Biomacromolecules 2002, 3, 601-608
601
Resorbable and Highly Elastic Block Copolymers from 1,5-Dioxepan-2-one and L-Lactide with Controlled Tensile Properties and Hydrophilicity Maria Ryner and Ann-Christine Albertsson* Department of Polymer Technology, Royal Institute of Technology, SE-100 44 Stockholm, Sweden Received December 10, 2001; Revised Manuscript Received March 11, 2002
New resorbable and elastomeric ABA tri- and multiblock copolymers have been successfully synthesized by combining ring-opening polymerization with ring-opening polycondensation. Five different poly(L-lactideb-1,5-dioxepan-2-one-b-L-lactide) triblock copolymers and one new poly(L-lactide-b-1,5-dioxepan-2-one) multiblock copolymer have been synthesized. The triblock copolymers were obtained by ring-opening polymerization of 1,5-dioxepan-2-one (DXO) and L-lactide (LLA) with a cyclic tin initiator. The new multiblock copolymer was prepared by ring-opening polycondensation of a low molecular weight triblock copolymer with succinyl chloride. The molecular weight and the composition of the final copolymers were easily controlled by adjusting the monomer feed ratio, and all of the polymers obtained had a narrow molecular weight distribution. It was possible to tailor the hydrophilicity of the materials by changing the DXO content. Copolymers with a high DXO content had a more hydrophilic surface than those with a low DXO content. The receding contact angle varied from 27 to 44°. The tensile properties of the copolymers were controlled by altering the PDXO block length. The tensile testing showed that all the polymers were very elastic and had very high elongations-at-break (b). The copolymers retained very good mechanical properties (b ≈ 600-800% and σb ≈ 8-20 MPa) throughout the in vitro degradation study (59 days). Introduction There is still a great need for improved properties and a better understanding of biomaterials, despite the numerous investigations and various synthetic pathways for new materials that have been achieved during the past decades. Aliphatic polyesters like polylactide,1-3 polyglycolide,4,5 and poly(-caprolactone)6,7 are the most extensively investigated polymers, essentially because of their good hydrolyzability and biocompatibility. Their homopolymers are however all semicrystalline, and hence, they offer only a limited range in tensile properties. Completely amorphous polymers like poly(1,5-dioxepan-2-one) (PDXO)8-10 are therefore very valuable for copolymerization to improve the elasticity of a polymer. In the past few years, our group has been focusing on tin alkoxide based initiators,11,12 since they offer diverse and rather simple synthetic pathways for producing aliphatic polyesters with advanced structures. For example, dihydroxyterminated homopolymers,13,14 triblock copolymers,15 macromonomers,16 and star-shaped polymers17 have been synthesized. Recently, Kricheldorf and co-workers18,19 reported that cyclic tin initiators could be used in the macrocyclic polymerization of lactones to obtain supermacrocycles that could further be polycondensed with dicarboxylic acid chlorides.20,21 This approach was used to obtain multiblock copolymers of -caprolactone and Bisphenol A in a one-pot procedure.22 The aim of this study has been to synthesize new degradable block copolymers (tri- and multiblock) that can
retain good tensile properties during several weeks of degradation. The aim has also been to tailor the hydrophilicity of the materials and to achieve a more hydrophilic polymer than the homopolymers from PLLA and poly(-caprolactone) (PCL). The polymers have been characterized by tensile testing, contact angle measurement, polarized optical microscopy, and AFM analysis. Preliminary results of this study were briefly described in a recent conference report.23 The tri- and multiblock copolymers have been subjected to in vitro degradation and the change in tensile properties with time has been monitored. Experimental Section Materials. L-lactide (L-LA) (Serva Feinbiochemica, 98%) was purified by recrystallization in dry toluene and subsequently dried under reduced pressure (10-2 mbar) at room temperature for at least 48 h prior to polymerization. 1,5Dioxepan-2-one (DXO) was synthesized according to the literature8 and purified by two subsequent distillations followed by drying over CaH2 and a final distillation under reduced pressure. Succinyl chloride (95%, Acros Organics) was used as received. Chloroform (Labora Chemicon) stabilized with 2-methyl-2-butene, was dried over calcium hydride for at least 24 h and distilled under reduced pressure in an inert atmosphere just before use. Initiators. The cyclic tin alkoxide 1,1,6,6-tetra-n-butyl1,6-distanna-2,5,7,10-tetraoxacyclodecane was synthesized from dibutyltin oxide and 1,2-ethanediol as described in the literature.14,24
10.1021/bm015658z CCC: $22.00 © 2002 American Chemical Society Published on Web 04/17/2002
602
Biomacromolecules, Vol. 3, No. 3, 2002
Polymerization. All reaction vessels were silanized and all glassware and syringes were flame-dried before use. The synthesis of triblock copolymers from L-LA and DXO was performed according to the literature.15 In principle, a silanized round-bottomed flask (50 mL) equipped with a magnetic stirring bar and closed by a three-way valve was used as reaction vessel. DXO, initiator and chloroform were added under an inert atmosphere and allowed to react completely by ring-expansion polymerization before the second monomer, L-LA was added. For the synthesis of triblock copolymers, the final cyclic block copolymer was precipitated into a methanol/hexane mixture, forming the linear triblock copolymer. The synthesis of a multiblock copolymer was accomplished by first synthesizing a low molecular weight cyclic triblock copolymer with the composition [DXO]/[I] ) 120 and [L-LA]/[I] ) 60. Succinyl chloride was added to this cyclic triblock copolymer in an equivalent amount to the initiator. Approximately three triblock units had formed one multiblock unit after 4 h of polycondensation at 60 °C, and the crude reaction mixture was precipitated into a methanol/ hexane mixture and dried to constant weight. Solution Casting. Triblock and multiblock copolymer films were prepared by dissolving the high MW triblock copolymers and the multiblock copolymer in chloroform and solution-cast them into thin films (0.1 mm) on glass plates. The solvent evaporated, and the films were dried under reduced pressure for at least 1 week before analysis. For the contact angle measurements, two additional films of the homopolymers of DXO and L-LA were cast in the same way. Sample Preparation and in Vitro Degradation. For tensile testing and degradation studies, rectangular plates, 8 × 0.5 cm in size, were punched from solution-cast films with a thickness of 0.1 mm. The samples were subjected to hydrolytic degradation in a saline buffer at pH 7.4 and 37 °C. The saline buffer contained the following per liter of water: 9 g of NaCl, 10.73 g of Na2HPO4‚7H2O, and 2.12 g of NaH2PO4. The pH of the buffer solution was adjusted to pH 7.4 by the addition of NaOH. To prevent microbial growth, 100 µL of a 0.04 wt % NaN3 solution was added to the saline solution. The in vitro degradation experiments were performed in Petri dishes filled with 30 mL of saline solution. The Petri dishes were placed in a thermostatically controlled chamber at 37 °C, and subjected to a gentle shaking motion. Samples were regularly withdrawn from the test environment and washed with deionized water before being dried under vacuum to constant weight. Instrumental Methods Nuclear Magnetic Resonance. The monomer conversion, polymer composition and molecular weight of the polymers were determined using 1H NMR spectroscopy. The measurements were performed using a Bruker Avance 400 Fourier transform nuclear magnetic resonance spectrometer (FTNMR) operating at 400 MHz, T ) 25 °C, with chloroformd1 (CDCl3) as solvent. The samples, 25 mg, were prepared in sample tubes with a diameter of 5 mm and dissolved in
Ryner and Albertsson Scheme 1
Scheme 2
0.5 mL CDCl3. Nondeuterated chloroform was used as an internal standard (δ ) 7.26 ppm). Size Exclusion Chromatography. The molecular weight and molecular weight distribution (PDI) were determined by size exclusion chromatography (SEC). The analysis was performed at room temperature using a Waters 717plus autosampler and a Waters model 510 apparatus equipped with three PLgel 10 µm mixed-B columns, 300 × 7.5 mm (Polymer Labs.). Spectra were recorded with an PL-ELS 1000 evaporative light scattering detector (Polymer Labs.) connected to an IBM-compatible PC. Millenium32 software version 3.20, was used to process the data. Chloroform was used as eluent at a flow rate of 1.0 mL/min. Narrow polystyrene standards in the 1700-706 000 range were used for calibration. Differential Scanning Calorimetry. The thermal properties of the synthesized triblock and multiblock copolymers were determined by differential scanning calorimetry (DSC), using a Mettler-Toledo DSC instrument with a DSC 820 module. A scanning rate of 10 °C/min was used and the samples were heated in a nitrogen atmosphere. The second
Biomacromolecules, Vol. 3, No. 3, 2002 603
Resorbable and Highly Elastic Block Copolymers Table 1. Characterization of Synthesized Tri- and Multiblock Copolymers
% convn name
[DXO]/[I]:[LLA]/[I]a
triblock 400:200 triblock 300:150 triblock 200:100 triblock 300:100 triblock 400:100 triblock 120:60 multiblock 120:60
400:200 300:150 200:100 300:100 400:100 120:60 3×(120:60)
Mn(theor)
Mwb
PDIb
75 200 56 400 37 600 49 200 60 800 22 600 67 700
198 000 124 000 77 100 93 400 122 900 50 400 117 700
1.26 1.25 1.07 1.13 1.18 1.07 1.19
DXO
(%)c
65 73 74 84 88 66 69
DXOd
LLAd
99.6 99.7 97.7 99.6 97.3 92.0
90.2 90.4 93.1 83.7 80.0 94.0
a Calculated from the feed ratio. b Determined by SEC analysis, calibration with polystyrene standards. c DXO content (mol %) in triblock copolymer determined by 1H NMR analysis on precipitated samples. d Determined by 1H NMR analysis on crude reaction mixture.
scan was used to record the heat of fusion. In evaluating the crystallinity of the copolymers, it was assumed that the only contribution to the heat of fusion was from the PLLA segments. According to earlier results obtained by DSC, PDXO is totally amorphous, having a Tg between -36 °C.8 Tensile Testing. The tensile testing of the polymers was carried out in an Instron 5566 equipped with pneumatic grips and controlled by a Dell 466/ME personal computer. The tensile measurements were performed with a cross-head speed of 100 mm/min and an initial grip separation of 32 mm. The elongation-at-break (b) was calculated from the grip separation, due to the large elongation of the samples. The films had a thickness of approximately 0.1 mm and were preconditioned (48 h at 50 ( 5% RH and 23 ( 1 °C) before testing. Five samples from the same film were tested for each copolymer. The mean thickness of each sample was calculated from five different measurements with a Mitutoyo micrometer. All procedures were according to ASTM D88295A. Contact Angle. The contact angle measurements were conducted on a Rame´ Hart goniometer. Deionized water (Millipore, resistivity: 18.4 MΩ cm) was used. The receding angle was obtained by withdrawing water from the droplet until the three-phase line started to recede. The receding angle was calculated from the mean value of 10 contact angle measurements at five different places on the film. Polarized Optical Microscopy. A polarized optical microscopy (Leitz Ortholux POL-BK II) was used to examine the morphology of the solution-cast films. Atomic Force Microscopy. Atomic force microscopy (AFM) tapping mode phase imaging were performed at the Institute for Surface Chemistry, Stockholm, Sweden, using a Nanoscope IIIA multimode atomic force microscope. The scan rate was in the range of 1.0-2.3 Hz. Some films had been crystallized from melt at 160 °C, before analysis. Results and Discussion Polymerization. Poly(L-lactide-b-1,5-dioxepan-2-one-btriblock and poly(L-lactide-b-1,5-dioxepan-2-one) multiblock copolymers were synthesized according to Schemes 1 and 2. Part of the synthesis has been described earlier in the literature.15 In principle, the middle block is first synthesized by ring-expansion polymerization of DXO with the cyclic initiator 1. The second monomer, L-LA is then added to the reaction, and the side blocks are formed simultaneously. A triblock copolymer consisting of a middle L-lactide)
Figure 1. SEC chromatograms from the multiblock synthesis.
Figure 2.
13C
NMR from the carbonyl region of multiblock 120:60.
block of 200 DXO units and two side blocks of 50 lactide units each, is named “triblock 200:100”. Six triblock copolymers of different block compositions were successfully synthesized (Table 1). The PDXO block was varied between 120 and 400 DXO units, and the PLLA side block between 30 and 100 lactide units. The theoretical molecular weight ranged from 22 600 to 75 200. The number-average molecular weight (Mn) obtained from the SEC analysis on precipitated polymers was approximately twice as high as the theoretical Mn. This was expected and can be subscribed to the use of polystyrene standards for calibration of the molecular weight. The molecular weight distribution (PDI) was narrow in all cases, 1.07-1.26. In the synthesis of the new multiblock copolymer, a low MW triblock copolymer (triblock 120:60) was first prepared, and then directly polycondensated in a one-pot procedure with succinyl chloride to form a multiblock copolymer. The
604
Biomacromolecules, Vol. 3, No. 3, 2002
Figure 3. Receding contact angle decreasing with increasing DXO content: ([) triblock copolymers; (4) multiblock copolymers.
multiblock copolymer is named “multiblock 120:60”, meaning that the polylactide blocks within the multiblock are 60 (actually 30 + 30) lactide units long, whereas the end blocks consist of 30 lactide units, Scheme 2. The synthesis is similar to the multiblock synthesis recently described by Kricheldorf and co-workers.21,25 The SEC analysis showed a significant increase in molecular weight after the polycondensation reaction (Figure 1), indicating a successful polycondensation. The shoulder to the right of the SEC traces can be subscribed to the formation of a small amount of oligomers. From the SEC analysis, it was calculated that approximately three triblock copolymers formed one multiblock copolymer, Table 1. The 13C NMR analysis showed only two distinct carbonyl peaks at 171.3 and 169.6 ppm, assigned to DXO-DXODXO (DDD) and LLA-LLA-LLA (LLL) confirming that a multiblock copolymer had been formed. Transesterification reactions would have resulted in carbonyl peaks from the copolymers DDL, LDL, LLD, and DLD. These carbonyl
Ryner and Albertsson
peaks have been reported to occur at 170.8, 170.7, 170.1, and 169.7 ppm, respectively.26 No carbonyl groups from transesterification reactions were detected, Figure 2. In the synthesis of high MW triblock copolymers, the reaction mixture became very viscous toward the end of the reaction. This can be explained by favorable Sn-O donoracceptor interactions,27 which might act as physical crosslinks in the reaction mixture. These interactions limited the upper molecular weight for the triblock synthesis. In the multiblock synthesis this problem was avoided, which was a great advantage. After the synthesis of low MW triblock copolymers, the dicarboxylic acid chloride was added. The viscosity dropped immediately and stayed low during the ring-opening polycondensation reaction. Hydrophilicity. Contact angle measurements were performed to investigate the hydrophilicity of the films. The receding angle is plotted vs the DXO content of the film in Figure 3. The higher the DXO content, the more hydrophilic the surface became. It was possible, with both triblock and multiblock copolymers, to tailor the hydrophilicity of the film surface by changing the composition of the copolymer. A DXO content of 65% gave a receding angle of 44 ( 3°, and with a DXO content of 88% the receding angle decreased to 27 ( 1°. These are all very hydrophilic surfaces compared to PLLA and PCL, which have been reported to have contact angles of 74 and 64° respectively.28,29 A hydrophilic material generally has a better biocompatibility than a hydrophobic material.30 Morphology. The morphology was examined using a polarizing optical microscope, Figure 4. The PLLA blocks of the triblock copolymers were able to phase separate into spherulites separated by large amorphous domains. PLLA is known to form spherulites upon crystallization.31,32 All triblock copolymers had spherulites with a diameter varying between 20 and 80 µm. The multiblock had much shorter
Figure 4. Polarized optical microscopy on (a) triblock 400:200, 65% DXO, (b) triblock 300:150, 73% DXO, (c) triblock 400:100, 88% DXO, and (d) multiblock 120:60, 66% DXO.
Resorbable and Highly Elastic Block Copolymers
Figure 5. AFM analysis on triblock 400:200: (a) solution-cast film; (b) film after crystallization; (c) 3D representation of surface after crystallization.
PDXO blocks and did not phase separate in the same distinct manner, Figure 4d. It follows from Figure 4 that the crystallinity of the films increased with decreasing DXO content, as expected. The solution-cast films of triblock 400:100, triblock 400: 200, and multiblock 120:60 were studied with AFM to examine the surface morphology before and after crystallization from melt. The solution-cast films were smooth and showed no phase separation at the surface (Figures 5a, 6a, and 7a). After crystallization, the AFM analysis showed completely different surfaces (Figures 5b, 6b, and 7b). As expected, crystalline domains were present on all film surfaces and the triblock 400:200 had the most ordered structure. Triblock 400:200 had a higher L-LA content than triblock 400:100 and therefore crystallized to a greater extent. This can be observed if Figures 5b and 6b are compared, agreeing with the results of the polarized optical microscope.
Biomacromolecules, Vol. 3, No. 3, 2002 605
Figure 6. AFM analysis on triblock 400:100: (a) solution-cast film, (b) film after crystallization; (c) 3D representation of surface after crystallization.
Thermal Analysis. The thermal analysis of the copolymers gave the glass transition and melting temperatures (Tg and Tm) listed in Table 2. The homopolymer of DXO is a completely amorphous polymer with a reported Tg around -36 °C.8 PLLA has been reported to have Tg ) +55 °C and Tm ) 169 °C.31 The synthesized copolymers are all block copolymers, and therefore the Tg and Tm values are expected to be additive and not average values. A large glass transition at approximately -33 °C was detected from the PDXO block. The Tg from the PLLA blocks was to small to be observed, as a consequence of the low L-LA content (600%), plastic deformation of the copolymers gave rise to the saw tooth features visible in Figure 8. These block copolymers had better mechanical properties than the random DXO/L-LA copolymer described in the literature.36 The mechanical
Resorbable and Highly Elastic Block Copolymers
Figure 9. Change in stress-at-break during in vitro degradation of triblock 300:150 and multiblock 120:60.
Biomacromolecules, Vol. 3, No. 3, 2002 607
were successfully synthesized in a controlled manner. The multiblock copolymer was synthesized in a one-pot procedure by ring-opening polycondensation of a nonterminated low molecular weight triblock copolymer. This reaction pathway had the great advantage of low reaction viscosity during the complete synthesis, which is a limitation in the triblock synthesis of high MW polymers. It was possible to tailor the hydrophilicity of the copolymers by changing the composition, and the hydrophilicity increased with increasing DXO content. All the polymers were very strong and had very high elongations at break. By changing the PDXO block length, the tensile properties of the materials could easily be controlled. The copolymers retained good mechanical properties during several weeks of in vitro degradation. Acknowledgment. The authors thank the Swedish Research Council for Engineering Sciences (TFR), Grant No. 1999-658, for financial support for this work. References and Notes
Figure 10. Change in elongation-at-break during in vitro degradation of triblock 300:150 and multiblock 120:60.
properties were comparable with the DXO/CL triblock copolymer37 or natural rubber (b ≈ 600-700%). Triblock 300:150 and multiblock 120:60 were subjected to in vitro degradation and the tensile properties were tested regularly with time. At the onset of the hydrolysis, the stressat-break immediately started to decrease, Figure 9. After 21 days of in vitro degradation, the stress-at-break had decreased to 68% of its initial value, for both polymers. The elongation at break was still very high, 930 ( 100% for the triblock and 680 ( 44% for the multiblock, Figure 10. After 59 days, triblock 300:150 had a stress-at-break of 22.1 ( 2 MPa and multiblock 120:60 had a stress-at-break of 7.9 ( 1.5 MPa, corresponding to 52% and 30% of their initial values. No significant weight percent loss was observed during the studied period of time. However, the Mn of the polymers started to decrease immediately after immersion into the saline buffer. The change in molecular weight was consistent with an earlier study on these materials.38 After 59 days of degradation, the Mn of the triblock copolymer had decreased with 40% (Mn ) 59 500) and the multiblock copolymer with 55% (Mn ) 44 500). In conclusion, the mechanical properties of both the triblock and the multiblock copolymers decreased during the hydrolysis. However, both copolymers still had relatively high stress-at-break as well as high elongation-at-break (approximately 800% and 600%, respectively), after 59 days of in vitro degradation. The mechanical properties were retained longer for the triblock copolymer than for the multiblock copolymer. This was probably due to the differences in morphology and molecular architecture in the block copolymers. Conclusions A new poly(LLA-b-DXO) multiblock copolymer and five different poly(LLA-b-DXO-b-LLA) triblock copolymers
(1) Dubois, P.; Jacobs, C.; Jerome, R.; Teyssie, P. Macromolecules 1991, 24, 2266-2270. (2) Kowalski, A.; Libiszowski, J.; Duda, A.; Penczek, S. Macromolecules 2000, 33, 1964-1971. (3) Kowalski, A.; Duda, A.; Penczek, S. Macromolecules 2000, 33, 7359-7370. (4) Kricheldorf, H. R.; Jonte´, J. M.; Berl, M. Makromol. Chem., Suppl. 1985, 12, 25-38. (5) Li, Y. X.; Kissel, T. Polymer 1998, 39, 4421-4427. (6) Kowalski, A.; Duda, A.; Penczek, S. Macromol. Rapid Commun. 1998, 19, 567-572. (7) Duda, A.; Florjanczyk, Z.; Hofman, A.; Slomkowski, S.; Penczek, S. Macromolecules 1990, 23, 1640-1646. (8) Mathisen, T.; Masus, K.; Albertsson, A.-C. Macromolecules 1989, 22, 3842-3846. (9) Lo¨fgren, A.; Albertsson, A.-C.; Dubois, P.; Jerome, R.; Teyssie, P. Macromolecules 1994, 27, 5556-5562. (10) Lo¨fgren, A.; Albertsson, A.-C. Polymer 1995, 36, 3753-3759. (11) Ryner, M.; Stridsberg, K.; Albertsson, A. C.; von Schenck, H.; Svensson, M. Macromolecules 2001, 34, 3877-3881. (12) von Schenck, H.; Ryner, M.; Albertsson, A.-C.; Svensson, M. Macromolecules 2002, 35,1556-1562. (13) Stridsberg, K.; Ryner, M.; Albertsson, A.-C. Macromolecules 2000, 33, 2862-2869. (14) Stridsberg, K.; Albertsson, A.-C. J. Polym. Sci., Polym. Chem. 1999, 37, 3407-3417. (15) Stridsberg, K.; Albertsson, A.-C. J. Polym. Sci., Polym. Chem. 2000, 38, 1774-1784. (16) Ryner, M.; Finne, A.; Albertsson, A.-C.; Kricheldorf, H. R. Macromolecules 2001, 34, 7281-7287. (17) Finne, A.; Albertsson, A.-C. Manuscript in preparation. (18) Kricheldorf, H. R.; Lee, S. R. Macromolecules 1995, 28, 67186725. (19) Kricheldorf, H. R.; Lee, S. R.; Bush, S. Macromolecules 1996, 29, 1375-1381. (20) Kricheldorf, H. R.; Eggerstedt, S. J. Polym. Sci., Part A Polym. Chem. 1998, 36, 1373-1378. (21) Kricheldorf, H. R.; Langanke, D.; Spickermann, J.; Schmidt, M. Macromolecules 1999, 32, 3559-3564. (22) Kricheldorf, H. R.; Eggerstedt, S. Macromolecules 1998, 31, 64036408. (23) Ryner, M.; Albertsson, A.-C. Macromol. Symp. 2001, 175, 11-18. (24) Gsell, R.; Zeldin, M. J. Inorg. Nucl. Chem. 1975, 37, 1133-1137. (25) Kricheldorf, H. R.; Fechner, B. Macromolecules 2001, 34, 35173521. (26) Albertsson, A.-C.; Lo¨fgren, A. J. Macromol. Sci.sPure Appl. Chem. 1995, A32, 41-59. (27) Davies, G. A. Organotin Chemistry; VCH Publishers: Weinheim, Germany, and New York, 1997; Chapter 12. (28) Tsuji, H.; Muramatsu, H. J. Appl. Polym. Sci. 2001, 81, 2151-2160.
608
Biomacromolecules, Vol. 3, No. 3, 2002
(29) Averous, L.; Moro, L.; Dole, P.; Fringant, C. Polymer 2000, 41, 4157-4167. (30) Hunt, J. A.; Meijs, G.; Williams, D. F. J. Biomed. Mater. Res. 1997, 36, 542-549. (31) Vasanthakumari, R.; Pennings, A. J. Polymer 1983, 24, 175-178. (32) Tsuji, H.; Ikada, Y. Polymer 1995, 36, 2709-2716. (33) Tsuji, H.; Ikada, Y. J. Appl. Polym. Sci. 1998, 67, 405-415. (34) Tsuji, H.; Mizuno, A.; Ikada, Y. J. Appl. Polym. Sci. 2000, 77, 14521464.
Ryner and Albertsson (35) Andronova, N.; Ryner, M.; Albertsson, A.-C. Manuscript in preparation. (36) Lo¨fgren, A.; Albertsson, A.-C.; Zhang, Y. Z.; Bjursten, L. M. J. Biomater. Sci.sPolym. Ed. 1994, 6, 411-423. (37) Lo¨fgren, A.; Renstad, R.; Albertsson, A. C. J. Appl. Polym. Sci. 1995, 55, 1589-1600. (38) Stridsberg, K.; Albertsson, A.-C. Polymer 2000, 41, 7321-7330.
BM015658Z
