Increased Impact Strength of Biodegradable Poly(lactic acid)-Poly(butylene succinate) Blend Composit
Increased Impact Strength of Biodegradable Poly(lactic acid)/Poly(butylene succinate)Blend Composites by Using Isocyanate as a Reactive Processing Agent
Masaki Harada,1Tsubasa Ohya,2Kouji Iida,1Hideki Hayashi,1Koji Hirano,1Hiroyuki Fukuda1 1Nagoya Municipal Industrial Research Institute,Atsuta-ku Rokuban,Nagoya456-0058,Japan
2Aichi Institute of Technology,Yachigusa,Yakusa-cho Toyota470-0356,Japan
Received21November2006;accepted3April2007
DOI10.1002/app.26717
Published online18July2007in Wiley InterScience(https://www.wendangku.net/doc/5618163929.html,).
ABSTRACT:Poly(lactic acid)(PLA)blended with poly (butylene succinate)(PBS)was prepared in the presence of lysine triisocyanate(LTI)by using a twin-screw extruder and injection molding machine.The physical properties, rheological behavior,compatibility,and morphology were investigated by using a tensile test,a Charpy impact test, melt mass-?ow rate(MFR)measurements,size exclusion chromatography(SEC),and laser scanning confocal micros-copy(LSCM).The impact strength of PLA/PBS(90/10wt%) blend composite was about18kJ/m2in the absence of LTI, and it increased to50–70kJ/m2in the presence of LTI at0.5 wt%.The MFR value of PLA/PBS(90/10wt%)decreased from25g/10min at2008C in the absence of LTI to approxi-mately3g/10min in the presence of LTI.These results imply that isocyanate groups of LTI reacted with both terminal hydroxyl or carboxyl groups of the polymers.Spherical par-ticles at1m m were observed by using LSCM in the presence of LTI.These results indicate that the LTI is a useful reactive processing agent to increase the compatibility of PLA/PBS blend composites to increase the impact strength of PLA.ó2007Wiley Periodicals,Inc.J Appl Polym Sci106:1813–1820,2007 Key words:PLA;biodegradable;biopolymers;reactive processing;compatibility
INTRODUCTION
Poly(lactic acid)(PLA)is a promising synthetic bio-polymer derived from biomass through bioconversion and polymerization.1Today PLA is not expensive,as with other engineering plastics,but it used to be ex-pensive because of complex processing procedures. The disadvantage of PLA is that it is too soft above its glass transition temperature(608C)and has low toughness,which limits its applications.Its brittleness is also a disadvantage for its application in various commercial items.The soft biodegradable plastic poly (caprolactone)(PCL)has been blended to PLA to increase the impact strength to solve the brittleness.2 However,soft biodegradable plastics decrease the ten-sile strength of blend composites at high impact strengths(30–50wt%blend).Copolymerization of PLA/PCL also decreases the brittleness of PLA.2–4 Various compatibilizers,such as PLA-co-PCL,have been studied to increase the miscibility of the blend composites to increase the impact strength.5,6 PLA-based urethane composites(PEU)have been syn-thesized by reacting PLA with diisocyanate,such as1,6-hexamethylene diisocyanate(HMDI).7–9PEU as a compatibilizer or an additive increases the impact strength.Kenaf at30wt%10or poly(ethyleneglycol)at 10wt%11are also useful to improve the physical properties.The impact strength of PLA/poly(butylene adipate-co-terephthalate)(PBAT)is also increased when20wt%of PBAT is added.12
Miscibility,crystallization behavior,and morphol-ogy of the PLA/poly(butylene succinate)(PBS)com-posites have been reported.13,14The average size of the PLA domain is about30l m as seen by polarized optical micrographs.13Synchrotron small-angle X-ray scattering data shows well-de?ned double-scattering peaks during crystallization,indicating that this sys-tem has dual lamellar stacks.14However,we believe the physical properties and characterization of PLA/ PBS blend composites using a reactive processing agent have not been reported.
Reactive processing is a frequently used compati-bilization strategy for cost-effective production of new multiphase polymeric materials with outstand-ing physical and chemical properties.15Successful application of the principles of reactive processing has produced considerable technological opportuni-ties for compatibilization of immiscible polymer blends.Terminal or graft groups react with the ter-minal group of the other polymer in the extrusion process during melt reactions.Copolymers of poly (propylene)having poly(maleic anhydride)can react
Correspondence to:K.Iida(iida.kouji@nmiri.city.nagoya.jp). Contract grant sponsor:Japan Keirin Association,Tokyo, Japan.
Journal of Applied Polymer Science,Vol.106,1813–1820(2007)
with polyamide(PA),6and with poly(styrene)having an amino group.Because blends of acrylonitrile–butadiene–styrene(ABS)and PA are highly immisci-ble,compatibilization is achieved by using maleic anhydride-functionalized ABS.15In the biodegrad-able plastics?eld,maleic anhydride,16–18triphenyl phosphite(TPP),19and dicumylperoxide(DCP)20are added to PLA,PCL,and PBS.
We hoped that isocyanates may be useful for reac-tive processing of PLA/PBS blend composites in the early stage of this study,because hydroxyl and car-boxylic groups that are hydrolyzed in PLA and PBS under heat and high shear stress in an extruding equipment19may react with isocyanate groups.Fig-ure1shows the predicted reaction of PLA,PBS,and reactive processing agents lysine triisocyanate(LTI, 2-isocyanate ethyl-2,6-diisocyanate ohexanoate,Fig.
1)and lysine diisocyanate(LDI,methyl2,6-diisocya-natohexanoate,Fig.1).We searched various isocya-nates for biodegradable material,and we found that LTI and LDI may be most suitable due to the main chain of atoms in lysine that is a kind of amino acid residue.PLA-based polyurethanes are synthesized by using HMDI,so we focused on LTI as a reactive processing agent because the LTI molecules have three isocyanate groups in,and we expected that these three isocyanate groups are useful to increase the chance of reaction at the PLA/PBS interface to increase its physical properties.
In this study,PLA was blended with PBS in the presence of LTI and LDI(Fig.1),and their physical properties,rheological behavior,phase behavior and morphology were investigated by using the tensile test,Charpy impact test,melt mass-?ow rate(MFR) measurements,size exclusion chromatography(SEC), and laser scanning confocal microscopy(LSCM).In particular,this study focused on making the impact test more effective by using LTI compared with LDI (Fig.1)and the observation of morphology in reac-tive processing to have a better understanding of the reactively compatibilized structure of PLA/PBS.
MATERIALS AND METHODS Materials
PLA(Toyota Motor)(Catalog No.CSL40529)and PBS(Showa Highpolymer)(Catalog No.1020)was dried at808C for12h before use.LTI and LDI (Kyowa Hakko Kogyo)were used as received. Blending
The blends were processed in a twin-screw extruder (KZW15-30TGN,L/D530,Technovel)at1908C(die) and300rpm.The extrusion rate was1kg/60min and the accumulation time was approximately2min. Three hundred and sixty gram PLA and40g PBS were mixed in a1-L beaker and then2g LTI or LDI was added.The mixed pellets were extruded and dried for4h at808C in an oven.
Injection molding
The test specimens were prepared from the extruded blends by using an injection molding machine(SE 18S Sumitomo Heavy Industry)with a nozzle at 2208C.The temperature of the mold was308C and the peak pressure of injection was40–60MPa.The cycle time was approximately50s.
Tensile strength
The tensile tests were done by using an Instron4505 tensile testing machine equipped with a computer adapted to the standard ISO527-1993.The samples were stored at rest for48h at238C and50%RH before the test.The crosshead speed was5mm/min. Impact strength
Charpy impact tests were done by using a Zwick impact tester(JT TOHSI Inc.)and a pendulum of8J (hanmer weight1.2kg)adapted to the standard ISO 179-1993(E)at238C and50%RH.Testing was done on unnotched specimens of8031034mm3.
The
impact blow was in the edgewise direction.Samples were stored at rest for 88h at 238C and 50%RH before test.MFR
The MFR was used to characterize the melt viscosity of PLA/PBS blend composites.The MFR was mea-sured by using a Ray–Ran apparatus (Imoto Seisa-kusho,Kyoto;MB-1)at standard conditions of 2008C and 2.16kg nominal load adapted to the standard ISO 1133-1997.SEC
Molecular weights (M w and M n )and molecular weight distributions (M w /M n )were determined by using a JASCO (PU-2080,CO-2065,and RI-2031)operating with CHCl 3.Samples (10mg)were dis-solved in 10mL CHCl 3and were ?ltered by using a 0.5l m pore ?lter (Advantec,DISMIC-13).The molecular weights were calibrated using polystyrene standards (Showa Denko K.K.,SM105).Observation of morphology
The morphology of the impact tested surface of injection-molded specimens was examined by using a LSCM (Olympus corporation,OLS-1200).The LSCM images were in the re?ecting mode.
RESULTS AND DISCUSSION
Physical properties of PLA/PBS blend composites in the presence of LTI
PLA was blended with viscous PBS with LTI as a re-active agent and the mechanical properties,rheological
properties and compatibility of the blend composites were investigated.Blends of PLA with 5,10,15,20wt %of PBS were prepared.LTI was added at 0.0–1.0wt %to the PLA/PBS blend composites.
Figure 2shows stress-strain curves of PLA/PBS (90/10wt %)blend composites in the presence of LDI and LTI,and the PLA in the absence of a reac-tive processing agent.Figure 3shows the tensile strength and ultimate strain of PLA/PBS (90/10wt %)blend composites in the presence of LTI.Adding LTI by 0.3wt %did not improve the tensile strength,but gradually decreased it (Fig.3).This result agreed with that of Hiljanen-Vainio et al.2who reported that the tensile strength of PLA/PCL blend compo-sites,with PLA–PCL copolymer as a compatibilizer,gradually decreased.However,the ultimate strain was higher in the presence of LTI than its absence and increased with increasing LTI to 0.3wt %LTI.The effect of 0.5wt %LDI was as same as that of 0.15wt %LTI (Fig.2).These results imply that LTI is effective for PLA/PBS blending,and agrees with the results of Semba et al.,20who reported that the tensile strength of PLA/PCL blend composites with 0.1–0.2wt %DCP as a compatibilizer gradually decreases and the ultimate strain increases.
Figure 4shows the dependence of the impact strength on the PBS content (5–20wt %)in both the presence and absence of LTI.The impact strength increased with increasing PBS content in the pres-ence of LTI (0.5wt %).The impact strength of 10–15wt %PBS content was more or less saturated at 50–70kJ/m 2.The impact strength of over 20wt %PBS content could not be measured in the presence of LTI because the test specimens were not breakable.Although 30wt %PBS was added to PLA in the absence of LTI,the impact strength did not increase (18kJ/m 2).In the presence of LTI,the impact strength of 10wt %PBS increased 2–3times
and
Figure 2Stress–strain curves of PLA/PBS (90/10wt %)blend composites in the presence of LTI and LDI,and PLA in the absence of LTI and LDI.The inset gives details of stress–strain of the blends in the neighborhood of yield
points.
Figure 3Tensile strength,ultimate strain and LTI (wt %)in PLA/PBS (90/10wt %)blend composites.(n )Tensile strength (MPa).(~)Ultimate strain (%).
BIODEGRADABLE PLA/PBS REACTIVE PROCESSING 1815
Journal of Applied Polymer Science DOI 10.1002/app
was the same as for normal ABS (Technopolymer,Catalog No.350)(unnotched,60–80kJ/m 2).Adding LDI to PLA with 20wt %PBS was not as effective for in?uencing the impact strength (31kJ/m 2,Fig.4),indicating that LTI is the more useful reactive processing agent for impact strength.In the case of blending of PCL as viscous biodegradable polymer in the absence of LTI to PLA,the impact strength of PLA clearly increases with 20wt %of PCL.2In this study,using PBS with LTI as a reactive agent,the impact strength clearly increased with the lower amount of the viscous polymer PBS.Preliminary results showed that the impact strength of PLA/PCL blend composites in the presence of LTI also increases.21,22
Figure 5shows the effect of LTI on the impact strength at PBS 5,10,and 20wt %in PLA.At 5wt %PBS,the impact strength increased gradually with increasing LTI and was saturated at 1.0wt %LTI.At 10and 20wt %PBS,the impact strength sud-denly increased and was saturated at 0.15wt %LTI.
Blending of high molecular weight PBS (Catalog No.1001;M w 5128,000)in the absence of LTI did not increase the impact strength in PLA blend compo-sites (data not shown).Thus an increase in impact strength does not depend on the molecular weight of PBS but depends on the presence of LTI.The reac-tion might occur between isocyanate groups of LTI and hydroxyl or carboxyl groups at the terminals of the polymer (Fig.1)so that the interfacial adhesion between PLA and PBS might increase the impact strength.
Rheological properties of PLA/PBS blend composites in the presence of LTI
Adding LTI was effective to increase the ultimate strain and impact strength of PLA/PBS blend com-posites,but was not effective to increase the tensile strength.The addition level of LTI required to opti-mize physical properties of PLA/PBS (90/10wt %)is approximately 0.15wt %,as well as for the impact strength of ABS.To investigate the rheological prop-erties of reactive processing,the MFR was measured.Figure 6shows the dependence of adding LTI at 5,10,and 20wt %PBS with PLA on the MFR,which decreased with increasing LTI,implying that the vis-cosity increased,indicating that LTI has a crucial role in the rheology of PLA/PBS blending.The total trend of MFR level decreased with increase in PBS amount,implying that PBS might also be important for reactive blending.LDI,which has two isocyanate groups,was not effective to decrease MFR of PLA/PBS (80/20wt %)(Fig.6),implying that three,not two,isocyanate groups in the reactive processing agent decrease MFR.The viscosity of PLA and PCL cross-linked blend composites by TPP 19and DCP,20respectively,increases when the torque data is mea-sured by using an internal mixer.In this study,the sudden decrease in MFR by adding LTI at 3g/10
min
Figure 4Impact strength and PBS (wt %)in PLA.Ab-sence of reactive processing agent (&),presence of LTI (0.5wt %,l ),and presence of LDI (0.5wt %,n ).NB:not
breakable.
Figure 5Impact strength and LTI (wt %)of PLA/PBS blend composites (PLA/PBS 595/5,90/10,80/20wt
%).Figure 6MFR and LTI (wt %)in PLA/PBS blend compo-sites (PLA/PBS 595/5,90/10,80/20wt %).
1816HARADA ET AL.
Journal of Applied Polymer Science DOI 10.1002/app
suggests cross-linking as in the cases of TPP or DCP. Although the impact strength was more or less satu-rated for10wt%PBS with0.15wt%LTI(Fig.5), the MFR was not too low(17g/10min),indicating that10wt%PBS with0.15wt%LTI in PLA is suit-able for required properties and productivity of blends by injection molding.
In this study using SEC,the absence of LTI,the CHCl3solution of blend composites of PLA/PBS were easily?ltered by using a0.5l m pore?lter before the SEC measurements.However,in the pres-ence of LTI,the?lter was gradually plugged up, indicating that the blend composites of PLA/PBS in the presence of LTI might also produce cross-linking parts.So,SEC was used to measure only the soluble parts.Table I shows molecular weights of PLA/PBS blend composites by using SEC.The molecular weight of PLA in the presence of LTI was almost same for PLA without LTI.However,the molecular weight of the blend composites of PLA/PBS in the presence of LTI slightly increased to compare in the absence of LTI,suggesting that LTI might react in the presence of both PLA and PBS.These results imply that LTI may produce both high molecular weight parts and insoluble cross-linking parts(Fig.
7)in PLA/PBS blend composites.In the synthesis of PLA-based polyurethanes,the molecular weight of telechelic prepolymers that have low molecular weight increases by reaction with HMDI.7–9In this study,the molecular weight also increased during reactive processing.
Morphology of PLA/PBS blend composites in the presence of LTI by LSCM
To investigate the reactive compatibility of PLA/PBS blend composites,the fracture surface of impact tested specimens was observed by using an LSCM. The LSCM image of the fracture of PLA with0.5wt %LTI showed that the fracture surface was brittle [Fig.8(a)]and that it was similar in the absence of LTI.Figure8(b–d)show LSCM images of PLA/PBS (90/10wt%)blend composites in the presence of0 and0.15wt%LTI.Spherical particles of0.9–1.4l m diameter were in the presence of LTI[Fig.8(c,d)].In this case,PBS was a smaller amount(10–20wt%), and so PLA and PBS might form a matrix and islands,respectively,in the blend composites.In the absence of LTI[Fig.8(b)],the morphology was unstable and irregularly shaped.Spherical particles of1l m diameter were fewer[Fig.8(b),white arrows S]and domains of2–5l m[Fig.8(b),white arrows L] were many,agreeing with the observation by using transmission electron microscopy.23Table II shows the size and number of spherical particles in a153 15mm2area of the composites.The standard devia-tion of size of particles in the absence of LTI was wider than in the presence of LTI,indicating that LTI stabilizes the morphology of PLA/PBS blend composites.Clearly the number of the particles was smaller at LTI0.15wt%and their size did not depend on the LTI and PBS concentrations.
The impact tested specimens of PLA/PBS(80/20 wt%)in the presence of0.15wt%LTI were not completely broken.The fracture surface was smooth (brittle)and white(ductile).The brittle surface[Fig. 9(a,b)]was far from the stress point and1l m diam-eter spherical particles were also in the?at surface [Fig.9(b)].The fracture surface of ductile surface had many spherical cavities[Fig.9(c,d)].Thus the PBS-rich domain might form spherical particles in the presence of LTI by phase separation and adsorb the impact strength.Our results suggest that LTI
TABLE I
Molecular Weight Determination of PLA/PBS Blend
Composites by Using SEC
Blend composition
(wt%)
M n(104)M w(104)M w/M n
PLA/PBS LTI
100/009.216.3 1.77
100/00.59.716.3 1.68
90/1007.014.4 2.06
90/100.57.115.7 2.21
80/200 6.814.2 2.09
80/200.5 6.917.9 2.59
0/1000 4.910.5
2.14
Figure7Predicted reaction pathways of PLA-LTI-PBS,
PLA,and PBS.
BIODEGRADABLE PLA/PBS REACTIVE PROCESSING1817
Journal of Applied Polymer Science DOI10.1002/app
changed the PBS domain size from 1.4–4.0l m to 0.8–1.4l m under shear stress in the extruder.The morphology development during dissipative mix-melting in these reactive blends may occur by a sheeting mechanism as described by Scott and Macosko.24
Figures 1and 7show proposed chemical reactions between LTI and biodegradable polymers.First,hydroxyl or carboxyl groups of PLA and PBS,respectively,might react with two isocyanate groups of the LTI or LDI molecule (Fig.1)to form grafting PLA.However,the impact strength did not effec-tively increase,and the MFR did not decrease in the presence of LDI.These results suggest that a third isocyanate group might have an important role in the formation of physical properties.
Second,the third isocyanate group of LTI might react with PLA (Fig.7,path a)or with PBS (Fig.7,path b)for grafting and cross-linking.Increasing mo-lecular weight and decreasing MFR with increasing PBS suggest that path b might react with these com-posites.Further reactions of the terminals of paths a and b induce cross-linking of the PLA/PBS blend
system.Several other reactions occur during the syn-thesis of polyurethane between diisocyanates and a polyol.25–29The major side reactions are dimeriza-tion,trimerization,and carbodiimide,biuret and al-lophanate formation.The dimerization,trimerization,and biuret and allophanate reactions might induce cross-linking in PLA/PBS blend composites.In this study,insoluble parts of over 500nm pore ?ltration were in the CHCl 3organic solvent and a spherical structure of about 1l m was found by using
LSCM.
Figure 8LSCM images of Charpy-fractured brittle surface of the blend composites.(a)PLA in the presence of LTI (0.5wt %).(b)PLA/PBS (90/10wt %)blend composite in the absence of LTI.(c)PLA/PBS (90/10wt %)blend composite in the presence of LTI (0.15wt %).(d)Inset of (c).[Color ?gure can be viewed in the online issue,which is available at https://www.wendangku.net/doc/5618163929.html,.]
TABLE II
Number and Size of Particles in 15?15mm 2of Brittle
Surface Imaging Area
Blend composition
(wt %)Number Size (l m)Standard deviation (l m)PLA/PBS LTI 90/100.08 2.60 1.290/100.1511 1.100.2090/100.326 1.260.1990/100.526 1.290.1380/20
0.5
20
0.92
0.17
1818HARADA ET AL.
Journal of Applied Polymer Science DOI 10.1002/app
Unfortunately,direct evidence of chemical reaction by using spectroscopic methods,such as infrared spectroscopy or nuclear magnetic resonance spec-troscopy could not be observed (data not shown).Our results suggest that control of the biodegradable polymer interfaces by chemical reaction may have a crucial role in the structure of isocyanates.The sys-tematic relationship of structure and properties of biodegradable polymers should be further studied in the presence of other isocyanates as reactive process-ing agents.
In conclusion,LTI was found to be a reactive proc-essing agent in the PLA/PBS blend composites that were processed by using a twin-screw extruder,and the impact strength of sample specimens increased to 2–4times that of PLA.The impact strength much increased at 10wt %PBS content.The LTI concen-tration was suf?cient at 0.15wt %to increase the impact strength.The decrease in MFR implies that LTI reacts with polymers.The morphology study showed that spherical structures induced by a reac-tive process contribute to an increase in impact strength.
References
1.Sodergard,A.;Stolt,M.Prog Polym Sci 2002,27,1123.
2.Hiljanen-Vainio,M.;Seppa
¨la ¨,J.;To ¨rma ¨ia ¨,P.;Varpomaa,P.Macromol Chem Phys 1996,197,1503.
3.Grijpma,D.W.;Zondervan,G.J.;Pennings,A.J.Polym Bull 1991,25,327.
4.Wehrenberg,R.H.Mater Eng 1981,94,63.
5.Aslan,S.;Calandrelli,L.;Laurienzo,P.;Malinconico,M.;Migliaresi,C.J Mater Sci Mater Med 2000,35,1615.
6.Maglio,G.;Malinconico,M.;Migliozzi, A.;Groeninckx,G.Macromol Chem Phys 2004,205,946.
7.Kylma
¨,J.;Seppa ¨la ¨,J.V.Macromolecules 1997,30,2876.8.Storey,R.F.;Wiggins,J.S.;Puckett,A.D.J Polym Sci Part A:Polym Chem 1994,32,2345.
9.Stolt,M.;Hiltunen,K.;So
¨derga ˙rd,A.Biomacromolecules 2001,2,1243.
10.Serizawa,S.;Inoue,K.;Iji,M.J Appl Polym Sci 2006,100,
618.
11.Jacobsen,S.;Fritz,H.G.Polym Eng Sci 1999,39,1303.
12.Jiang,L.;Wolcott,M.P.;Zhang,J.Biomacromolecules 2006,7,
199.
13.Park,J.W.;Im,S.S.J Appl Polym Sci 2002,86,647.14.Park,J.W.;Im,S.S.J Polym Sci B 2002,40,1931.
15.Baker,W.;Scott,C.;Hu,G.-H.,Eds.Reactive Polymer Blend-ing;Hanser:Munich,2001.
16.Pan,J.;Wang,Y.;Qin,S.;Zhan,B.;Luo,Y.J Biomed Mater
Res B Appl Biomater 2005,74,
476.
Figure 9LSCM images of Charpy-fractured brittle and ductile surface of the PLA/PBS (80/20wt %)blend composites.(a)Brittle surface in the absence of LTI.(b)Brittle surface in the presence of LTI (0.5wt %).(c)Ductile surface in the pres-ence of LTI (0.15wt %).(d)Ductile surface in the presence of LTI (0.5wt %).[Color ?gure can be viewed in the online issue,which is available at https://www.wendangku.net/doc/5618163929.html,.]
BIODEGRADABLE PLA/PBS REACTIVE PROCESSING 1819
Journal of Applied Polymer Science DOI 10.1002/app
17.Kim,C.-H.;Cho,K.Y.;Park,J.-K.Polymer2001,42,5135.
18.Naber,Y.;Raquez,J.;Dubois,P.;Narayan,R.Biomacromole-
cules2005,6,807.
19.Wang,L.;Ma,W.;Gross,R.A.;McCarthy,S.P.Polym Degrad
Stab1998,59,161.
20.Semba,T.;Kitagawa,K.;Ishiaku,U.S.;Hamada,H.J Appl
Polym Sci2006,101,1816.
21.Harada,M.;Hirano,K.;Iida,K.;Hayashi,H.;Fukuda,H.J pn
Kokai Tokkyo Koho346241(2004).
22.Harada,M;Hayashi,H;Iida,K;Hirano,K.Polym Prepr Jpn
2003,52,4331.23.Kamikura,M.;Imamura,S.;Toyoda,A.;Mihara,T.DIC Tech-
nical Review2004,10,1.Available at http://www.dic.co.jp/ rd/tech/no10/01.pdf.
24.Scott,C.E.;Macosko,C.W.Polymer1994,35,5422.
25.Gogolewski,S.Colloid Polym Sci1989,267,757.
26.Petrovic,Z.;Ferguson,J.Prog Polym Sci1991,16,695.
27.Hepburn, C.Polyurethane Elastomers,2nd ed.;Elsevier
Applied Science:London,1991.
https://www.wendangku.net/doc/5618163929.html,mba,N.M.K.;Woodhouse,K.A.;Cooper,S.L.Polyur-
ethanes in Biomedical Applications,2nd ed.;1998.
29.Koscielecka,A.Acta Polymerica1991,42,221.
1820HARADA ET AL. Journal of Applied Polymer Science DOI10.1002/app