solutions, the block copolymer in doubly distilled water was vigorously stirred at 70 °C for 3 h. Micellar solutions (Py-PCL-b-PAA28, Py-PCL-b-PAA45, and Py-PCL-b-MAA28, 10 mL, concentration ) 1 g/L) were mixed with each proton-accepting polymer (PEG, PEtOz, and PVP) at pH 7.4 to examine the effect of the polymer structure on the pH-induced micelle assembly through hydrogen bonding. The pH of the mixture of micellar solutions and polymers was then adjusted at specific pH in the range 2.0-7.4. The hybrid formation of the micelles with polymers at prefixed pH was examined by monitoring the change in pyrene fluorescence intensity (15) Harada, A.; Kataoka, K. Macromolecules 1998, 31, 288. (16) Wilhelm, M.; Zhao, C.; Wang, Y.; Xu, R.; Winnik, M. A.; Mura, J.; Riess, G.; Croucher, M. D. Macromolecules 1991, 24, 1033. (17) Kim, C.; Lee, S. C.; Kang, S. W.; Kwon, I. C.; Kim, Y.-H.; Jeong, S. Y. Langmuir 2000, 16, 4792. (18) Lee, S. C.; Kim, C.; Kwon, I. C.; Chung, H.; Jeong. S. Y. J. Controlled Release 2003, 89, 437-446. (19) Liu, S.; Weaver, J. V. M.; Save, M.; Armes, S. P. Langmuir 2002, 18, 8350-8357. (20) Fukushima, S.; Miyata, K.; Nishiyama, N.; Kanayama, N.; Yamasaki, Y.; Kataoka, K. J. Am. Chem. Soc. 2005, 127, 2810-2811. (21) Bronich, T. K.; Keifer, P. A.; Shlyakhtenko, L. S.; Kabanov, A. V. J. Am. Chem. Soc. 2005, 127, 8236-8237. (22) Allen, C.; Han, J.; Yu, Y.; Maysinger, D.; Eisenberg, A. J. Controlled Release 2000, 63, 275-286.
490 Langmuir, Vol. 23, No. 2, 2007 Scheme 1. Synthetic Route for Py-PCL-b-PAA or Py-PCL-b-PMAA Copolymers
of the aqueous mixture. For quantification of the micelles in the solid hybrid, the mixture at each pH was centrifuged at 3000 rpm for 40 min, and the supernatant was analyzed by fluorescence spectroscopy. The quantity of micelles in the solid hybrid was calculated by the ratio of pyrene fluorescence intensity (I) of the micelle/polymer mixture at each pH to the initial fluorescence intensity (Io) of the mixture prepared at pH 7.4. For the reversible release behavior of micelles from the hybrids, the micelle/polymer hybrid was obtained at pH 2, freeze-dried, and pressed as a disc with a diameter of 50 mm and a thickness of 1.4 mm at 50 MPa. The release behavior of the micelles from the hybrid was investigated in phosphate buffer solutions with pH values of 2.0 and 7.4. The concentrations of all the media were 100 mM with identical ionic strength. The disc-shaped hybrid was placed in 10 mL of release media. The sample was placed in a shaker bath maintained at 37 °C and shaken at 200 rpm. At a predetermined time interval, the whole volume of release media was withdrawn for assay and replaced by an equal volume of fresh media. The samples were assayed using a spectrofluorometer. In our assumption, the number of micelles is in a linear relationship with the concentration of block copolymer solutions above cmc, and the pyrene fluorescence intensity also has an identical relationship with the number of micelles in the solutions. The micelle release was quantified by the pyrene fluorescence. For quantitative analysis for the micelles release, the standard curves for polymer concentrations could be obtained by the calibration of pyrene fluorescence intensities in the micelle solutions of various block copolymer concentrations (0.0001-0.05 g/L). The concentration of the block copolymer solution (g/L) ) (intensity at 395 nm - 0.13)/3046, r2 ) 0.9999. The pH-dependent release behavior of the pyrene-loaded micelle from the hybrid was investigated by varying pH of the release media.
Results and Discussion Synthesis of Py-PCL-b-Poly(carboxylic acid)s. Scheme 1 illustrates the synthetic route for pyrene-labeled PCL-b-poly(acrylic acid) (Py-PCL-b-PAA) and PCL-b-poly(methacrylic acid) (Py-PCL-b-PMAA). Pyrene moieties were labeled at the terminal of the PCL block to provide polymer micelles with photophysical properties useful for assembly/disassembly assay in an aqueous phase. First, HO-PCL-Br was synthesized by the ring-opening polymerization of -caprolactone in the presence of HEBP as an initiator and Sn(Oct)2 as a catalyst. The number average molecular weight (Mn) of HO-PCL-Br was calculated by the peak integration ratio of CH in the 2-bromopropionate end group at 4.37 ppm to CH2 protons in the PCL repeating unit at 4.05 ppm. The Mn of HO-PCL-Br was estimated to be 1900 g/mol. The Mn of HO-PCL-Br macroinitiator estimated by GPC was 1950 g/mol, which is consistent with that measured by NMR. GPC analysis showed that the Mn of HO-PCL-Br macroinitiator was 2000 and had a narrow molecular weight distribution (Mw/ Mn ) 1.28). In the second step, the HO-PCL-Br macroinitiator
Lee and Lee
was used to produce block copolymers, HO-PCL-b-PtBA28, HOPCL-b-PtBA45, and HO-PCL-b-PtBMA28 by atom transfer radical polymerization (ATRP) of tert-butyl acrylate (tBA) or tert-butyl methacrylate (tBMA). ATRP was proceeded by heating a mixture of HO-PCL-Br and tBA or tBMA in the presence of an excess, relative to the initiating species, of equal amounts of Cu(I)Br and PMDETA at 85 °C. The block copolymers exhibited a narrow GPC trace with moderate polydispersity (Figure S-1 in Supporting Information and Table 1). This indicates that HO-PCL-Br was efficiently involved in initiating ATRP of tBA or tBMA, and there were no uncontrolled side reactions, particularly thermal polymerization. Pyrene labeling was successfully performed by DCC-mediated coupling reaction between the hydroxyl end group of the PCL block and the carboxyl group of pyrenebutyric acid. The percentage of pyrene labeling was calculated by the ratio of the resonance peak of PCL protons (t, J ) 6.6 Hz, CO(CH2)4CH2O) at 3.97 ppm to the peaks of pyrenyl protons at 7.9-8.3 ppm. The percent labeling of pyrene at the polymer termini was estimated to be 98%. The final product, Py-PCLb-poly(carboxylic acid)s, was obtained from the selective hydrolysis of the tert-butyl ester groups in copolymers by treatment with TFA. The resonance for the tert-butyl ester groups at 1.49 ppm completely disappeared after the hydrolysis (Figure S-2 in Supporting Information). The molecular weights, molecular weight distributions, and block compositions of the copolymer were determined by 1H NMR and GPC, and summarized in Table 1. The conversion of tBA28, tBA45, and tBMA28 to polymers was found to be 93%, 75%, and 70%, respectively. The molar composition ratios of PCL to PAA were 15:28 and 15:45, and for PCL to PMAA, the ratio was calculated as 15:28. The amphiphilic character of Py-PCL-b-poly(carboxylic acid)s gives an opportunity to self-associate in water to form a coreshell-type micelle structure. Thus, the polymer design in this work is useful for production of the well-defined polymer micelles with functionalized nanosurface with carboxylic groups of which protonation/ionization state can be controlled by pH. Micellization of Pyrene-Labeled Block Copolymers in Aqueous Media. The micelle formation of the block copolymer in an aqueous phase was confirmed by fluorescence spectroscopy and dynamic light scattering.23-25 The amphiphilic Py-PCLb-poly(carboxylic acid) copolymer is expected to self-associate in water to form polymer micelles of the hydrophobic PCL inner core and the hydrated poly(carboxylic acid) outer shell. Upon micellization in water, pyrene moieties labeled at the PCL termini preferably locate inside the hydrophobic microdomain of micelles, and thus consequently, they come into close contact to generate the excimers.26 Figure 1 shows the fluorescence emission spectra of Py-PCL-b-PMAA aqueous solutions at various concentrations. As the concentration of the block copolymer increases, the fluorescence intensity is gradually increased. It is noted that the fluorescence spectra of block copolymer solutions show the intrinsic pyrene fluorescence at 370-400 nm as well as the broad, featureless peaks at 400-550 nm corresponding to pyrene excitedstate dimers (excimers). This indicates that the block copolymer self-associated to form the polymer micelles with the inner core of the pyrene-labeled PCL and the outer shell of the anionic PMAA or PAA. The inserted plot shows the dependency of IE/IM (23) Kim, C.; Lee, S. C.; Shin, J. H.; Yoon, J.-S.: Kwon, I. C.; Jeong, S. Y. Macromolecules 2000, 33, 7448. (24) Kabanov, A. V.; Nazarova, I. R.; Astafieva, I. V.; Batrakova, E. V.; Alakhov, V. Y.; Yaroslavov, A. A.; Kabanov, V. A. Macromolecules 1995, 28, 2303. (25) Nagasaki, Y.; Okada, T.; Scholz, C.; Iijima, M.; Kato, M.; Kataoka, K. Macromolecules 1998, 31, 1473. (26) Jule, E.; Yamatomo, Y.; Thouvenin, M.; Nagasaki, Y.; Kataoka, K. J. Controlled Release 2004, 97, 407-419.
Assembly of Polymer Micelle Nanoparticles
Langmuir, Vol. 23, No. 2, 2007 491
Table 1. Characteristics of Py-PCL-b-Poly(carboxylic acid)sa feed ratio ([CL]:[tBA] or [CL]:[tBMA])
compositon ratiob ([CL]:[AA] or [CL]:[MAA])
conversionb
Mnc
Mw/Mnc
Py-PCL-b-PAA28
15:30
15:28
93%
3800
1.53
Py-PCL-b-PAA45
15:60
15:45
75%
5000
1.73
Py-PCL-b-PMAA28
15:40
15:28
70%
4200
1.60
copolymer
dd (µ2/Γ2)
de (µ2/Γ2)
51.5 nm (0.039) 29.9 nm (0.029) 36.3 nm (0.118)
53.6 nm (0.205) 31.4 nm (0.040) 41.2 nm (0.163)
a The copolymers were synthesized using the macroinitiator (HO-PCL-Br) with Mn of 1900 and polydispersity of 1.29, respectively. b Calculated by 1H NMR spectra. c Estimated by GPC. d Diameter of micelles before hybrid formation. e Diameter of micelles after redispersion from the hybrid at pH 7.4.
Figure 1. Emission spectra of Py-PCL-b-PMAA as a function of the concentration in aqueous media. The inserted plot represents the IE/IM ratios as a function of the concentration of aqueous block copolymer solutions.
(the intensity ratio of the excimer peak at 435 nm to the monomer peak at 374 nm) on the block copolymer concentration. The intensity ratio, IE/IM, is frequently used to provide information on the aggregation state and the aggregation number of the micelles. As shown in the inserted figure, IE/IM is almost zero below 1 mg/L, which is typically observed for fluorescence spectra of pyrene.26 On the other hand, IE/IM reached 0.67 at the critical concentration, reflecting the formation of excimers caused by micellization of block copolymers. Thus, the critical micelle concentration (cmc) of the block copolymers could be decided on the basis of the critical change in the IE/IM value. The cmc values of the block copolymers are almost in the similar range 0.5-1 mg/L, probably due to the fixed length of the hydrophobic PCL block. The influence of molecular weight distribution on micellar assembly was negligible. It is noteworthy that, above the critical concentration (1 mg/L), IE/IM is nearly constant over the wide range of concentrations. This suggests that the micelle structure is maintained over a wide range of concentration higher than cmc, and the increase in the concentration only leads to the increase of the micelle number. Thus, the number of the polymer micelles has a linear relationship with the polymer concentration. For analysis of polymer-mediated micelle assembly, this linear relationship is critical since it allows for the quantification of the numbers of assembled micelles in the hybrids by simple fluorescence analysis. As listed in Table 1, the hydrodynamic diameters of polymer micelles estimated by dynamic light scattering were in the range 36.3-51.5 nm at pH 7.4. For Py-PCL-b-PAA28, the micelle size at pH 7.4 was 51.5 nm, whereas it decreased to 35 nm at pH 3.0. In general, the micelle size decreased at acidic pH (below 4.0), which is probably due to the hydrogen bonding within the protonated micellar outer shell. This indicates that the structure of the hydrated corona of the micelles became more compact upon decreasing the pH. However, the narrow size distribution was maintained at a wide range of pH (pH 2.0-7.4) (data not
Figure 2. pH-Dependent phase behavior of the mixture of Py-PCLb-PAA28 micelles and PEtOz (a) and Py-PCL-b-PMAA28 micelles and PEtOz (b).
Figure 3. TEM images of micellar structure prepared from the aqueous mixture of PEtOz and Py-PCL-b-PMAA28 at pH 7.4 (a) and pH 4.5 (b). White arrows indicate the PEtOz-mediated assembled structure of micelles.
shown). This strongly supports that the micelles were stable without intermicellar aggregation even in the acidic pH. Polymer-Mediated Assembly of Polymer Micelles via pHControlled Complexation. The polymer-mediated micellar assembly based on complexation between Py-PCL-b-poly(carboxylic acid)s micelles and various types of proton-accepting polymers (PEG, PVP, and PEtOz) was investigated in an aqueous phase. Figure 2 shows pH-dependent phase behavior of the mixture of Py-PCL-b-PAA28 micelles and PEtOz, and Py-PCLb-PMAA28 micelles and PEtOz, respectively. Interestingly, two systems exhibit a distinct phase change by varying the solution pH. Upon decreasing pH from 7.4 to 2.0, the mixture of polymer micelles and PEtOz underwent a unique three-stage phase
492 Langmuir, Vol. 23, No. 2, 2007
Lee and Lee
Figure 4. pH-Dependent changes in the pyrene fluorescence intensity of the PEG/Py-PCL-b-PAA45 micelle mixture.
behavior at critical pH ranges: homogeneous micelle solutions, turbidity increase by polymer-mediated intermicellar assembly, and formation of solid hybrid assemblies. This unique pH-induced phase behavior was a totally reversible process. At neutral or basic pH, the polymer micelles and PEtOz existed separately, since most carboxyl groups of micellar shells are ionized, and thus, hydrogen-bonding interaction between PEtOz and micelles is negligible. As shown in Figure 2, depending on the micelle systems, the polymer-mediated micelle assembly begins around pH 3.5 or 4.5 to produce colloidal hybrids with enhanced turbidity. This assembly is caused by polymer-mediated intermicellar assembly at specific pH, where hydrogen bonding between hydrogen atoms in protonated carboxyl groups and oxygen and nitrogen in PEtOz becomes pronounced markedly. Interestingly, at critical pH, the white solid hybrid precipitates were formed due to the solubility loss of the polymer/micelle assembly in an aqueous media. It is of great interest to note that polymer micelles possessing PMAA shells underwent phase transition and formed the hybrid precipitates at higher pH than those with PAA shells. This phenomenon can be explained by considering the state of dissociation of each poly(carboxylic acid) as follows: Apparent dissociation constants pKa from the Henderson-Hasselbach equation are 7.3 for PMAA and 5.6 for PAA.27 These pKa values indicate that, at a certain pH, poly(carboxylic acid) with bigger pKa value makes more carboxylic groups protonated. Consequently, the micelles with PMAA shells could form hybrid with the proton-accepting polymer via hydrogen bonding at a higher pH than PAA did. Thus, it is confirmed that the type of anionic micellar shell is one design parameter that can control polymermediated micelle assembly by pH variation. Figure 3 shows the TEM images of the micelles obtained from the Py-PCL-b-PMAA28 micelles/PEtOz mixture at pH 7.4 (a) and pH 4.5 (b), respectively. At pH 7.4, individual, monodispese spherical micelles are observed, whereas the pH control from 7.4 to 4.5 led to the PEtOz-mediated intermicellar assembly to result in the coexistence of intermicellar aggregates with the larger particle size and single micelles, which was reflected in (27) Ikawa, T.; Abe, K.; Honda, K.; Tsuchida, E. J. Polym. Sci., Part A: Polym. Chem. 1975, 13, 1505. (28) Freichel, O. L.; Lippold, B. C. Int. J. Pharm. 2001, 216, 165-169.
turbidity increase in the solution mixture (Figure 2b). The further pH decrease facilitates the formation of larger aggregates and phase separation, thereby leading to the formation of solid hybrids. These nanoscale images strongly support the pH-dependent macroscopic phase behavior of the mixtures described in Figure 2. Figure 4 shows pH dependency of pyrene emission spectra of the mixtures of Py-PCL-b-PAA45/PEG obtained at various pH values. The mixture of the polymers and polymer micelles at various pH values was centrifuged, and the upper solutions were analyzed by pyrene fluorescence to quantify the complexed micelles in the solid hybrids. The pyrene fluorescence intensity is nearly constant in the pH range 7.4-3.0, indicating the stable micellar dispersion, but the pH decrease to 2.8 leads to the abrupt loss in the intensity, which reflects the initiation of the formation of the precipitated micelles/PEG hybrids. It is noteworthy that the pyrene intensity below pH 2.6 corresponds to about 1.5% of the initial intensity of the micelle solution, reflecting participation of almost all the micelles in hybrid formation. To provide detailed information on polymer-induced micellar assembly via pHinduced complexation, we examined all possible combinations of three polymer micelle systems and three proton-accepting polymers. Figure 5a shows the change of the intensity ratio (I/I0) of pyrene fluorescence (I) at various pH values to initial fluorescence of the micelles (I0) at pH 7.4. This plot is focused on the effect of the chemical structure and length of anionic micellar shells, while the proton-accepting polymer is fixed as PEtOz. For PyPCL-b-PAA28/PEtOz and Py-PCL-b-PAA45/PEtOz, the onset pH for assembly was found to be 5.0, whereas the critical pH for Py-PCL-b-PMAA28/PEtOz was 6.0. As discussed above, this difference can be ascribed to the higher pKa value of PMAA than PAA. However, there is little effect of the length of the micellar shell-forming poly(carboxylic acid) on the polymer-mediated micellar assembly. We note that the complex hybrids are obtained at a 41% yield at pH 4.5 for Py-PCL-b-PMAA28/PEtOz, and almost 100% of the micelles are complexed at pH 4.3 to form solid hybrids by polymer-mediated micelle assembly. On the other hand, for Py-PCL-b-PAA28/PEtOz and Py-PCL-b-PAA45/ PEtOz, the pH required for 100% recovery of micelles in solid
Assembly of Polymer Micelle Nanoparticles
Langmuir, Vol. 23, No. 2, 2007 493 Scheme 2. Postulated Mechanism of Reversible pH-Induced Association and Dissociation Behavior of Micelles with Polymers
Figure 5. (a) Changes in the intensity ratio (I/I0) of pyrene fluorescence (I) at various pH to initial fluorescence of the micelles (I0) at pH 7.4 for the complex hybrid formation of various types of polymer micelles and PEtOz. (b) Changes in the intensity ratio (I/I0) of pyrene fluorescence (I) at various pH to initial fluorescence of the micelles (I0) at pH 7.4 for the complex hybrid formation of Py-PCL-b-PMAA28 micelle and various types of proton-accepting polymers.
hybrids is lower than pH 2.7. This finding indicates that onset pH for assembly as well as pH for 100% assembly of micelles in solid hybrids can be controlled by the type of anionic polymer micelles. To gain an insight into the structural effect on assembly behavior, PEG, PVP, and PEtOz with different structures and hydrogen-bonding power with poly(carboxylic acid)s were utilized in inducing assembly of polymer micelles. The similar number of each repeating unit was considered for each polymer to minimize the effect of the polymer length. As illustrated in Scheme 2, the proton-accepting polymers are able to trigger the assembly of polymer micelles with ionizable poly(carboxylic acid)s by hydrogen bonding below critical pH ranges. PEG has one proton-accepting atom in the repeating unit, whereas PVP and PEtOz have two atoms possible for hydrogen bonding, nitrogen and carbonyl oxygen. Thus, it is likely that PEtOz and PVP are more effective in inducing polymer-mediated assembly of polymer micelles at higher pH ranges than PEG. It is also noted that the accessibility to nitrogen in PVP seems to be more limited than the nitrogen in PEtOz due to the five-membered ring structure of pyrrolidone. This intermolecular hydrogen bonding is well-reflected in the polymer effect on the pH-induced micellar assembly in Figure 5b. It is noteworthy that polymers began to induce the intermicellar assembly at critical pH in a descending pH order of PEtOz (pH ) 6.0), PVP (pH ) 5.0), and PEG (pH ) 4.0). It seems that, since PVP and PEtOz have two proton-accepting atoms in the repeating units, their hydrogen-bonding capacity is much stronger than that of PEG, thereby resulting in polymer/micelle hybrid formation at higher pH. As expected, the structure of PEtOz is more favorable for hydrogen bonding with poly(carboxylic acid)s
Table 2. Critical pH for the Complex Hybrid Formation
than PVP due to the steric freedom of hydrogen-accepting sites. Table 2 summarized data from a set of experiments giving detailed information on onset pH for micellar assembly as well as pH for 100% micelle hybridization. As listed in Table 2, each threshold pH is dependent on the pair of systematically designed polymer micelles and the proton-accepting polymers. In case we need the system that forms the polymer/micelle hybrids at the highest pH value, the best combination is Py-PCL-b-PMAA28 micelles and PEtOz, due to the higher pKa value of PMMA and high capacity for hydrogen bonding of PEtOz, respectively. In this system, micellar assembly by polymers was initiated at pH 6.0, and 100% of the micelles in solutions could be hybridized in the solid at pH 4.3. From pH-controlled assembly behavior, it can be suggested that the appropriate selection of a pair of the polymer micelle and the proton-accepting polymer can generate useful nanohybrid systems that can associate and dissociate at specific pH in a precisely controlled manner. Special emphasis is placed upon the observation that the micelle alone could not induce the long-range aggregation of the micelles, and the solid hybrid precipitation was not observed by simply lowering the pH of the micelle solution. Only in the case of using polymers such as PEG, PVP, and PEtOz did the polymer micelles form the solid hybrids at low pH through the polymer-mediated long-range micellar assembly. pH-Controlled Reversible Release of the Micelles from the Hybrids. The hybrid of polymer/micelles could be redispersed as individual micelles and polymers by pH increase as a result
494 Langmuir, Vol. 23, No. 2, 2007
of the breakage of hydrogen bonding between polymers and the outer shell of micelles. As listed in the Table 1, the redispersed solution of the hybrid at pH 7.4 contains the micelles of which the hydrodynamic diameters are comparable to those of micelles that are measured before hybrid formation. This indicates that the micelle structure is maintained during a pH-controlled assembly/disassembly process. The intermolecular interaction within hybrids is dependent on the power of the hydrogen bonding and the hydrophobicity of the systems. Thus, depending on the type of polymers and micelles utilized for hybrid formation, it is expected that the rate of micelle release from the hybrids can be controlled. Figure 6a shows the release profile of polymer micelles from the disc-shaped hybrid in buffer solutions with pH values of 2.0 and 7.4 at 37 °C. It is noted that the micelle release from the hybrid is completely prohibited at pH 2.0, due to the maintenance of the hydrogen bonding within the hybrid. It is noteworthy that the micelle release is triggered by placing the hybrids at pH 7.4. The micelle release by hybrid dissolution is controlled depending on the class of micelles. More specifically, the length and the hydrophobicity of the poly(carboxylic acid) shells are key factors to determine the dissolution rate of the hybrids. When the structure of the outer shell is fixed as PAA, the micelles with longer PAA shells may have the stronger interaction with proton-accepting polymers within the hybrids. Figure 6a clearly shows the slower micelle release from the hybrids of Py-PCL-b-PAA45 micelles than Py-PCL-b-PAA28 micelles. Besides, the micelle release may be controllable by the strength of hydrophobicity between micelles and poly(carboxylic acid)s. PMAA is comparatively more hydrophobic than PAA due to R-methyl groups in its main chain. Thus, it seems that the hybrid from the PMAA shell containing micelles and polymers is more cohesive and stable, compared to the hybrids of micelles with PAA shells and polymers. Interestingly, the micelle release from Py-PCL-b-PMAA28/PEG is much more retarded, compared with Py-PCL-b-PAA28/PEG and Py-PCL-b-PAA45/PEG. The hybrid released only about 20% micelles even after 7 h, whereas 100% micelles were released in around 1-2 h for the hybrids of Py-PCL-b-PAA/PEG. This suggests that the structure of shell-forming poly(carboxylic acid)s is one of the key parameters to determine the rate of reversible micelle release. Figure 6b shows the dependency of the micelle release kinetics on the type of proton-accepting polymers used for micellar assembly. For all the hybrid systems, the profile follows the zero-order release pattern. Interestingly, the micelle release from the Py-PCL-b-PAA28/PEG hybrid was found to be much faster than the Py-PCL-b-PAA28/PEtOz and Py-PCL-b-PAA28/PVP hybrids. The micelle release from the hybrid fabricated using PEG was completed in 40 min, whereas the micelle release was retarded by up to 100 min for hybrids formed by PEtOz and PVP. This is caused by the relatively weak hydrogen-bonding capacity of PEG, compared with PEtOz or PVP. PEtOz and PVP would generate much more cohesive hybrids due to the multiple hydrogen-bonding sites in the repeating units. Thus, the rate of reversible disassembly of micelles from the hybrids could be controlled depending on the design parameters such as the structure of micellar shell-forming poly(carboxylic acid)s and the class or type of polymers that induced the micellar assembly via pH variation. The polymer components, PEG, PEtOz, PVP, PCL, PAA, and PMAA, used for the hybrid formation in this work are widely used as biocompatible materials with low toxicity. In case drugloaded PCL-b-poly(carboxylic acid) micelles are used for the hybrid, this type of pH-controlled reversible assembly can provide
Lee and Lee
Figure 6. (a) pH-Dependent micelle release from hybrids estimated by pyrene fluorescence assay, and the effect of the structure and length of the micellar shell-forming polymer chains on micelle release (n ) 3). (b) The structural effect of proton-accepting polymers on the micelle release rate from hybrids (n ) 3).
a new platform methodology in the area of oral drug delivery. The main limitation for developing successful oral formulations is the rapid release of drugs unstable at an acidic pH during transit in the stomach (pH ≈ 2.0).28,29 The hybrid formed by pH-induced assembly of the micelles and polymers can provide a promising matrix that can release the drug-loaded micelles selectively at the pH of the small intestine (pH ≈ 7.0), thereby maximizing the oral bioavailability of drugs. The precise control of polymer-mediated micellar assembly, the key finding of this work, can offer useful oral formulation methods tailored for a specific drug. As an example, for acid-labile drugs, the assembly of Py-PCL-b-PMAA28 micelles and PEtOz is recognized as the most promising system, since their properties assembled at relatively high pH may guarantee the stability of drugs in resulting hybrids. Besides, the micelles released in the small intestine are expected to adhere to the intestinal wall to deliver drugs efficiently into the blood stream due to the strong mucoadhesive property of poly(carboxylic acid) micellar outer shells.
Conclusions pH-Controlled reversible assembly between Py-PCL-b-poly(carboxylic acid)s micelles and proton-accepting polymers was investigated in an aqueous medium. The assembly/disassembly behavior as well as phase transitions of the micelle/protonaccepting polymer could be conveniently and precisely controlled by adjusting pH. These behaviors depended on the rationally designed parameters such as the chemical structure of micellar shell-forming poly(carboxylic acid)s and the class of protonaccepting polymers. The relationship between design parameters and the polymer-mediated reversible micellar assembly presented in this study can be extended to the design for a variety of novel (29) Sakuma, S.; Hayashi, M.; Akashi, M. AdV. Drug DeliVery ReV. 2001, 47, 21-37.
Assembly of Polymer Micelle Nanoparticles
Langmuir, Vol. 23, No. 2, 2007 495
intelligent polymer/nanoparticle assembly systems useful for applications in the area of drug delivery and chemical sensing systems.
Technology Development Project” by Korea Ministry of Commerce, Industry and Energy.
Acknowledgment. This study was supported by a grant (code #10006921) from R&D Program for “Development of Sol-gel Hybrid Biomaterials and Its Application Technology” and a grant (code #10024816) from R&D Program for “International Joint
Supporting Information Available: Gel permeation chromatograms and NMR spectra. This material is available free of charge via the Internet at http://pubs.acs.org. LA0623580
