Article pubs.acs.org/Biomac
Controlled Thermoresponsive Hydrogels by Stereocomplexed PLAPEG-PLA Prepared via Hybrid Micelles of Pre-Mixed Copolymers with Different PEG Lengths Daniel G. Abebe and Tomoko Fujiwara* Department of Chemistry, The University of Memphis, 213 Smith Chemistry Building, Memphis, Tennessee 38152, United States ABSTRACT: The stereocomplexed hydrogels derived from the micelle mixture of two enantiomeric triblock copolymers, PLLA-PEG-PLLA and PDLA-PEG-PDLA, reported in 2001 exhibited sol-to-gel transition at approximately body temperature upon heating. However, the showed poor storage modulus (ca. 1000 Pa) determined their insufficiency as injectable implant biomaterials for many applications. In this study, the mechanical property of these hydrogels was significantly improved by the modifications of molecular weights and micelle structure. Co-micelles composed of block copolymers with two sizes of PEG block length were shown to possess unique and dissimilar properties from the micelles composed of single-sized block copolymers. The stereomixture of PLA-PEG-PLA comicelles showed a controllable sol-to-gel transition at a wide temperature range of 4 and 80 °C. The sol−gel phase diagram displays a linear relationship of temperature versus copolymer composition; hence, a transition at body temperature can be readily achieved by adjusting the mixed copolymer ratio. The resulting thermoresponsive hydrogels exhibit a storage modulus notably higher (ca. 6000 Pa) than that of previously reported hydrogels. As a physical network solely governed by self-reorganization of micelles, followed by stereocomplexation, this unique system offers practical, safe, and simple implantable biomaterials.
■
INTRODUCTION Thermoresponsive biodegradable hydrogels have been actively studied for drug delivery systems and temporary implants.1−4 In general, poly(lactide) (PLA) or poly(lactide-co-glycolide) (PLGA) is copolymerized with poly(ethylene glycol) (PEG) to procure amphiphilic micelles in aqueous solution, which transform into a hydrogel due to physical cross-linking influenced by swelling, hydration, and self-assembly of polymers responding to temperature. Kim and Lee developed injectable microparticles for DDS from various AB and BAB type block copolymers consisting of PLLA and PEG, or PLGA and PEG.5,6 Of those gels reported, the sol−gel phase diagrams are highly complicated, and thus numerous factors require adjustment to find the useful hydrogel at the temperature of interest. Sol−gel transition behavior and mechanical properties of hydrogels are particularly important for clinical application. Block length, block type,7 and the crystallinity8 of PLA-PEG copolymers were systematically studied by mixing ABA type and AB type copolymers or by mixing crystalline PLLA-PEG and amorphous PDLLA-PEG copolymer systems. Several techniques are used for photo-cross-linking hydrogels of PLAPEG block copolymer systems. As an example, methacrylatecapped triblock copolymer MA-PLLA-PEG-PLLA-MA and its’ degradation properties were studied, and as reported, the degradability of this high modulus gel can be controlled by tailoring the composition.9,10 © 2012 American Chemical Society
The use of stereocomplexation of PLLA and PDLA was previously applied to form hydrogels using p(HEMA)-graf toligo(lactide)11 and dextran-graf t-oligo(lactide) systems.12,13 In 2001, this stereocomplex mechanism was first introduced to prepare specifically the “thermosensitive hydrogels” from triblock copolymers, PLA-PEG-PLA by Fujiwara and Kimura.14 Aqueous micelle solutions from enantiomeric triblock copolymers with a particular molecular weight (MW), PLLA-PEGPLLA (1300−4600−1300) and PDLA-PEG-PDLA (1100− 4600−1100), exhibited sol-to-gel transition between 25 and 37 °C, which was considered to be a promising injectable hydrogel. The mechanism of this gelation was believed to occur as an interaction of micelles by exchange of PLA blocks in the micelle core and formation of stereocomplexation between PLLA and PDLA blocks. This PLA-PEG-PLA stereocomplex hydrogel, however, showed a relatively low storage modulus (ca. 1000 Pa) and thus was insufficient for many applications as an injectable implant material. There have been a few reports using similar materials for this hydrogel system. Vert et al. obtained hydrogels from increased MWs of PLA-PEG-PLA triblock copolymers but did not observe thermosensitive sol-to-gel transition.15 Furthermore, graft polymers16 and star-block copolymers of enantiomeric PLAs were synthesized17,18 and used with other chemical crossReceived: February 29, 2012 Revised: April 22, 2012 Published: April 26, 2012 1828
dx.doi.org/10.1021/bm300325v | Biomacromolecules 2012, 13, 1828−1836
Biomacromolecules
Article
CHOH, end group CH of the lactate repeating units), and δ 5.1−5.2 (q, CHCH3, of lactate repeating units). Micelle and Hydrogel Preparation. Figure 1 illustrates the micelle and hydrogel formation procedure. The micelle solutions of
linking for the gelation process to obtain more robust gels.19,20 Despite many studies, there was no report for mechanically strong and controllable thermoresponsive hydrogels driven solely by stereocomplexation of PLAs. Hydrogels from stereomixture of enantiomeric PLA-b-PEG are promising safe biomaterials without the use of other materials or chemical cross-linking; therefore, an improvement of physical properties and deployable gelation behaviors is crucial. Micelle formation mechanisms and the properties of hybrid micelles comprising more than two different block copolymers have been reported. As an example discussed by Munk, thermodynamic stability of hybrid micelles depends on many factors, such as core block sizes, total MW, medium, hydrophilic−lipophilic balance (HLB), and an interaction parameter of two blocks.21 Chaibundit studied a mixture of two different corona block sizes of Pluronic (PEG-PPG-PEG) to form comicelles and then hydrogels.22 In achievement of novel properties and functions, designing hybrid micelles from multiple components is an attractive approach. In this work, we developed the hybrid micelles formed from two different sizes of “premixed” PLA-PEG-PLA triblock copolymers with the aim to obtain a tunable sol-to-gel transition at a wide range of temperatures and superior mechanical properties of injectable hydrogels for practical use. The structures and properties of the aqueous micelles consisting of longer and shorter PEG blocks in the corona with a constant size of PLA in the core have been studied.
■
EXPERIMENTAL SECTION
Materials. The monomers, L-lactide and D-lactide with the trademarked names of PURASORB L and PURASORB D, respectively, were purchased from Purac Biochem (Netherlands). PEG with a number-average molecular weight (Mn) of 2000 (PEG2000) and 3350 Da (PEG-3350), toluene (extra dry), and tin(II) 2ethylhexanoate (SnOct2) was purchased from Sigma Aldrich (St. Louis, MO). The monomers, L-lactide and D-lactide were purified by recrystallization from dry toluene, then dried in a vacuum oven at room temperature for 48 h and stored at −20 °C under N2 environment before use. The macroinitiator PEG was lyophilized from benzene and stored at −20 °C under N2 atmosphere prior to use. All other reagents were used as received. Synthesis of Poly(L-lactide)-poly(ethylene glycol)-poly(lactide) Triblock Copolymers. The ABA-type triblock copolymers of PLLA-PEG-PLLA and PDLA-PEG-PDLA were prepared by the ordinary ring-opening polymerization (ROP) of L-lactide and D-lactide, respectively, in the presence of PEG-diol by the catalysis of Sn(Oct)2 (5 mol % relative to each hydroxyl group of PEG). The typical bulk polymerization was carried out as follows: PEG-2000 (5 g, 2.5 mmol) and L-lactide (4 g, 28 mmol) were placed in a reaction flask and equipped with a reflux condenser, a stir bar, and an addition funnel. The reaction flask was sealed and evacuated to remove moisture, then converted to N2 gas. After homogenizing the reactant at 100 °C for 1 h, the catalyst, Sn(Oct)2 (0.1 g. 0.25 mmol) at a 0.1 g/mL toluene solution was added to the reaction mixture. The reaction temperature was raised to 120 °C and allowed to react overnight. The reaction mixture was cooled to room temperature, and the mixture solidified to a cloudy solid at room temperature. The solid was dissolved in dichloromethane (50 mL) and added dropwise to 1:1 diethyl ether/ hexane (500 mL) to precipitate the product. The product was recovered by vacuum filtration and dried in vacuo at 60 °C for 24 h. The reprecipitation process was repeated more than twice to remove completely the catalyst, which was confirmed by 500 MHz NMR. The number-average molecular weight (Mn) and PLA/PEG block ratio were calculated from the 1H NMR spectra obtained in CDCl3. The chemical shifts are assigned as follows: δ 1.56−1.6 (d, CHCH3), δ 3.6−3.7 (m, OCH2CH2), δ 4.2−4.3 (m, COOCH2 for the oxymethylene connected to lactate repeating units), δ 4.3−4.4 (m,
Figure 1. Micelle and hydrogel preparation procedure for both homocopolymer and mixed copolymer micelles. PLLA-PEG-PLLA and PDLA-PEG-PDLA triblock copolymers were separately prepared. As depicted in the schematic (Figure 1a), the copolymer was dissolved in tetrahydrofuran (THF), a water miscible solvent. The transparent solution was added dropwise into deionized water with an ultrasonic wave applied at ∼4 °C. THF was then evaporated from the suspension under reduced pressure at 10 °C to acquire an aqueous micelle solution. Alternatively the “premixed copolymer micelle” preparation procedure is depicted in Figure 1b; the copolymers composed of the same PLA stereoisomer (e.g PLLA) but different PEG sizes (MW: 800−2000−800 + 800−3350−800) were combined and dissolved in THF to yield a homogeneous solution. This is the only difference in otherwise identical procedure, as described above. Micelle concentrations ranging from 0.01 to 20% were prepared for characterization and hydrogel studies. Controlled hydrogel formation is achieved by blending equal volumes of the PLLA-PEG-PLLA and PDLA-PEG-PDLA micelles. The micelle solutions of PLLA-PEG-PLLA and PDLA-PEG-PDLA were equilibrated at 4 °C using an ice bath initially. To an empty scintillation vial immersed in 4 °C ultrasonic bath, we added equal volume of each micelle solution, followed by gentle ultrasonic agitation. The mixture was allowed to sonicate at 4 °C for 30 min or until a homogeneous solution was obtained. The vial was then transferred to a temperature-controlled circulating water bath, in which the temperature was gradually ramped from 4 to 80 °C with a 1 h hold at each temperature interval. The vials were sealed with a screw cap to prevent water evaporation. The physical state of the mixture was reported at each temperature interval by tilting the vial. If the mixture flowed, then it was reported as a solution, and if it did not flow for at least 8 s, it was reported as a gel. The results were plotted as phase diagrams to display the sol−gel transition temperatures. Instrumentation. Nuclear magnetic resonance (1H NMR) was used to characterize all copolymers in CDCl3 at room temperature on a FT-NMR-Joel GSX-270 MHz. Gel permeation chromatography (GPC) was performed on a Shimadzu LC-20AD with two Jordi DVB 1829
dx.doi.org/10.1021/bm300325v | Biomacromolecules 2012, 13, 1828−1836
Biomacromolecules
Article
Table 1. Molecular Weight and Polydispersity of PLA-PEG-PLA Block Copolymers ID
block copolymer
(block MW)
theoretical MW(g/mol)
Mn (NMR)
Mn (GPC)
Mw (GPC)
PDI
L7E45L7 D7E45D7 L11E45L11 D11E45D11 L11E76L11 D11E76D11
PLLA-PEG-PLLA PDLA-PEG-PDLA PLLA-PEG-PLLA PDLA-PEG-PDLA PLLA-PEG-PLLA PDLA-PEG-PDLA
(500−2000−500) (500−2000−500) (800−2000−800) (800−2000−800) (800−3350−800) (800−3350−800)
3000 3000 3600 3600 4950 4950
2870 2850 3610 3620 4960 4950
2500 2470 3700 4020 5810 5880
2680 2680 4130 4650 6630 6550
1.07 1.08 1.11 1.15 1.14 1.11
Table 2. Thermal Properties of PLA-PEG-PLA Block Copolymers by DSC (2nd Scan)a ID (wt %) L7E45L7 L11E45L11 L11E76L11 L7E45L7 + D7E45D7 (50:50) L11E45L11 + D11E45D11 (50:50) L11E76L11 + D11E76D11 (50:50)
Tm (PEG) (°C) 37 33 45 35−50d 31 47
Tm (PLLA) (°C)
Tm (sc‑PLA) (°C)
ΔHm (PLLA) (J/g)
nd 132 156
0 0 0.8 0 0 0
c
nd nd 117 nd nd nd
ΔHm (sc‑PLA) (J/g)
Xc (PLLA) (%)b
Xc (sc‑PLA) (%)b
0 3.1 7.9
0 0 0.6 0 0 0
0 2.1 5.4
a All data shown are detected at the second scan from −50 to 190 (LEL) or to 240 °C (stereoblend) at the heating rate 10 °C/min. bCrystallinities of homo- and stereocomplex crystals were calculated using reported ΔHm for the crystals having an infinite thickness, ΔHm(100%) = 135 J/g for PLLA23 and 146 J/g for PLA stereocomplex.24 cnd: not detected dMultiple peaks observed
500 Å (250 × 10 mm) columns calibrated with polystyrene standards at 35 °C. The flow rate was 1.0 mL/min with THF as the mobile phase. Differential scanning calorimetry (DSC) was performed on a NETZSCH DSC 200 PC Phox. Polymers as synthesized were first heated to 190 (PLLA-PEG-PLLA) or 240 °C (1:1 blend of PLLAPEG-PLLA and PDLA-PEG-PDLA) at the rate of 20 °C/min, then cooled to −50 °C at the cooling rate of 100 °C/min and reheated (2nd scan) from −50 to 190 (PLLA-PEG-PLLA) or 240 °C (1:1 blend of PLLA-PEG-PLLA and PDLA-PEG-PDLA) at the rate of 10 °C/min. Dynamic light scattering (DLS) was used to measure a micelle size on a Malvern Zetasizer Nano ZS90 at 1, 0.1, and 0.01% concentrations at room temperature in triplicate. Wide-angle X-ray diffraction (WAXS) was performed on a Bruker AXS D8 advance Xray diffractometer for lyophilized powder samples. The instrument was operated at 40 kV and 40 mA and scanned in the 2θ range of 5−40°. Rheology measurement was performed on a TA AR 550 rheometer using a cone-and-plate geometry with a 4° cone angle, 40 mm diameter plate, and 61 mm gap. A frequency sweep was performed from 0.01 to 40 Hz at 37 °C. Synchrotron small-angle X-ray diffraction (SAXS) experiment was conducted at the 12-ID-C beamline at the Advanced Photon Source (APS) in the Argonne National Laboratory, Argonne, IL. The incident beam was monochromatized with Si (111) crystals. Data were collected with q range of 0.001 to 0.2 Å−1. The micelle solution sample was placed in a fluid cell with a path length of 1 mm and Kapton film window. All measurements were done at room temperature. The obtained SAXS data were corrected for the air/water scattering. Igor software (WaveMetrics) was used with Irena program provided by APS for unified fitting processes.
PEG blocks were designed. The triblock copolymers were obtained by the ROP of L- or D-lactide from PEG (MW: 2000 and 3350) macroinitiator at 120 °C in the presence of Sn(Oct)2. Table 1 summarizes the block copolymers used for this study, the number-average (Mn) and weight-average (Mw) molecular weights, and polydispersity index (PDI) were calculated from GPC (polystyrene conversion) and 1H NMR. Narrow polydispersity was obtained for all block copolymers. Thermal properties of synthesized copolymers were examined by DSC. Table 2 lists the melting temperatures and enthalpies of crystals formed via a second scan at the heating rate of 10 °C/min. From these short PLA blocks in the copolymers, the melting points (Tm) of the imperfect crystals were detected at much lower temperatures compared with the Tm of typical high MW homopolymers, PLLA (160−170 °C), and stereocomplex PLA crystal (220−230 °C). From the first scan of as-synthesized and vacuum-dried copolymers, for example, L11E45L11 and L11E76L11, a broad melting peak of PLLA was clearly detected at the range of 106−110 °C with crystallinities ca. 2−4% (data not shown). Given the short PLA blocks with a unit repeating number of 7 or 11, the crystallization process from the melt-quench state (second scan) was slow, and all copolymers displayed no or low PLA crystallinities. de Jong et al.25,26 prepared nonblended and equimolarly blended PLA oligomers and found that PLLA or PDLA were crystallizable for a repeating monomer unit number above 11, whereas an equimolar mixture formed stereocomplex crystallites for a unit number above 7, which means a high stability of the stereocomplex crystallites compared with that of homocrystallites. Pertaining to DSC data of the block copolymers, our studies showed consistent results as short PLLA copolymer (L7E45L7) formed no crystals and L11E45L11 crystallized very slowly, whereas stereocomplex formation was favored because crystallinity of blend polymers was higher than crystallinity of corresponding PLLA copolymer based on enthalpies of crystal melting calculated using theoretical values.23,24 Characterization of Micelles. Flower-type micelles with a spherical morphology are generally produced from PLA-PEGPLA block copolymers. The hydrophobic PLA segments
■
RESULTS AND DISCUSSION PLA-PEG-PLA Block Copolymers. On the basis of the MW of previously reported triblock copolymers, PLA-PEGPLA (e.g., 1100−4600−1100),14 several groups examined a series of larger sized copolymers.15 We have also prepared larger triblock copolymers using PEG (MW: 10000) as an initiator, for example, PLLA-PEG-PLLA and PDLA-PEGPDLA (1300−10000−1300, 2500−10000−2500, etc). Similarly to other attempts, a “temperature-responsive hydrogel” by self-organized PLA stereocomplexation mechanism was not observed, although gels were formed more easily by using a mixture of enantiomeric copolymer micelles rather than from single homopolymer micelles at same concentration. In this study, relatively low-MW copolymers with both short PLA and 1830
dx.doi.org/10.1021/bm300325v | Biomacromolecules 2012, 13, 1828−1836
Biomacromolecules
Article
associate in the micelle core surrounded by the hydrophilic PEG segments in an aqueous medium. The three types of micelle structures specifically focused on for this hydrogel study are illustrated in Figure 2. Among copolymers with constant
Table 3. Hydrodynamic Diameter of the Micelles Determined by DLS (0.1 wt %) micelle ID Mshort-L Mlong-L Mmix-L (10%) Mmix-L (20%) Mmix-L (30%) Mmix-L (40%) Mmix-L (50%) Mmix-L (60%) Mmix-L (70%) Mmix-L (80%) Mmix-L (90%)
Figure 2. Proposed micelle structures influenced by PEG block length.
length of PLA blocks (800), the shorter PEG copolymer (L11E45L11 or D11E45D11) and the longer PEG copolymer (L11E76L11 or D11E76D11) form different micelles, which are abbreviated as (Mshort) and (Mlong), respectively. The third micelle (Mmix) in Figure 2 was prepared by the premixture of shorter and longer PEG copolymers (i.e., L11E45L11+ L11E76L11 or D11E45D11+ D11E76D11). As Figure 2 depicts, the “premixed copolymer micelle” (Mmix) is considered to be coassociated by both sizes of copolymers. Co-micellization of two block copolymers has been studied by several researchers.21,22,27 Munk discussed thermodynamic stability of hybrid micelles with di- and triblock copolymers, which are dependent on various factors such as core block sizes, total MW, medium, HLB, interaction parameter of two block, and so on.21 Some combinations prevented hybridization, particularly at low core stability, whereas some formed uniform hybrid micelles. In such a situation, the solvent phase that is saturated by both unimers may become supersaturated with respect to the hybrid micelles, and their nucleation may possibly occur. Chang et al. used different corona blocks of copolymers with similar MW and block ratio on a silicon wafer. It is a clever experiment using AFM and pH change to demonstrate that hybrid micelles are formed by a premixed copolymer method, and postmixed micelles consist of the mixture of two different micelles.27 For (Mshort) and (Mlong) micelles, micelle size and stability are shown to be primarily influenced by the PEG block length. The hydrodynamic diameters of (Mshort-L) and (Mlong-L) by DLS gave 63 and 42 nm, respectively, as shown in Table 3. The (Mshort-L) micelle from PLLA-PEG-PLLA with the shorter PEG (800−2000−800) displayed a larger hydrodynamic size than the (Mlong-L) micelle from the copolymer with the longer PEG (800−3350−800). The increase in PEG block length leads to a smaller micelle, which means that a fewer number of polymers associate in the micelle core due to the large corona volume. The hydrodynamic diameters of (Mmix) micelles with a varied copolymer ratio measured by DLS are listed in Table 3. The sizes of “premixed copolymer micelles” were all similar but with some indication of aggregation. No specific trend was observed in micelle sizes; however, polydispersity indices of “mixed copolymer micelles” were slightly larger than those of single polymer micelles, (Mshort) and (Mlong). The similar average sizes and DLS curves obtained for the mixtures (Mmix) indicate comicellization of the two copolymers. Chaibundit et al. reported the micelle properties and gelation behavior of mixed copolymer micelles of Pluronic (PEG-PPG-PEG) with two different lengths of PEG blocks.22 Their mixed copolymer micelles with varied ratios have shown slightly larger hydro-
copolymers (weight ratio, %) L11E45L11 L11E76L11 (10%) L11E45L11 L11E76L11 (20%) L11E45L11 L11E76L11 (30%) L11E45L11 L11E76L11 (40%) L11E45L11 L11E76L11 (50%) L11E45L11 L11E76L11 (60%) L11E45L11 L11E76L11 (70%) L11E45L11 L11E76L11 (80%) L11E45L11 L11E76L11 (90%) L11E45L11 L11E76L11
size (nm)
polydispersity index
+ (90%)
63 43 81
0.112 0.106 0.193
+ (80%)
24, 116
0.298
+ (70%)
53
0.206
+ (60%)
51
0.150
+ (50%)
70
0.231
+ (40%)
66
0.205
+ (30%)
61
0.195
+ (20%)
75
0.219
+ (10%)
134
0.217
dynamic diameters and wider polydispersity than the single copolymer micelles, which is consistent with the results of our PLA-PEG-PLA system. The static light scattering study of their Pluronic copolymers revealed that the association number of “mixed copolymer micelles” linearly increased as the ratio of longer corona polymer decreased, whereas the micelle diameters for all mixed copolymer micelles were slightly larger than both of the single polymer micelles. They concluded that a stable “comicellization” occurred for the Pluronic system with constant-sized micelle core blocks. This is a similar system to our “pre-mixed copolymer micelles”, except these are BAB type copolymers (as A: hydrophobic core block). For our ABA type copolymers, the size of hydrophobic PLA is constant (800), and only the length of PEG is different (2000 and 3350). It is reasonable to form hybrid micelles with a stable PLA core and two sizes of PEG corona layer for premixed copolymers. The fourth micelle system investigated but not shown in Figure 2 is the “postmixed micelles” (Mshort + Mlong) that are prepared by simply mixing existing micelle solutions. The response of mixed micelles, however, was not as unique as that of the “premixed copolymer micelles” (M mix ). These “postmixed micelles” behaved similarly to either (Mshort) or (Mlong) depending on the major component. Thermoresponsive Hydrogel Formation. Hydrogels were prepared by mixing equal volumes of the enantiomeric micelle solutions at various polymer concentrations. As the sol−gel transition temperature is reached, the micellar solution changes from a solution to a gel state. This phenomenon is easily witnessed by tilling the vial. Stereocomplexed PLA-PEGPLA thermoresponsive hydrogels are considered to form by physical cross-linking induced by chain exchange between micelles, which is governed by temperature increase, as illustrated in Figure 3.14 These hydrogels are irreversible due to the robustness and insolubility of the PLA stereocomplex crystals formed inside micelle cores. The sol−gel transition temperatures for single copolymer micelle systems are summarized in phase diagram plots in Figure 4. The mixture of enantiomeric micelles composed of short PLA blocks, L7E45L7 (MW: 500−2000−500) and D7E45D7 (MW: 500−2000−500), did not form a hydrogel at 1831
dx.doi.org/10.1021/bm300325v | Biomacromolecules 2012, 13, 1828−1836
Biomacromolecules
Article
Figure 3. Previously proposed stereocomplexed hydrogel mechanism.
any given temperature or concentration, as shown in Figure 4a. It is known that the PLA block length must be greater than 7 repeating monomer units for successful formation of strereocomplexation between PLLA and PDLA.28 The block copolymers represented in Figure 4a have slightly below seven repeating lactate units; therefore, no gelation was observed. This result agrees with DSC data shown in Table 2, in which no stereocomplex formation has been detected from the polymer blend of L7E45L7 and D7E45D7. Figure 4b,c shows the phase diagrams for the enantiomeric mixture of the micelle solutions from constant PLA size (MW: 800) polymers with different PEG sizes, (Mshort-L + Mshort-D) and (Mlong-L + Mlong-D), respectively. The sol-to-gel transition temperature for the (Mshort-L + Mshort-D) micellar solutions is as low as 4 °C at polymer concentration 8% or higher. The micellar solutions of (Mlong-L + Mlong-D), however, showed the contrast response upon temperature treatment. These micelles were stable over the dehydration temperature of the PEG block, which is about 60−70 °C. The increase in PEG block length from MW of 2000 to 3350 had a significant effect on micelle stability. The stability of the micelles (Mlong) is because of their smaller hydrodynamic diameter and a larger PEG corona providing increased aqueous solubility. Consequently, an exchange of PLA blocks between micelles was suppressed by a relatively thicker corona layer. To be noted, the control experiments for all types of PLLA-PEG-PLLA block copolymer micelles (single enantiomer micelles) did not have transition into a gel state at any of the temperatures and concentrations shown in Figure 4, although some solutions were viscous at high concentrations. The stereomixture of the “premixed copolymer micelles” (Mmix-L and Mmix-D) dramatically exhibited different sol-to-gel phase transition behaviors. The sol-to-gel phase diagram in Figure 5 shows a linear response in sol−gel transition temperature as a factor of short-PEG and long-PEG block copolymer ratio. The concentration of the micelles was kept at
Figure 5. Sol−gel transition temperature of equimolar enantiomeric blends of the premixed copolymer micelles (Mmix-L + Mmix-D). The x axis shows the weight fraction of the short PEG copolymer (MW:800−2000−800) in the premixed copolymer micelle (MW:800−2000−800 + 800−3350−800). The micelle solutions are maintained at 10% concentration. Plots represent solution state (open squares) and gel state (filled triangles).
10% for this series. Most interestingly, a sol−gel transition around body temperature can be achieved by simply adjusting the copolymer ratio. These mixed copolymer micelles have physiological relevant thermoresponsive properties and can be used as injectable thermosensitive hydrogels. Again, the mixed copolymer micelle solutions from single enantiomer, Mmix-L, did not exhibit any hydrogel formation under the same given conditions. The SEM image of the resulting hydrogel (Mmix-L and Mmix-D) showed a porous network that is vastly interconnected (Figure 6, left). Using the tilt method, this
Figure 6. Hydrogel obtained by the mixture of enantiomeric PLAPEG-PLA from premixed copolymer (50:50% short/long PEG copolymers) micelle solutions (Mmix-L + Mmix-D). An SEM micrograph of the freeze-dried hydrogel (left) and a photo of the hydrogel in a vial immediately after heat treatment above body temperature (right).
Figure 4. Phase diagram of the 1:1 stereomixtures of micelle solutions prepared by different sized block copolymers by tilt method upon increase in temperature for each concentration. PLLA-PEG-PLLA and PDLA-PEG-PDLA micelle solutions: (a) MW: 500−2000−500, (b) MW: 800−2000− 800 (Mshort-L + Mshort-D), and (c) MW: 800−3350−800 (Mlong-L + Mlong-D). Plots represent solution state (open squares) and gel state (filled triangles). 1832
dx.doi.org/10.1021/bm300325v | Biomacromolecules 2012, 13, 1828−1836
Biomacromolecules
Article
mixed copolymer hydrogel did not flow at all immediately after the heat treatment of micelle solutions, which is a promising result for a robust injectable gel. The mechanism of sol−gel transition and mechanical strength of the resulting hydrogels is further discussed below. As a contrast experiment, the enantiomeric mixture from the postmixed micelles ((Mshort+Mlong)-L and (Mshort+Mlong)-D) was investigated using the same method. This stereomixture containing different types of micelles did not have any unique response at any given ratios and behaved similarly to the single sized micelle of the major copolymer. The resulting hydrogels were apparently weaker than “premixed copolymer micelle” gels by tilt observation. Therefore, the exceptional properties of the “premixed copolymer micelles” can be attributed to the unique micellar structures formed and not on the sole presence of both block copolymers. Chain exchange process between micelles has been studied by researchers. Previously, Wang et al. studied polystyrene-bpoly(ethylene oxide) micelle chain exchange kinetics experimentally and theoretically and raised the complexity of factors.29 Lund30,31 and more recently Bates32 reported micelle chain exchange of model block copolymers using time-resolved SANS technique. Although factors of chain-exchange kinetics have not reached a conclusion from such studies, it is believed that the micelle core plays a key role, such as distribution of core block lengths, interactions between core block segments and solvent, and core block mobility. The chain exchange kinetics of our system, Mshort, Mlong, and Mmix (premixed), is unknown presently and may not directly correspond to the sol−gel phase behavior. However, the difference of phase behavior between a stereoblend of premixed micelles and that of postmixed micelles will be based on kinetics and thermodynamics of micelle core dynamics, which requires further investigation. We believe that another significant effect is a stabilization of the core polymers after the chain exchange. The PEG blocks in the triblock copolymers act as bridges between micelles by chain exchange, and the formation of stable stereocomplex crystals is an essential driving force of this hydrogel. Crystallization of Hydrogels. As depicted in Figure 3, the typical stereomixture gelation for (Mshort) and (Mlong) micelles proceeds through the stereocomplexation between the PLLA and PDLA segments after chain exchange between micelles occurs.14 At some particular temperature, the polymer mobility in the micelle increases, and PLA blocks start to exchange between the micelles. PEG acts as a physical cross-linker to establish a hydrogel network. Although chain exchange occurs for PLLA-PEG-PLLA micelles, single micelle solutions do not exhibit sol-to-gel transition at any given concentrations and temperatures. As our DSC results have agreed with literatures, stereocomplex formation is more favorable than homocrystals formation in terms of limitation of chain length. Additionally, and as reported, the critical concentration and critical temperature of stereocomplex formation are lower than those of homocrystals. The crystallization of PLLA and PDLA has been studied using WAXS patterns. Enantiomeric PLLA and PDLA are both crystallizable in an orthorhombic or pseudo-orthorhombic unit cell with a 103 helical conformation (α-form). It is known that a racemic mixture forms the stereocomplex between the PLLA and PDLA by an alternate packing of β-form 31-helices of the opposite absolute configuration (left- and right-handed, respectively) side-by-side with van der Waals contact.33,34
The homopolymer crystal of PLLA or PDLA displays the XRD diffraction peaks at 2θ = 17 and 19°, whereas the stereocomplex crystal of the stereomixture shows the diffraction peaks at 2θ = 12 and 21°. Figure 7 shows the WAXS profiles of the freeze-
Figure 7. (a) WAXS profile for (Mshort-L) micelle (I) and (Mshort-L + Mshort-D) 50/50% mixture at 10 °C (II), 50 °C (III), and 80 °C (IV). (b) WAXS profile for (Mmix-L) micelle (V) and (Mmix-L + Mmix-D) 50/50% mixture at 10 °C (VI), 50 °C (VII), and 80 °C (VIII). (H) and (S) represent homopolymer crystal and stereocomplex crystal, respectively. Copolymer compositions in each micelle are Mshort-L: L11E45L11 (800−2000−800); Mshort-D: D11E45D11 (800−2000−800); Mmix-L: L11E45L11 (800−2000−800) + L11E76L11 (800−3350−800), 50/50 wt %; and Mmix-D: D11E45D11 (800−2000−800) + D11E76D11 (800−3350−800), 50/50 wt %.
dried micelles used in this study. The triblock copolymer hydrogels generated from (Mshort) and (Mlong) gave comparable diffraction peaks. The diffraction spectra in Figure 7a show the XRD profile for (Mshort) micelle and hydrogels. The diffraction spectra (I) for the (Mshort) micelle show PLLA homocrystal peaks at 2θ = 17 and 19° as well as crystalline PEG at 2θ = 19 (overlapped with PLLA peak) and 23°. The diffraction at 2θ = 19° is primarily from PEG; therefore, the peak at 2θ = 17° will be used to monitor PLLA or PDLA homocrystal changes. The diffraction spectra (Figure 7a, II−IV) showed stereocomplex peaks at 2θ = 12 and 21°. Both peaks are shown to increase as temperature increases, indicating a continued chain exchange, which leads to increased interaction between PLLA and PDLA segments. The increase in stereocomplex intensity exhibited sigmoid curve as a function of temperature, which was consistent with the previously reported result for the original hydrogel system by synchrotron WAXS.35 Above 80 °C, no further increase was seen in these systems. The peak at 2θ = 17° for the homocrystal of PLLA or PDLA blocks was shown to decrease as the stereocomplex crystal increased. It is proposed that the new micellar systems composed of mixed sized copolymers form hydrogels through the same 1833
dx.doi.org/10.1021/bm300325v | Biomacromolecules 2012, 13, 1828−1836
Biomacromolecules
Article
Figure 8. SAXS profile for (a) Mmix-L(70%) micelle solution and (b) (Mmix-L(70%) + Mmix-D(70%)) 50:50 wt % mixture at room temperature. Polymer concentrations are 10 wt % for both samples. Copolymer compositions in each micelle are Mmix-L(70%): L11E45L11 (800−2000−800) + L11E76L11 (800−3350−800), 70/30 wt %, and Mmix-D(70%): D11E45D11 (800−2000−800) + D11E76D11 (800−3350−800), 70/30 wt %.
systematic SAXS studies as well as other analysis techniques such as microscopies and surface labeling. Mechanical Properties of Hydrogels. Mechanical properties of these hydrogels were characterized by rheology. Figure 9 shows the storage modulus (G′) and the loss modulus
stereocomplexation mechanism, as described above. Figure 7b shows the XRD profiles for the mixed copolymer micelles and resulting hydrogel. The diffraction spectrum (V) for (Mmix-L) micelle (L11E45L11/L11E76L11 = 50/50) exhibited the PLLA homocrystal and crystalline PEG diffraction peaks equivalent to Figure 7a (I). The diffraction spectra (VI−VIII) for the racemic mixture (Mmix-L(50%) + Mmix-D(50%)) taken at varying temperatures showed stereocomplex crystal diffraction peaks at 2θ = 12 and 21°. Chain exchange between the (Mmix-L and Mmix-D) micelles was shown to occur at as low as 10 °C. However, as shown in the phase diagram in Figure 5, sol−gel transition for the racemic mixture of L11E45L11/L11E76L11 = 50/ 50 premixed copolymer micelles and corresponding D copolymer micelles was not observed until ∼40 °C. The stereocomplex peaks are shown to increase and the PLLA or PDLA homocrystal peaks are shown to decrease as temperature rises. In parallel to the (Mshort) hydrogels in Figure 7a, the chain exchange between the (Mmix-L) and (Mmix-D) micelle, followed by stereocomplexation, is a driving factor for cross-linking and hydrogel formation in the mixed copolymer systems. Nanostructures of Hydrogels. As hypothesized in illustration (Figure 3), we suggest that the micelle form retains after a successful chain exchange between micelles and stereocomplex formation in the cores. To examine structural information, we performed the SAXS measurement for the micelle solution and hydrogel samples using a synchrotron radiation. For the amphiphilic block copolymer micelles, SAXS data are expected to exhibit hydrophobic PLA core information because fully hydrated and low-electron-density PEG corona would not be detected. Figure 8 shows SAXS profiles of (a) 10 wt % of Mmix-L(70%) and (b) 10 wt % of (Mmix-L(70%) + Mmix-D(70%)) (50:50 racemic mixture), which are the solution and hydrogel states at room temperature, respectively. The experimental intensity plots are correlated to the equation q = 4π sin θ/λ, where q is the scattering vector, 2θ is the scattering angle, and λ is the wavelength of the X-ray. The radius of gyration (Rg) was estimated by Guinier and Porod fittings. The calculated core mean diameters for Mmix-L(70%) micelle and (Mmix-L(70%) + Mmix-D(70%)) micelle were 16 and 7 nm, respectively. Although the size of the stereocomplexed PLA core in the hydrogel is smaller than the core of the single enantiomer micelle, which is in a solution state for these particular set of samples, the distinctive peak in SAXS profile indicates that cross-linked the PLA core structure retains after the gelation process. To discuss particle sizes and the gelation mechanism of the hybrid micelles in detail, we will need further
Figure 9. Plots of storage modulus (G′) and loss modulus (G″) as a function of frequency at 37 °C for 1:1 stereomixture of micelle solutions: (closed diamond) G′ of (Mmix-L + Mmix-D); (open diamond) G″ of (Mmix-L + Mmix-D); (closed triangle) G′ of (MlongL + Mlong-D); (open triangle) G″ of (Mlong-L + Mlong-D); (closed square) G′ of (Mshort-L + Mshort-D); and (open square) G″ of (Mshort-L + Mshort-D). Concentration of micelles: Mmix: 20%, Mlong: 20%, and Mshort: 15%. Copolymer composition in each micelle is Mmix-L: L11E45L11 (800−2000−800) + L11E76L11 (800−3350−800), 50/50 wt %; Mmix-D: D11E45D11 (800−2000−800) + D11E76D11 (800−3350− 800), 50/50 wt %; Mlong-L: L11E76L11 (800−3350−800); Mlong-D: D11E76D11 (800−3350−800); Mshort-L: L11E45L11 (800−2000−800); and Mshort-D: D11E45D11 (800−2000−800).
(G″) of the stereocomplexed hydrogels as a function of frequency (Hz) at 37 °C. The racemic micelle mixture of the short PEG copolymers (Mshort-L + Mshort-D) has the lowest mechanical modulus (square). Both G′ and G″ are less frequency-dependent, which means that this hydrogel is stable even though it is the weakest. It is reasonable that the short PEG micelles turn to gel immediately after mixing of L and D under any condition higher than 8% concentration. The micellar blend of the longer PEG copolymers (Mlong-L + Mlong-D) holds modest mechanical strength (triangle). Initial higher values of G″ explain the fluid-like characteristic of this 1834
dx.doi.org/10.1021/bm300325v | Biomacromolecules 2012, 13, 1828−1836
Biomacromolecules
Article
20% turns to gel at a lower temperature than the 10% solution shown in the phase diagram in Figure 5. As seen in Figure 11,
hydrogel; however, upon further stereocomplexation, both moduli become stable. The racemic mixture of “premixed copolymer micelles” (Mmix-L + Mmix-D) exhibited the highest mechanical strength as the storage modulus immediately reached ∼6000 Pa (diamond). The higher value of G′ indicates that this “premixed copolymer micelle” hydrogel possesses a more solid-like, rigid microstructure. The loss modulus (G″) showed an unstable profile at relatively lower frequencies, which was reproducible for this particular hydrogel. This result indicates that chain exchange and structural reorganization occur in a different way from a single size copolymer hydrogel. The mechanical strength of this gel is a great improvement from the originally reported stereocomplexed PLA-PEG-PLA thermoresponsive hydrogel (ca. 1000 Pa).14 Alternatively, Figure 10 displays a comparison of the moduli (G′ and G″) of the racemic mixture of “premixed copolymer
Figure 11. Stress sweep modulus profiles of the “premixed copolymer micelle” hydrogel (Mmix-L + Mmix-D, 50:50 wt %) with 20% concentration measured at 25, 37, 42, and 50 °C.
there was no stress-hardening or weakening observed for this hydrogel. It is observable that the complete destruction of the gel structure (crossover of G′ and G′′ plots) occurred at oscillatory stress values of ∼100 Pa at 37 °C and higher than 100 Pa at room temperature. The rheology results in this study support the fact that the unique hybrid structure from “premixed copolymer micelles” offers additional factors on the gelation mechanism and the network structure toward a controlled thermoresponsive and robust hydrogel system. Theoretical analysis of the gelation mechanism, thermodynamics, and physical properties of the “premixed copolymer” micelles will be further investigated in future studies. Additionally, biodegradability and drug loading and release studies of new hydrogels from premixed copolymer hybrid micelles are underway.
Figure 10. Plots of storage modulus (G′) and loss modulus (G″) as a function of frequency at 37 °C for 1:1 stereomixture of micelle solutions: (closed diamond) G′ of (Mmix-L + Mmix-D); (open diamond) G″ of (Mmix-L + Mmix-D); (closed triangle) G′ of ((Mshort+Mlong)-L and (Mshort+Mlong)-D); and (open triangle) G″ of ((Mshort+Mlong)-L and (Mshort+Mlong)-D). Concentration of micelles: 20%. Polymer compositions in each micelle are Mmix-L: L11E45L11 (800−2000−800) + L11E76L11 (800−3350−800), 50/50 wt %; MmixD: D11E45D11 (800−2000−800) + D11E76D11 (800−3350−800), 50/ 50 wt %; (Mshort+Mlong)-L: L11E45L11 (800−2000−800) + L11E76L11 (800−3350−800), 50/50 wt %; and (Mshort+Mlong)-D: D11E45D11 (800−2000−800) + D11E76D11 (800−3350−800), 50/50 wt %.
■
CONCLUSIONS The physical and mechanical properties of PLA-PEG-PLA micelles and hydrogels were shown to change in response to variations of PEG block length and micellar structures. A unique micelle structure comprising two different sizes of hydrophilic corona blocks was achieved from PLA-PEG-PLA triblock copolymers. Hydrogels prepared from the racemic mixture of these hybrid micelles can be tailored to have sol-togel transition at desired temperature by simply adjusting the copolymer ratio and concentration. These mixed copolymer micelles possess physiologically relevant thermoresponsive properties and improved mechanical strength, which can be used as a promising injectable thermosensitive hydrogels in the biomedical field. As solely physical-network gels governed by self-reorganization of biodegradable micelles, this new system will be of great interest for practical use as biomaterials. Moreover, the hybrid micelles comprising two or more copolymers are important systems to study the new theory and mechanism of gel formation.
micelles” (Mmix-L + Mmix-D) (diamond) and “postmixed micelles” ((Mshort+Mlong)-L + (Mshort+Mlong)-D) (triangle). The postmixed micelle hydrogel consists of two different sizes of micelles (50:50 wt %) and showed similar rheological behavior to the (Mshort-L + Mshort-D) micelles shown in Figure 10, as both G′ and G″ have low values and are frequencyindependent. Although polymer composition of these two mixtures in Figure 10 is exactly same, the postmixture of different sized micelles does not show a high mechanical strength (ca. 2000 Pa) compared with the premixed copolymer micelles (diamond, ca. 6000 Pa). The difference of “premixed copolymer micelle” and “postmixed micelle” was also clearly seen for the sol−gel phase transition behavior (vide supra). Furthermore, the stress sweep of this “premixed copolymer micelle” gel (Mmix-L + Mmix-D) was examined at various temperatures to establish the linear viscoelastic region, which is a characteristic of a gel material. The racemic mixture (50:50 wt %) of “mixed copolymer micelle” solutions at concentration
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. 1835
dx.doi.org/10.1021/bm300325v | Biomacromolecules 2012, 13, 1828−1836
Biomacromolecules
Article
Notes
(26) de Jong, S. J.; van Dijk-Wolthuis, W. N. E.; Kettenes-van den Bosch, J. J.; Schuyl, P. J. W.; Hennink, W. E. Macromolecules 1998, 31, 6397−6402. (27) Chung, B.; Choi, H.; Park, H. W.; Ree, M.; Jung, J. C.; Zin, W. C.; Chang, T. Macromolecules 2008, 41, 1760−1765. (28) Tsuji, H. Macromol. Biosci. 2005, 5, 569−597. (29) Wang, Y. M.; Kausch, C. M.; Chun, M. S.; Quirk, R. P.; Mattice, W. L. Macromolecules 1995, 28, 904−911. (30) Lund, R.; Willner, L.; Richter, D.; Dormidontova, E. E. Macromolecules 2006, 39, 4566−4575. (31) Lund, R.; Willner, L.; Pipich, V.; Grillo, I.; Lindner, P.; Colmenero, J.; Richter, D. Macromolecules 2011, 44, 6145−6154. (32) Choi, S. H.; Bates, F. S.; Lodge, T. P. Macromolecules 2011, 44, 3594−3604. (33) Okihara, T.; Tsuji, M.; Kawaguchi, A.; Katayama, K.; Tsuji, H.; Hyon, S. H.; Ikada, Y. J. Macromol. Sci., Part B: Phys. 1991, B30, 119− 140. (34) Brizzolara, D.; Cantow, H. J.; Diederichs, K.; Keller, E.; Domb, A. J. Macromolecules 1996, 29, 191−197. (35) Fujiwara, T.; Yamaoka, T.; Kimura, Y. In Handbook on Bioapplications of Hydrogels; Ottenbrite, R., Ed.; Springer: New York, 2010; pp 157−177.
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was partially funded by FedEx Institute of Technology at the University of Memphis. T.F. thanks Prof. Sanjay Mishra in the Physics Department and Prof. William S. Janna in the Mechanical Engineering Department at the University of Memphis for equipment support on XRD and Rheology measurements, respectively. Special thanks to Prof. Robert Lamb and Dr. Alex Wu at the University of Melbourne in collaboration at Synchrotron SAXS experiment and for their guidance. T.F. acknowledges the APS user facility grant, the proposal ID: GUP-24921.
■
REFERENCES
(1) Lin, C. C.; Metters, A. T. Adv. Drug Delivery Rev. 2006, 58, 1379− 1408. (2) Hoare, T. R.; Kohane, D. S. Polymer 2008, 49, 1993−2007. (3) Van Tomme, S. R.; Storm, G.; Hennink, W. E. Int. J. Pharm. 2008, 355, 1−18. (4) He, C. L.; Kim, S. W.; Lee, D. S. J. Controlled Release 2008, 127, 189−207. (5) Jeong, B.; Bae, Y. H.; Lee, D. S.; Kim, S. W. Nature 1997, 388, 860−862. (6) Jeong, B.; Kim, S. W.; Bae, Y. H. Adv. Drug Delivery Rev. 2002, 54, 37−51. (7) Aamer, K. A.; Sardinha, H.; Bhatia, S. R.; Tew, G. N. Biomaterials 2004, 25, 1087−1093. (8) Agrawal, S. K.; Sanabria-DeLong, N.; Tew, G. N.; Bhatia, S. R. Macromolecules 2008, 41, 1774−1784. (9) Bryant, S. J.; Bender, R. J.; Durand, K. L.; Anseth, K. S. Biotechnol. Bioeng. 2004, 86, 747−755. (10) Rice, M. A.; Anseth, K. S. J. Biomed. Mater. Res., Part A 2004, 70A, 560−568. (11) Lim, D. W.; Choi, S. H.; Park, T. G. Macromol. Rapid Commun. 2000, 21, 464−471. (12) de Jong, S. J.; De Smedt, S. C.; Wahls, M. W. C.; Demeester, J.; Kettenes-van den Bosch, J. J.; Hennink, W. E. Macromolecules 2000, 33, 3680−3686. (13) de Jong, S. J.; De Smedt, S. C.; Demeester, J.; van Nostrum, C. F.; Kettenes-van den Bosch, J. J.; Hennink, W. E. J. Controlled Release 2001, 72, 47−56. (14) Fujiwara, T.; Mukose, T.; Yamaoka, T.; Yamane, H.; Sakurai, S.; Kimura, Y. Macromol. Biosci. 2001, 1, 204−208. (15) Li, S. M.; Vert, M. Macromolecules 2003, 36, 8008−8014. (16) Takami, K.; Watanabe, J.; Takai, M.; Ishihara, K. J. Biomater. Sci., Polym. Ed. 2011, 22, 77−89. (17) Nagahama, K.; Nishimura, Y.; Ohya, Y.; Ouchi, T. Polymer 2007, 48, 2649−2658. (18) Nagahama, K.; Ouchi, T.; Ohya, Y. Adv. Funct. Mater. 2008, 18, 1220−1231. (19) Hiemstra, C.; Zhong, Z. Y.; Li, L. B.; Dijkstra, P. J.; Jan, F. J. Biomacromolecules 2006, 7, 2790−2795. (20) Hiemstra, C.; Zhou, W.; Zhong, Z. Y.; Wouters, M.; Feijen, J. J. Am. Chem. Soc. 2007, 129, 9918−9926. (21) Tian, M. M.; Qin, A. W.; Ramireddy, C.; Webber, S. E.; Munk, P.; Tuzar, Z.; Prochazka, K. Langmuir 1993, 9, 1741−1748. (22) Chaibundit, C.; Ricardo, G.; Costa, V.; Yeates, S. G.; Booth, C. Langmuir 2007, 23, 9229−9236. (23) Miyata, T.; Masuko, T. Polymer 1998, 39, 5515−5521. (24) Tsuji, H.; Horii, F.; Nakagawa, M.; Ikada, Y.; Odani, H.; Kitamaru, R. Macromolecules 1992, 25, 4114−18. (25) de Jong, S. J.; van Nostrum, C. F.; Kroon-Batenburg, L. M. J.; Kettenes-van den Bosch, J. J.; Hennink, W. E. J. Appl. Polym. Sci. 2002, 86, 289−293. 1836
dx.doi.org/10.1021/bm300325v | Biomacromolecules 2012, 13, 1828−1836