Biodegradable Shape-Memory Polymers Exhibiting Sharp Thermal

May 8, 2009 - To date, various types of biodegradable SMPs based on poly(dl-lactide) and poly(ε-caprolactone) have been prepared for the development ...
0 downloads 0 Views 187KB Size
Download PDF
Biomacromolecules 2009, 10, 1789–1794

1789

Biodegradable Shape-Memory Polymers Exhibiting Sharp Thermal Transitions and Controlled Drug Release Koji Nagahama,† Yuichi Ueda,† Tatsuro Ouchi,† and Yuichi Ohya*,†,‡ Department of Chemistry and Materials Engineering, Faculty of Chemistry, Materials and Bioengineering, and High Technology Research Center, Kansai University, Suita, Osaka 564-8680, Japan Received February 18, 2009; Revised Manuscript Received April 13, 2009

Biodegradable shape-memory polymer networks prepared by cross-linking star shape branched oligo(ε-caprolactone) (bOCL) with hexamethylene diisocyanate are introduced. The thermal and mechanical properties of these networks were investigated using differential scanning calorimetry and tensile testing, respectively, and the morphology of the phase structure was characterized by polarized optical microscopy. The shape-memory properties of the networks were quantified using thermomechanical tensile experiments and showed strain fixity rates Rf higher than 97% and strain recovery rates Rr as high as 100%. Of note, networks of OCL segments with a lower degree of polymerization (DP; 10) exhibited significantly improved temperature-sensitive shape recovery: 90% of the permanent shape was recovered upon heating to within a 2 °C range (37-39 °C). The networks exhibited complete shape recovery to the permanent shape within 10 s at 42 °C. Theophylline-loaded (10 and 20 wt %) shapememory materials, prepared by cross-linking bOCL with hexamethylene diisocyanate in the presence of theophylline, are also described as a model for a controlled drug release device. The 10 wt % loaded material was sufficiently soft and flexible for complex shape transformation and also showed high Rf (98%) and Rr (99%). Sustained release of loaded theophylline was achieved over 1 month without initial burst-release in a phosphate buffer solution (PBS; pH 7.4) at 37 °C.

Introduction Shape-memory polymers (SMPs) are a class of smart materials with the ability to change shape on demand in response to an environmental stimulus.1,2 This shape-changing capacity has enabled SMPs to be used as smart medical devices3-7 and as sensors and actuators.8,9 The shape-memory effect is typically initiated by a change in temperature. Thermoresponsive SMPs generally consist of two components: cross-links determine the permanent shape and a thermally reversible phase fixing the temporary shape below the switching temperature (Ttrans).10 Although a variety of thermoresponsive SMPs have been developed, biodegradability of often required for those intended for use as biomedical materials in the body. One example of their application is in implant devices for minimally invasive surgery. Degradable implants can be inserted into the body in a compressed (temporary) shape through a small incision. When the device is placed in the correct position, it achieves its application (permanent) shape when its temperature reaches Ttrans. The device degrades after a defined time period, thus eliminating the need for a second surgery for removal. To date, various types of biodegradable SMPs based on poly(DL-lactide) and poly(ε-caprolactone) have been prepared for the development of smart implant devices.3,11-18 In addition to biodegradability, thermoresponsive SMPs used in implant devices during minimally invasive surgery should exhibit highly temperaturesensitive shape recovery within a narrow temperature range that is slightly above normal human body temperature (37 °C) to prevent migration of the device from its fine inserted position, and to minimize trauma to cells during deployment by treatment at excessively high temperatures. Although a few biodegradable * To whom correspondence should be addressed. Tel.: +81-6-6368-0818. Fax: +81-6-6330-4026. E-mail: [email protected]. † Department of Chemistry and Materials Engineering. ‡ High Technology Research Center.

SMPs with suitable Ttrans have been developed,14-17 they have required treatment at temperatures considerably higher than their corresponding Ttrans to affect complete shape recovery. In this study, we have developed novel biodegradable SMPs with strongly temperature-sensitive shape recovery within a 2 °C range (37-39 °C). The SMPs consist of starshape branched oligo(ε-caprolactone) (bOCL) and hexamethylene diisocyanate as the thermally reversible crystalline phase and cross-linker, respectively. The structural concept is based on polymer networks synthesized by coupling welldefined star-shape branched oligomers with low-molecular weight junctions. Poly(ε-caprolactone) is a biodegradable polyester with a relatively low Tm around 60 °C, and medical devices containing poly(ε-caprolactone) have been safely used in humans. Therefore, ε-caprolactone (ε-CL) was selected as the monomer for this system. Branched polylactides show a lower Tm than linear polylactides;19-24 therefore, a branched architecture was used in this poly(ε-caprolactone) system, and relatively short poly(ε-caprolactone) segments were employed as another means of lowering Tm. One of the polymer networks exhibits remarkably temperature-sensitive shape recovery from the temporary shape to permanent shape over a narrow temperature range (within a 2 °C range from 37 to 39 °C). To the best of our knowledge, such a sharp thermal transition could be significantly narrower than those of previously reported shape-memory polymers.13-15 We have also developed drug-eluting SMPs to develop controlled drug release devices for minimally invasive surgery systems. From this investigation, the preparation of bOCL networks, their thermal and mechanical properties, phase structure, shape-memory properties, degradation behavior, and patterns of controlled drug release are described.

10.1021/bm9002078 CCC: $40.75  2009 American Chemical Society Published on Web 05/08/2009

1790

Biomacromolecules, Vol. 10, No. 7, 2009

Nagahama et al.

Table 1. Characterization of bOCL Used in this Study sample

M/OHa

yield (%)

DPb

Mnc

Mw/Mnd

bOCL-10 bOCL-20

10 20

95 95

10 20

10300 18700

1.2 1.6

a Molar ratio of given ε-caprolactone to one hydroxyl group of oligoglycerin. b Degree of polymerization of one ε-caprolactone segment was an averaged value estimated by 1H NMR in CDCl3. c Mn calculated using Mn ) (Ir/It + 1) × 114 × 8 + 500, where Ir and It are the 1H NMR spectral intensities of the repeating and terminal methylene peaks of caprolactone units, respectively. The values 114, 8, and 500 are the molar mass of a caprolactone unit, the number of hydroxyl groups of oligoglycerin used, and the molecular weight of oligoglycerin, respectively. d Estimated by GPC (eluent: DMF).

Experimental Section Materials. Oligoglycerin (polyglycerine No. 500, anhydrous, Lot No. 55041, hydroxyl value ) 977) was supplied by Sakamoto Yakuhin Kogyo Co. Ltd. (Osaka). All other reagents were purchased from Wako Pure Chemical Co. (Tokyo) and used without further purification. Preparation of Polymer Networks. Oligoglycerin with an average of eight hydroxyl groups (Mw ) 500 g/mol; 200 mg, 0.40 mmol) was freeze-dried in a flask under vacuum for 12 h before use. ε-Caprolactone (ε-CL; 3648 mg, 32.0 mmol) and tin 2-ethylhexanoate (13 mg, 0.032 mmol) were added to the flask and the mixture was stirred at 120 °C for 12 h. The reaction mixture was then dissolved in chloroform and the resulting solution was precipitated in excess cold methanol. The precipitate obtained was collected by centrifugation and dried under vacuum overnight to yield branched oligo(ε-caprolactone) (bOCL) as a white powder. The degree of polymerization of ε-CL segments was calculated from the integral ratios of repeating methylene peaks (δ ) 3.65 ppm) and terminal methylene peaks (δ ) 4.07 ppm) of the caprolactone units (CDCl3, JEOL GSX-400). Based on these assignments, the number average molecular weights (Mn) of the bOCLs were calculated according to Mn ) (Ir/It + 1) × 114 × 8 + 500, where Ir and It are the 1H nuclear magnetic resonance (NMR) spectral intensities of the repeating and terminal methylene peaks, respectively, and 114, 8, and 500 are the molar mass of a caprolactone unit, the number of hydroxyl groups of oligoglycerin used, and the molecular weight of oligoglycerin, respectively. The molecular weight distribution (Mw/Mn) of bOCLs was determined by gel permeation chromatography (GPC; Tosoh GPC-8020, column: TSK-GEL R-5000 × 2, eluent: dimethyl formamide (DMF), detector: refractive index (RI), standard: poly(ethylene glycol) (PEG)). bOCL (900 mg, 0.68 mmol) was dissolved in CHCl3 (4 mL) and hexamethylene diisocyanate (HMDI; 57 mg, 0.34 mmol) was added to the stirred solution. The mixture was then poured into a Teflon dish (diameter 50 mm) and slowly dried at room temperature for 12 h, after which the cast film formed in the dish was heated to 80 °C and incubated for 24 h under nitrogen. The crude film obtained was washed thoroughly with CHCl3 and DMF to remove uncoupled HMDI and bOCLs, followed by drying overnight under vacuum at 100 °C to afford cross-linked bOCL (XbOCL) films (thickness: 500 ( 30 µm).

Characterization. The thermal properties of these films were analyzed using differential scanning calorimetry (DSC; Shimadzu DSC60, TA-60WS) under nitrogen gas from -50 to 100 °C at a heating rate of 10 °C/min. Fourier transform-infrared (FT-IR; Perkin-Elmer 1600) spectra of the polymer films were also recorded. The crystalline structure of the polymer films was analyzed using wide-angle X-ray diffractometry (WAXD; Rigaku RINT2000), and the optical properties were examined using polarized optical microscopy (POM; Olympus BHS-751P with crossed polarizers). Tensile testing (Shimadzu Autograph AGS-J) of the polymer films (dumbbell shape, 2 × 18 mm) was conducted using an elongation rate of 10 mm/min at 25 and 37 °C. Shape-Memory Property Analysis. The shape-memory properties of the XbOCL films were examined by thermocycling tests on a tensile machine (Shimadzu AG-2000E). The XbOCL films were cut into rectangular strips (original shape, 5 × 25 mm) and stretched to 100% elongation from the original shape using a stretching rate of 10 mm/ min at 60 °C. The strips were then cooled to 25 °C and incubated for 10 min under loading to fix the temporary shape. After removal of the load, the length of the samples was measured to calculate the strain fixity rate (Rf). The Rf values were calculated using Rf (%) ) lf,25/lst,60 × 100, where lst,60 and lf,25 are the respective lengths of the stretched samples at 60 °C and the temporary shape fixed at 25 °C after removal of the load. The samples with fixed temporary shape were immersed in a water bath at a certain temperature for 3 min to allow shape recovery (shrinking). After removal from the water bath, the lengths of the shrunken samples were measured to calculate the strain recovery rate (Rr). This procedure was performed at different temperatures between 30 and 55 °C. The Rr values were calculated using Rr (%) ) lor/lsh,t × 100, where lor and lsh,t are the lengths of the original shape and the recovered (shrunken) shape, respectively, at a given temperature. Biodegradation Test. The degradation behavior of the XbOCL films was investigated by periodically measuring the sample weight. XbOCL films (7 × 7 mm) were weighed (W0) after thorough drying and incubated in 4 mL of phosphate buffer solution (PBS; pH 7.4, I ) 0.14) at 37 °C for 120 days. After predetermined time periods, the films were washed with pure water and dried under vacuum. The films were then weighed (Wt) and the weight loss was calculated each time using weight loss (%) ) [(W0 - Wt)/W0] × 100. Preparation and Characterization of Theophylline-Loaded XbOCL Films. The following is a typical preparation procedure: bOCL (1000 mg, 0.056 mmol) and HMDI (71 mg, 0.42 mmol) were dissolved in CHCl3 (4 mL), and theophylline (111 mg, 0.62 mmol) was added to the stirred solution. After 30 min, the resulting homogeneous solution was poured into a Teflon dish (diameter 50 mm) and slowly dried at room temperature for 12 h, after which the cast film was heated to 80 °C and incubated for 24 h under nitrogen to afford a theophyllineloaded XbOCL film with a thickness of 500 ( 50 µm. The thermal and shape-memory properties of the obtained theophylline-loaded films were then analyzed. In Vitro Theophylline Release Test. The 10 wt % theophyllineloaded XbOCL films (5 × 10 mm) were placed in a dialysis tube

Scheme 1. Chemical Structure of Branched Oligo(ε-caprolactone) and the Polymer Network Structure Formed by Cross-Linking Reactions

Biodegradable Shape-Memory Polymers

Biomacromolecules, Vol. 10, No. 7, 2009

1791

Table 2. DSC Data and Shape-Memory Properties of bOCL and XbOCL Films sample

Tm (°C)

∆Hm (J/g)

Xca (%)

bOCL-10 XbOCL-10 bOCL-20 XbOCL-20

46.9 42.9 53.1 48.6

-62.6 -43.4 -65.2 -58.0

44.1 30.6 45.9 40.8

Rfb (%)

Rrb (%)

97.5 ( 0.6

100

99.1 ( 0.6

100

a

Figure 1. (A) DSC curves of (O) bOCL-10, (4) bOCL-20, (b) XbOCL10, and (2) XbOCL-20 films. These curves were recorded on the second heating run. (B) Wide-angle X-ray diffraction patterns of (a) bOCL-10, (b) bOCL-20, (c) XbOCL-10, and (d) XbOCL-20 films recorded at 25 °C.

Degree of crystallinity of OCL crystal formed in the films was calculated using Xc ) [∆Hm]/142 × 100, where 142 (J/g) is the melting enthalpy of a perfect PCL crystal. b The strain fixity rate (Rf) and the strain recovery rate (Rr,t) were calculated using Rf ) lf,25/lst,60 × 100 and Rr ) lor/lsh,t × 100. lf,25, length of the temporary shape fixed at 25 °C after removal of the load; lst,60, length of the stretched shape at 60 °C; lor, length of the original shape; lsh,t, length of the recovered (shrunken) shape at specific temperatures. Rr values shown in this table are the maximum value for each XbOCL film.

(MWCO: 1500) and into a vial and then immersed in 4 mL of PBS (pH 7.4, I ) 0.14) at 37 °C. At given time intervals, 20 µL of PBS (outside the tube) was withdrawn and the absorbance of the PBS at 274 nm was measured with a UV-vis spectrometer. After the addition of 20 µL of fresh PBS to the vial, incubation was continued at 37 °C. The amount of theophylline released from the films was determined using a calibration curve of previously prepared standard solutions.

Results and Discussion Preparation of Cross-Linked Branched OCL Films. Table 1 shows the characteristics of the bOCLs obtained. Ring-opening polymerization of ε-CL in the presence of oligoglycerin with eight hydroxyl groups as a macroinitiator, and tin 2-ethylhexanoate as a catalyst, afforded bOCL (Scheme 1). The DPs of ε-CL segments in bOCL, based on the integral ratios of the 1H NMR spectrum (Figure S1, Supporting Information) were calculated to be 10 and 20 for bOCL-10 and bOCL-20, respectively. Based on the GPC elution profile, it was confirmed that bOCL-10 and bOCL-20 were uncontaminated by unreacted oligoglycerin or ε-CL (Figure S2, Supporting Information). Both bOCL-10 and bOCL-20 precursors for the polymer networks were then reacted with certain amounts of hexamethylene diisocyanate (HMDI) to yield cross-linked bOCL-10 (XbOCL10) and bOCL-20 (XbOCL-20) films (Scheme 1). A 1:1 molar ratio of the bOCL hydroxyl end groups to isocyanate groups was used. Figure S3 (Supporting Information) shows FTIR spectra of bOCL and XbOCL films. After the cross-linking reaction, a new characteristic band corresponding to the N-H bending vibration of the urethane bonds appeared at 1520 cm-1 (red arrows) for both the XbOCL-10 and XbOCL-20 films, indicating the formation of a urethane network based on bOCL. The resultant films were white and opaque, as shown in Figure S4 (Supporting Information), and did not dissolve in typical organic solvents. Phase Structure of Cross-Linked OCL Films. The thermal properties of bOCL and XbOCL films were characterized using DSC. As shown in Figure 1A, the crystalline melting temperatures Tm were observed at 46.9 and 53.1 °C for bOCL-10 and bOCL-20, respectively, confirming the presence of crystalline phases in the films. Tm decreased with the decrease in the DP of the OCL segments. The bOCL-10 film also showed a lower fusion enthalpy ∆Hm and degree of crystallinity Xc compared to those of the bOCL-20 films, suggesting a lower quantity of crystalline phase in bOCL-10 (Table 2). Notable decreases in Tm, and in the corresponding ∆Hm and Xc values, were observed upon network formation. In particular, XbOCL-10 displayed a more marked decrease than XbOCL-20, indicating that network

Figure 2. Polarized optical microscopic images of (a) bOCL-10, (b) XbOCL-10, (c) bOCL-20, and (d) XbOCL-20 films recorded at 25 °C.

formation had a strong influence on the morphology of the crystalline phase, which was composed of short OCL segments. Figure 1B shows WAXD patterns of the bOCL and XbOCL films recorded at 25 °C. The diffraction peaks at 2θ ) 21.4 and 23.7° were attributed to OCL crystals,18 and these characteristic peaks were observed in all films, which indicates that there was no difference in the crystalline structure between bOCL and XbOCL. On the other hand, there were significant differences in the morphology of their phase structures, as observed by POM. Figure 2 shows POM images of the bOCL and XbOCL films at 25 °C, where the dark regions and birefringent domains indicate the isotropic amorphous phase and anisotropic crystalline phase, respectively. The bOCL-20 film has larger individual birefringent domains occupying a significantly larger area than the bOCL-10 film, which could be attributed to the larger crystal size in bOCL-20 due to its longer OCL segments. The XbOCL-20 film has a reduced area of birefringent domains compared to bOCL-20, which is caused by the decrease in crystallinity upon network formation. Interestingly, XbOCL-10 has well-defined continuous birefringent domains of macroscopic size in all areas of the film. The crystallinity of the XbOCL-10 film was 30.6%; therefore, the individual birefringent domains are expected to include some amorphous material. Although the birefringent domains probably consist of both amorphous and crystalline parts, most of the OCL chains are thought to be arranged in position via crosslinking, with the short length of the OCL chains in the XbOCL10 films being favorable for such arrangement. Mechanical Properties of Cross-Linked OCL Films. The mechanical properties of XbOCL films were investigated by tensile testing. Figure 3 shows stress-strain curves of the XbOCL-10 and XbOCL-20 films at 25 and 37 °C. Both films displayed distinct yield points at 25 °C and relatively high

1792

Biomacromolecules, Vol. 10, No. 7, 2009

Figure 3. Stress-strain curves of (a) XbOCL-10 film at 25 °C, (b) XbOCL-10 film at 37 °C, (c) XbOCL-20 film at 25 °C, and (d) XbOCL20 film at 37 °C.

Figure 4. (A) Thermally induced shape-memory behavior of XbOCL10 film. (a) Permanent spiral shape, (b) temporary rod shape fixed at 25 °C, and (c) the recovered shape after 10 s immersion in water at 42 °C. (B) Shape recovery ratio vs temperature for (b) XbOCL-10 and (O) XbOCL-20 films.

strengths (12 and 15 MPa for XbOCL-10 and XbOCL-20, respectively). When the samples were heated to 37 °C, the polymer films became relatively soft and exhibited larger elongations at break than at 25 °C. The XbOCL-10 and XbOCL20 films also displayed constant tensile strength during elongation until reaching strains of 240 and 290%, respectively. The soft and flexible character of these shape-memory polymers was deemed favorable for the transformation from the original shape to the desired temporary shape. Shape-Memory Properties of Cross-Linked OCL Films. The thermally induced shape-memory behavior of XbOCL films is shown in Figure 4A. The images in Figure 4A (a, b, and c) show the permanent shape of a XbOCL-10 film (2 × 120 mm and 510 µm thick), the temporary shape, and the recovered shape after heat treatment, respectively. Thus, a corkscrew shape was stretched into a straight rod at 60 °C and the deformation was fixed at 25 °C for 10 min. When placed in water at 42 °C, the shape was recovered, finally reforming to the original corkscrew shape in 10 s (Movie S1, Supporting Information). The XbOCL20 film also exhibited typical shape-memory behavior. The shape-memory properties of the XbOCL films were quantified by thermomechanical tensile experiments. Rf and Rr were calculated to quantify the fixation of the temporary shape and the recovery to permanent shape of the samples, as listed

Nagahama et al.

in Table 2. Rf values of both samples were higher than 97% when the OCL segment content (forming the triggering phase of the networks) was relatively high. The final Rf values of both samples reached values as high as 100%. Both samples had the same Rr value (100%) throughout five consecutive cyclic tests. Figure 4B shows the strain recovery curves of the XbOCL-10 and XbOCL-20 films as a function of temperature. Both the measured curves were S-shaped, and Ttrans for XbOCL-10 and XbOCL-20 were around 38 and 50 °C, respectively, with Ttrans decreasing with decrease in the OCL chain length. Ttrans could be reduced to approximate body temperature by the use of short OCL chains. The XbOCL-10 film displayed strongly temperature-sensitive shape memory, recovering 90% of its permanent shape upon heating over a 2 °C range (37-39 °C). To the best of our knowledge, the stimulus-responsive behavior of the XbOCL-10 film is significantly better than that of any previously reported SMP.3,13-15,25,26 The XbOCL-20 film displayed complete shape recovery when heated over a 6 °C range (47-53 °C), while displaying a still more pronounced response. Ttrans of thermosensitive SMPs is typically governed by reversible phase transition temperatures such as Tm or Tg of the triggering segments. The Ttrans of XbOCL-20 corresponds well with its Tm (49 °C). In contrast, Ttrans of XbOCL-10 was significantly lower than its Tm (43 °C), which indicates a different process of shape recovery from that of XbOCL-20. POM observations were carried out during shape recovery of the XbOCL-10 and XbOCL-20 films because these films exhibited quite different birefringence morphologies at 25 °C. XbOCL-20 has microscopic, mostly crystalline, birefringent domains, as shown in Figure 2d. When the sample was heated to 40 at 10 °C/min, the number of birefringent domains clearly decreased, and these domains eventually changed, one by one, to dark domains at 55 °C (data not shown). This thermal phase transition occurred in essentially the same temperature range where an endothermic peak was observed in the DSC thermogram, indicating gradual melting of the crystalline phase. Thus, the melting is revealed to be a trigger that induces shape recovery in XbOCL-20. In contrast, no decrease in the number of birefringent domains was observed for XbOCL-10, even when the temperature was increased to 32 °C, where an endothermic peak was observed for XbOCL-10. Interestingly, XbOCL-10 showed cooperative and macroscopic transitions, with the birefringent domains persisting despite the increased temperature (Figure S5, Supporting Information). When the temperature was raised to 43 °C, some parts of the anisotropic birefringent domains began to change into isotropic domains. This temperature was the almost the same as that when the shape recovery rate reached 100%, which indicates that macroscopic transitions of the continuous anisotropic domains triggered the rapid shape recovery of XbOCL-10 within the 2 °C range. The significantly sharp thermal transition of XbOCL-10 induced near body temperature could be an important advantage for implant systems in minimally invasive surgery. These features of the material would (a) prevent implant migration from the correct preinserted position and (b) minimize trauma to cells during in vivo deployment. Degradation of Cross-Linked OCL Films. XbOCL films are expected to gradually degrade by hydrolysis of the ester bonds in the OCL segments, forming relatively lower molecular weight OCL that would be converted to water-soluble degradation products as hydrolysis progresses. The rate of degradation was influenced by the degree of crystallinity of XbOCL. The degradation behavior of XbOCL-10 and XbOCL-20 films was investigated by measuring their weights after incubation in PBS

Biodegradable Shape-Memory Polymers

Figure 5. Time course curves of weight remaining for (b) XbOCL-10 and (O) XbOCL-20 films in PBS at 37 °C.

at pH 7.4 and 37 °C, as shown in Figure 5. As expected, XbOCL-10 degraded faster than XbOCL-20. The XbOCL-10 and XbOCL-20 films retained 96 and 99% of their original weight, respectively, at the end of the 120 day test period. After the degradation test for 120 days, it was confirmed that both films retained their shape-memory properties. Shape-Memory Properties of Theophylline-Loaded CrossLinked OCL Films. Based on the favorable shape-memory properties of the XbOCL films for implant devices, an attempt was made to encapsulate drug-like molecules into the films, because biodegradable SMPs with sustained drug release properties could enable the development of novel types of medical devices and therapies. Theophylline we selected as a model drug. Theophylline-loaded XbOCL films (500 ( 50 µm thick) were prepared by cross-linking bOCL with HMDI in the presence of certain amounts (10 and 20 wt %) of theophylline. Figure S6 (Supporting Information) shows typical SEM cross sectional images of an XbOCL-10 film with 10 wt % theophylline; a phase-separated structure composed of OCL and theophylline phases was observed. Although chloroform used in cross-linking was a poor solvent for theophylline, theophylline domains were well-dispersed in the XbOCL-10 films. The thermal properties of theophylline-loaded XbOCL were characterized by DSC, as listed in Table S1 (Supporting Information). The loaded XbOCL films showed slightly higher Tm, ∆Hm, and Xc values than those of the parent XbOCL films; therefore, crystallization of the OCL segments may have been facilitated by the presence of theophylline molecules. The shape-memory properties of theophylline-loaded XbOCL films were investigated by the same protocol used for XbOCL films, and the results are summarized in Table S1 (Supporting Information). Shape-memory behavior was observed for all of the theophylline-loaded XbOCL films, but the degree of elongation of the 20 wt % theophylline-loaded XbOCL films was poor and the films completely broke at elongation below 100%. In contrast, the 10 wt % theophylline-loaded XbOCL films were soft and sufficiently flexible for shape transformation, and also exhibited high Rf (98%) and Rr (99%), only slight less than those of the corresponding XbOCL films. Figure S7 (Supporting Information) shows time-course photographs of shape recovery for a 10 wt % theophylline-loaded XbOCL-10 film at 45 °C. In this test, the permanent shape was a corkscrew and the temporary shape was a straight rod (2 × 120 mm, 520 µm thick). The shape recovery time from temporary to permanent shape was expected to be strongly dependent on the size, shape, and thickness of the sample. The corkscrew was transformed into a rod at 60 °C, and the resulting temporary shape was fixed by cooling to 25 °C. The complex transformation from rod back to corkscrew took approximately 300 s.

Biomacromolecules, Vol. 10, No. 7, 2009

1793

Figure 6. In vitro release profile of theophylline from (b) XbOCL-10 and (O) XbOCL-20 films in PBS at 37 °C.

In Vitro Theophylline Release Behavior. The release of theophylline from the films was measured in PBS at 37 °C, and the cumulative release of theophylline is shown in Figure 6. Both films exhibited slow sustained release of theophylline over one month, and importantly, no initial burst-release was observed. Although it was anticipated that XbOCL-20 would have larger pores in its polymer network than XbOCL-10, based on the longer OCL chain length of the former, similar release curves were observed for both films. This indicated that release was mainly depended on the diffusion of entrapped theophylline, due to the defined phase-separated structure of the films. Both films showed shape-memory behavior after the release testing for 35 days. The theophylline-loaded (20%) XbOCL films did not exhibit sufficient flexibility for shape-memory. However, we should know the marginal value of drug loading (wt %) to maintain the shape-memory effects. We think that the results are important with respect to providing information regarding the drug loading capacity. In this study, theophylline was used as an example of a conventional model drug. Other drugs with higher activity may be used in smaller amount.

Conclusions Novel biodegradable shape-memory materials based on chemically cross-linked polymeric networks of branched oligo(ε-caprolactone) were developed. The shape-memory materials exhibited strain fixity rates higher than 97% and strain recovery rates approaching 100%. The shape-memory materials containing oligo(ε-caprolactone) segments with lower DP (10) showed significantly pronounced temperature-sensitive shape recovery, regaining 90% of their permanent shape when heated to within a 2 °C range (37-39 °C). In addition, theophylline-loaded (10 and 20 wt %) shape-memory materials were prepared by crosslinking branched oligo(ε-caprolactone) with hexamethylene diisocyanate in the presence of specific amounts of theophylline. The 10 wt % theophylline-loaded material was soft and sufficiently flexible for shape transformation, and displayed high strain fixity (98%) and strain recovery (99%) rates. Slow sustained release of loaded theophylline was achieved over one month without initial burst-release in PBS at 37 °C. These results should be relevant and important for the developments of minimally invasive surgery systems using biodegradable SMPs. Acknowledgment. This work was carried out as a study in the High-Tech Research Center Project supported by the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan. A part of this work was financially supported by a Grant-in-Aid for Scientific Research on Priority Areas

1794

Biomacromolecules, Vol. 10, No. 7, 2009

(20034055) from MEXT. The authors thank Sakamoto Yakuhin Kogyo Co. Ltd. for supplying oligoglycerin. Supporting Information Available. 1H NMR spectra, GPC charts, FTIR spectra, POM images, SEM image, and time course photographs and a movie of shape recovery. This material is available free of charge via the Internet at http://pubs.acs.org.

References and Notes (1) Wei, Z. G.; Sandstrom, R.; Miyazaki, S. J. Mater. Sci. 1998, 33, 3743– 3762. (2) El Feninat, F.; Laroche, G.; Fiset, M.; Mantovani, D. AdV. Eng. Mater. 2002, 4, 91–104. (3) Lendlein, A.; Langer, R. Science 2002, 296, 1673–1675. (4) Metcalfe, A.; Desfaits, A. C.; Salazkin, I.; Yahia, L.; Sokolowski, W. M.; Raymond, J. Biomaterials 2003, 24, 491–497. (5) Lendlein, A.; Kelch, S. Clin. Hemorheol. Microcirc. 2005, 32, 105– 116. (6) Gall, K.; Yakacki, C. M.; Liu, Y.; Shandas, R.; Willett, N.; Anseth, K. S. J. Biomed. Mater. Res. 2005, 73, 339–348. (7) Yakacki, C. M.; Shandas, R.; Lanning, C.; Rech, B.; Eckstein, A.; Gall, K. Biomaterials 2007, 28, 2255–2263. (8) Metzger, M. F.; Wilson, T. S.; Schumann, D.; Matthews, D. L.; Maitland, D. J. Biomed. MicrodeVices 2002, 4, 89–96. (9) Small, W.; Wilson, T. S.; Benett, W. J.; Loge, J. M.; Maitlad, D. J. Opt. Express 2005, 13, 8204–8213. (10) Lendlein, A.; Kelch, S. Angew. Chem., Int. Ed. 2002, 41, 2034–2057.

Nagahama et al. (11) Alteheld, A.; Feng, Y.; Kelch, S.; Lendlein, A. Angew. Chem., Int. Ed. 2005, 44, 1188–1192. (12) Lendlein, A.; Schmidt, A. M.; Langer, R. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 842–847. (13) Wang, W. S.; Ping, P.; Chen, X. S.; Jing, X. B. Eur. Polym. J. 2006, 42, 1240–1249. (14) Zini, E.; Scandola, M.; Dobrzynski, P.; Kasperczyk, J.; Bero, M. Biomacromolecules 2007, 8, 3661–3667. (15) Cao, Q.; Liu, P. Polym. Bull. 2006, 57, 889–899. (16) Venkatraman, S. S.; Tan, L. P.; Joso, J. F. D.; Boey, Y. C. F.; Wang, X. Biomaterials 2006, 27, 1573–1578. (17) Ping, P.; Wang, W.; Chen, X.; Jing, X. Biomacromolecules 2005, 6, 587–592. (18) Luo, H.; Liu, Y.; Yu, Z.; Zhang, S.; Li, B. Biomacromolecules 2008, 9, 2573–2577. (19) Finne, A.; Albertsson, A.-C. Biomacromolecules 2002, 3, 684–690. (20) Numata, K.; Srivastava, R. K.; Finne, A.; Albertsson, A.-C.; Doi, Y.; Abe, H. Biomacromolecules 2007, 8, 3115–3125. (21) Li, Y.; Kissel, T. Polymer 1998, 39, 4421–4427. (22) Choi, Y. K.; Bea, Y. H.; Kim, S. W. Macromolecules 1998, 31, 8766– 8774. (23) Korhonen, H.; Helminen, A.; Seppa¨la¨, J. V. Polymer 2001, 42, 7541– 7549. (24) Tasaka, F.; Ohya, Y.; Ouchi, T. Macromolecules 2001, 34, 5494– 5500. (25) Liu, G. L.; Ding, X.; Cao, Y.; Zheng, Z.; Peng, Y. Macromolecules 2004, 37, 2228–2232. (26) Choi, N.; Lendlein, A. Soft Matter 2007, 3, 901–909.

BM9002078