Available online at www.sciencedirect.com

Acta Biomaterialia 5 (2009) 2901–2912

www.elsevier.com/locate/actabiomat

Synthesis, characterization and surface modification of low modulipoly(ether carbonate urethane)ureas for soft tissue engineering

Feng Wang a, Zhenqing Li a, John L. Lannutti a, William R. Wagner b, Jianjun Guan a,*

a Department of Materials Science and Engineering, The Ohio State University, 2041 College Road, Columbus, OH 43210, USAb Departments of Surgery, Bioengineering & McGowan Institute for Regenerative Medicine, University of Pittsburgh, Pittsburgh, PA 15219, USA

Received 2 February 2009; received in revised form 16 April 2009; accepted 16 April 2009Available online 3 May 2009

Abstract

Flexible scaffolds are of great interest in engineering functional and mechano-active soft tissues as such scaffolds might allow mechan-ical stimuli to transfer effectively from the scaffolds to cells during tissue development. Towards this end, we have developed a family offlexible poly(ether carbonate urethane)ureas (PECUUs) with a triblock copolymer poly(trimethylene carbonate)–poly(ethylene oxide)–poly(trimethylene carbonate) (PTMC–PEO–PTMC) or pentablock copolymers PTMC–PEO–PPO–PEO–PTMC (PPO, polypropyleneoxide) as soft segments, linked by 1,4-diisocyanatobutane and putrescine. All of the PECUUs had low glass transition temperatures(<�46 �C). The PTMC–PEO–PTMC-containing PECUUs had low tensile strength and breaking strain. Replacing PEO with the similarlength PEO–PPO–PEO resulted in highly flexible and soft PECUUs possessing breaking strains of 362–711%, tensile strengths of 8–18 MPa and moduli of 5.5–7.4 MPa at room temperature in air. Under aqueous conditions at 37 �C, these polymers remained flexiblewhile their moduli were decreased to 3.4–4.0 MPa. PECUUs based on PTMC–PEO–PPO–PEO–PTMC were thermosensitive as thewater content at 37 �C was lower than that at 4 �C. PECUU using PTMC–PEO–PTMC as a soft segment showed 30% weight loss over6 weeks in PBS at 37 �C, while that using PTMC–PEO–PPO–PEO–PTMC as a soft segment had weight loss <6%. Degradation productswere found to lack cytotoxicity. The mechanical stresses and moduli of PECUUs based on PTMC–PEO–PPO–PEO–PTMC wereunchanged during the degradation. To enhance cell adhesion, PECUUs were surface modified with Arg-Gly-Asp-Ser (RGDS). Smoothmuscle cell adhesion was 114% of tissue culture polystyrene for unmodified PECUU and >180% for RGDS-modified PECUUs, with cellviability on both surfaces increasing during culture. These low moduli polyurethanes may find applications in engineering cardiovascularor other soft tissues.� 2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Poly(trimethylene carbonate); Poly(ether carbonate urethane)urea; Surface modification; Biodegradable

1. Introduction

With the increasing appreciation of the role of mechan-ics in tissue development and remodeling, designing scaf-folds with mechanical properties approximating the targettissues is often considered in addition to scaffold degrada-tion properties and the support of cell adhesion and growth[1–6]. It is believed that an appropriate combination ofchemical, biological and mechanical properties within thescaffold is more likely to create an instructive microenvi-

1742-7061/$ - see front matter � 2009 Acta Materialia Inc. Published by Else

doi:10.1016/j.actbio.2009.04.016

* Corresponding author. Tel.: +1 614 292 9743.E-mail address: guan.21@osu.edu (J. Guan).

ronment for cells to develop into functional soft tissues,either in vitro or in vivo. This is particularly true whenengineering mechano-active soft tissues such as cardiacmuscle and blood vessels, as such scaffolds would allowmechanical stimuli transfer effectively from the environ-ment to the scaffold and from the scaffold to the cells todevelop mechanically appropriate soft tissues [1–6].

As researchers have sought to explore mechanical effectsin tissue engineering, the use of the relatively stiff, simplepolyesters polylactide (PLA), polyglycolide (PGA) andtheir copolymers has been joined by new biodegradablepolymers that are better suited for soft tissue applications.Flexible behavior for scaffolding materials has been of par-

vier Ltd. All rights reserved.

2902 F. Wang et al. / Acta Biomaterialia 5 (2009) 2901–2912

ticular interest to biomaterials scientists and engineersseeking to generate novel materials. Polyesters [7,8] andpolyurethanes [9–11] have most commonly been used tocreate flexible scaffolds, with flexibility resulting fromeither chemical or physical crosslinking. One of the disad-vantages of chemically crosslinked flexible polymers is theinability to process them into three-dimensional scaffoldswith convenient solvent-related techniques such as solventcasting and salt leaching, phase separation and electrospin-ning. In contrast, physically crosslinked flexible polymersare attractive in that they are thermoplastic and can beprocessed either by solvent-based strategies or melt pro-cessing. In the past several years, we and others have devel-oped a variety of families of physically crosslinkedpolyurethanes through the molecular design of soft andhard segments [12–16]. These polymers have been success-fully fabricated into scaffolds for engineering soft tissueslike blood vessels, cardiac patches and heart valves [17–19], but there is arguably a need to provide options formaterials that have even lower stiffness (<17 MPa), whilemaintaining high flexibility. Furthermore, most of the flex-ible polymers rely on polyester linkages for hydrolytic lia-bility and they undergo bulk degradation, limiting theperiod during which scaffold-based mechanical supportcan dominate [20].

The objective of this work was to generate biodegrad-able and low moduli flexible polymers that could be usedto engineer low moduli soft tissues such as cardiac muscle(<0.5 MPa) [21] and blood vessel (�1.4 MPa) [17]. Inspiredby the highly soft nature of poly(trimethylene carbonate)(PTMC), which has a modulus of �2.9 MPa [22], PTMCwas chosen as one of the components of the soft segment.Besides its low modulus, PTMC has been shown toundergo surface degradation, which may allow the synthe-sized polyurethanes to retain mechanical properties forprolonged periods as compared with polycaprolactone-based polyurethanes. Previous work demonstrated thatthe utilization of pure PTMC as a soft segment yieldedhydrophobic polyurethanes with rather slow degradationrates and moderate to weak cell adhesive properties[23–26]. To overcome these limitations, PTMC has beencopolymerized with other biodegradable polymers toincrease hydrophilicity and degradation rate [23–26]. Inthis work, we sought to introduce variably hydrophilic seg-ments into the polyurethane to vary polyurethane hydro-philicity and degradation properties. The moderatelyhydrophilic poly(ethylene oxide) (PEO)–polypropyleneoxide (PPO)–PEO and the highly hydrophilic PEO wereincorporated into the polymers. We found that the moder-ately hydrophilic PEO–PPO–PEO preserved the slowerhydrolytic degradation characteristic of the PTMC, whilethe highly hydrophilic PEO allowed the polymer toundergo faster degradation. To enhance cell adhesion,polymers were surface modified with the cell adhesive pep-tide Arg-Gly-Asp-Ser (RGDS). Smooth muscle cells werecultured on the polymer surfaces to evaluate cytocompati-bility of the synthesized polymers.

2. Materials and methods

2.1. Materials

Pluronic L31 (EO2-PO16-EO2, mol. wt. �1100, BASF),PEO (ol. wt. 1000, Sigma) and trimethylcarbonate (TMC;Boehringer Ingelheim) were vacuum dried overnight priorto use. Butanediisocyanate (BDI; Fluka) and putrescine(Aldrich) were vacuum distilled before synthesis. Stannousoctoate (Sigma) and dimethyl sulfoxide (DMSO) weredried over 4 A molecular sieves. RGDS (Sigma) was usedas received.

2.2. Synthesis of block copolymer diols

The block copolymer diols (TMC)n–PEO–PPO–PEO–(TMC)n and (TMC)n–PEO–(TMC)n were synthesized byring-opening polymerization of TMC using either PEO–PPO–PEO or PEO as an initiator and stannous octoate as acatalyst (Scheme 1) [5,27]. The polymerization was conductedat 110 �C for 18 h under a nitrogen atmosphere. The yieldedcopolymers were washed with ethyl ether and hexane, thendried in a vacuum oven at 50 �C for 24 h. Copolymers withvariable TMC length were synthesized by utilizing differentTMC/PEO or TMC/PEO–PPO–PEO ratios (Table 1).

2.3. Polyurethane synthesis and film preparation

The poly(ester carbonate urethane)ureas (PECUUs) weresynthesized using a two-step solution polymerizationmethod (Scheme 1) [5,27]. In brief, the synthesis was con-ducted in a 250 ml three-necked flask equipped with a nitro-gen inlet and outlet. A 1 wt.% BDI (7.98 mmol) solution inDMSO was added into the flask followed by a 10 wt.%copolymer diol (3.99 mmol) solution in DMSO. Two dropsof stannous octoate were then added and served as a catalyst[28]. The reaction was carried out at 80 �C for 3.5 h undercontinuously stirring. The solution was then cooled to roomtemperature, and a 1 wt.% putrescine solution (3.99 mmol)added. The reaction was continued for 12 h at room temper-ature. The polymer solution was precipitated in a saturatedpotassium chloride solution. The polymer was thenimmersed in deionized (DI) water for 24 h to leach out thesalt, and dried under vacuum at 50 �C for 24 h. The synthe-sized polymers are abbreviated as PECUU–PPPX orPECUU–PEOX, where PPP and X represent PEO–PPO–PEO and number of TMC units, respectively.

2.4. Polyurethane film preparation

PECUU films were prepared by solution casting andsolvent evaporation. The polymers were dissolved indimethylformamide (DMF) to form a 3 wt.% solution.The solution was cast in a Teflon dish, and the solventwas evaporated in a vacuum oven at 50 �C. The formedfilm was further dried for 2 days before being used forcharacterization.

O O

O

H2N(CH2)4NH2

+

+

+

OCN(CH2)4NCO

22 16CHCH2O CH2CH2OCH2CH2O(CH2)3 HC(OCOCO 2)3O

OH

OHO

CH3

n n

CHCH2O CH2CH2OCH2CH2O HOHCH3

16 22

CHCH2O CH2CH2OCH2CH2O(CH2)3 HC(OCOCO 2)3OO

CNH(CH2)4NCOO

OCN(CH2)4NHCOOO CH3

2 216n n

nn 16 22CHCH2O CH2CH2OCH2CH2O(CH2)3OCO CO(CH2)3O

OCNH(CH2)4NHCNH(CH2)4NH

ONH(CH2)4NHCNH(CH2)4NHCO

OO OO.... ....

CH3

Scheme 1. Synthesis of PTMC–PEO–PPO–PEO–PTMC-based PECUUs.

Table 1Feed ratio and molecular composition of the copolymer diols.

Feed ratio (molTMC/moldiol) Block length of PTCM–PEO–PTMCPTMC–PEO–PPO–PEO–PTMC

PTMC–PEO–PTMC 8/1 460–1000–460PTMC–PEO–PPO–PEO–PTMC 4/1 240–88–928–88–240PTMC–PEO–PPO–PEO–PTMC 8/1 460–88–928–88–460

F. Wang et al. / Acta Biomaterialia 5 (2009) 2901–2912 2903

2.5. Polymer characterization

1H NMR spectra were obtained on a Bruker NMR spec-trometer (300 MHz) [23]. Fourier transform infrared(FTIR) spectra were recorded on a Nicolet FTIR spec-trometer. The inherent viscosity of the synthesized PECU-Us was measured using a previously described method [25].Each sample was dissolved in 1,1,1,3,3,3-hexafluoroisopro-panol (HFIP) to form a solution with a concentration of0.4 dl per 100 ml. After filtration employing a 0.45 lmpolytetrafluoroethylene filter, the measurements were con-ducted at room temperature using an Ubbelohde capillary0c. At least three measurements were made for each sam-ple. The inherent viscosity was calculated as follows:

Inherent viscosity ¼ lnðtp=tsÞ=Cp;

where tp is the passage time for the polymer solution, ts isthe passage time for the pure solvent and Cp is the polymerconcentration.

Differential scanning calorimetry (DSC) was performed ona Thermal Analysis (TA Instruments) DSC 2920 differential

scanning calorimeter. Both dry and wet samples were mea-sured. The dry samples were tested in the range of �100 to200 �C at a heating rate of 20 �C min�1. The results fromthe first run were presented. Wet samples were impregnatedwith DI water for 24 h before being tested over a temperaturerange of �10 to 90 �C at a heating rate of 10 �C min�1.

Water absorption of the polymer films at 4 and 37 �Cwere evaluated respectively. The films were soaked in Dul-becco’s phosphate-buffered saline (PBS, without calciumand magnesium, pH 7.4) at the corresponding temperature(4 or 37 �C) for 24 h. They were then taken out, gentlywiped with tissue paper and weighed. Water absorptionwas defined as the normalized difference of the wet massand dry mass of the film. Six independent measurementswere performed.

Uniaxial tensile testing was performed for PECUU filmson a TestResources 1000R load frame (model 1322)equipped with a 50 lb load cell. The films were cut withan ASTM D638-98 die into dumb-bell shape strips withlengths of 20 mm and widths of 2.5 mm [5,16,27]. Across-head speed of 10 mm/min was used. Two testing

2904 F. Wang et al. / Acta Biomaterialia 5 (2009) 2901–2912

conditions were employed for each polymer composition,i.e. in air at room temperature, and in a water bath at37 �C. Samples to be tested in the water bath wereimmersed in the 37 �C water overnight before being tested.At least four samples were evaluated for each polymercomposition at each testing condition. The initial modulusof the sample was taken from the initial linear part of thestress–strain curve [5,16,27]. The highest stress and strainat the highest stress were considered the tensile stress andbreaking strain of the sample.

2.6. Polymer degradation

To characterize polymer degradation properties, dryfilms of each polymer composition were cut into a rectan-gular shape, weighed (�20 mg) and loaded into a conicaltube containing 10 ml of PBS (pH 7.4). The degradationwas conducted in a water bath at 37 �C. At each predeter-mined time point, the samples were taken out, rinsed withDI water and weighed after drying in a vacuum oven at50 �C for at least 2 days. The weight remaining was definedas the weight after degradation (w2)/weight before degrada-tion (w1) � 100%. The mechanical properties of the sam-ples after degradation were tested in a 37 �C water bathusing the method described above.

2.7. Cytotoxicity of biodegradation products

Cytotoxicity of the degradation products of the PECU-Us after 8 weeks of degradation was evaluated as describedpreviously [29]. In brief, rat vascular smooth muscle cells(SMCs) were isolated according to Ray et al. [30]. SMCssuspended in a culture medium (Dulbecco’s modified eaglemedium (DMEM) supplemented with 10% fetal bovineserum (FBS)) were seeded into a 96-well polystyrene tissueculture plate (TCPS) at a density of 2 � 105 cells ml�1

(200 ll well�1). The degradation solution (20 ll) was thenadded to each well. After culture for 2 and 4 days, the cellviability was measured (n = 8) by colorimetric tetrazoliumsalt (MTT) assay [29]. For comparison purposes, the cul-ture medium not mixed with the degradation solutionwas used as a control.

2.8. Surface modification with RGDS peptide

To immobilize RGDS peptide on the PECUU surfacefor improved cell adhesion, PECUU was first modifiedwith O2 plasma to introduce reactive groups capable ofreacting with RGDS (Scheme 2). The plasma treatmentwas conducted in a Technics 800 Reactive Ion Etcher for2 min at 50 sccm and 100 W. The treated films were quickly

O2 plasmaCOOH

COOH

OH

OH

RGDS

RGDS

RGDS

RGDS

EDC activation

RGDS binding

Scheme 2. Scheme of PECUU surface modification with RGDS.

immersed in an 1-ethyl-3-[3-dimethylaminopropyl]carbodi-imide hydrochloride solution in PBS and remained in thesolution for 12 h to activate the reactive groups. Afterrinsing with PBS, the films were transferred into a RGDSsolution in PBS containing 20 mg ml�1 of RGDS. Thereaction was allowed to continue for 12 h at 4 �C. The filmswere rinsed extensively with PBS followed by freeze-drying.

To evaluate surface modification, X-ray photoelectronspectroscopy (XPS) spectra were acquired on a PerkinElmer (model 550) ESCA/Auger spectrometer with Mg Ka

excitation radiation and analyzed using XPSPEAKsoftware.

2.9. Smooth muscle culture

The cytocompatibility of the PECUUs with or withoutsurface modification was evaluated in terms of their abilityto support SMC adhesion [5,27,29]. Samples were punchedinto 6 mm diameter discs and sterilized with 75% ethanolfor 1 h. After rinsing with PBS, the discs were placed intothe bottom of 96-well tissue culture plates. Rat SMCs ata density of 2 � 105 cells ml�1 were seeded on the top ofeach disc. Cell adhesion was evaluated 24 h after cell seed-ing by quantifying cell viability using the MTT assay (n = 4per sample) and was defined as the percentage of opticalabsorption for the PECUU samples and for TCPS.

To evaluate cell viability changes, the seeded cells werefurther cultured. The culture medium (DMEM supple-mented with 10% FBS) was replenished every second day.At days 3 and 7, the MTT assay was performed (n = 4)for PECUU samples with or without surface modificationand the control TCPS. Fluorescent staining was conductedto show the cell morphology during the culture period. Thesamples were rinsed with PBS, fixed with 3.7 wt.% parafor-maldehyde and stained with rhodamine phalloidin (Molec-ular Probes) for F-actin and draq-5 (Biostatus Ltd) fornuclei. Images were taken on a Zeiss 510 META laser scan-ning confocal microscope.

2.10. Statistical methods

Data are expressed as means ± standard deviation (SD).Statistical comparisons were performed by analysis of var-iance with post hoc Neuman–Keuls testing.

3. Results

3.1. Synthesis of PECUUs

Fig. 1a shows a typical FTIR spectrum of the synthe-sized pentablock copolymer PTMC–PEO–PPO–PEO–PTMC diols. The copolymers exhibited characteristicabsorbance peaks of both PEO–PPO–PEO and PTMC.The peak at 1120 cm�1 was attributed to the ether bond(–C–O–C–) in PEO–PPO–PEO. The absorbance peaks at1760 and 1270 cm�1 were attributed to the carbonategroup in PTMC [31]. The copolymers also showed a broad

100015002000250030003500

Wavelength (cm-1)

-OH

-OCO-

O

-C-O-C-

4000

(3470)

(1760) (1120)

HO O OO

OO O O

HOCH3

O n 2162 n

a

b cd e

f

de

f ee d d

c

DMSO

1.03.0 2.5 2.0 1.54.04.5 3.5 ppm

H

b

d

e

f

1.03.0 2.5 2.0 1.54.04.5 3.5 ppm

a

(a)

(b)

Fig. 1. (a) FTIR spectrum of pentablock PTMC–PEO–PPO–PEO–PTMC copolymer. The wavenumbers for characteristic groups are presented in theparentheses; (b) 1H NMR spectrum of pentablock PTMC–PEO–PPO–PEO–PTMC copolymer.

F. Wang et al. / Acta Biomaterialia 5 (2009) 2901–2912 2905

peak at 3490 cm�1, belonging to the –OH groups flanked atboth ends of the copolymer diols.

Chemical compositions of the PTMC–PEO–PPO–PEO–PTMC and PTMC–PEO–PTMC copolymers were charac-terized by 1H NMR. A typical 1H NMR spectrum of thePTMC–PEO–PPO–PEO–PTMC diols is shown in

Fig. 1b. The peaks attributing to methyl (–CH3, a), methine(–CH, b) and methylene (–CH2, c) groups in the PPO seg-ment were seen at 1.03, 3.32, 3.51 ppm, respectively. Thepeak at 3.61 ppm was attributed to methylene (d) groupsin the PEO segment. The peaks belonging to PTMC seg-ments were found at 4.14 (e) and 1.95 ppm (f), respectively

18 ºC

-100 -50 0 50 100 150

Temperature ( oC)

PECUU-PPP8

PECUU-PPP4

PECUU-PEO8

End

othe

rmal

Fig. 3. DSC heating curves of dry PECUUs. The polymers were tested ina temperature range of �100 to 150 �C, with a heating rate of 20 �C min�1.

2906 F. Wang et al. / Acta Biomaterialia 5 (2009) 2901–2912

[23]. The PTMC length in the copolymer was determinedfrom the integration ratio of peaks f and d since the molec-ular weights of both PEO–PPO–PEO and PEO are known.Different PTMC lengths with numbers of TMC units rang-ing from 4 to 8 were obtained by varying the feed ratio ofTMC and PEO–PPO–PEO or PEO (Table 1).

The FTIR spectrum shown in Fig. 2 confirmed thestructure of the synthesized PECUU. The polymer dis-played a pronounced carbonyl peak of the carbonate at1760 cm�1. The absorbance at 1106 cm�1 was attributedto ether groups in the soft segment PTMC–PEO–PPO–PEO–PTMC. The urethane and urea groups showed absor-bance at 1550, 1670 and 3340 cm-1, respectively. Theabsence of absorbance at 2267 cm�1 indicated that the iso-cyanate groups were completely reacted.

The synthesized PECUUs had minimal crosslinkingwith greater than 99% solubility in DMF and DMSO.Inherent viscosity was used to characterize polymer chainlength since the PECUUs tend to stick to GPC columnsduring the molecular weight measurement due to thestrong hydrogen bonding. The polymer solutions in HFIPhad inherent viscosities of 0.87, 1.07 and 0.98 dl g�1 forPECUU–PPP4, PECUU–PPP8 and PECUU–PEO8,respectively.

3.2. Thermal properties

Fig. 3 represents DSC curves of the PECUUs in the drystate. None of the PECUUs displayed hard segment transi-tions. The polymers had glass transition temperatures below�46 �C. The lowest glass transition temperature (852.1 �C)was seen for PECUU–PPP4, which had the shortest PTMClength. The glass transition temperature increased to�46.9 �C when the PTMC length was increased to 8 forPECUU�PPP8. By comparing PECUU�PEO8 andPECUU�PPP8, both if which had the same PTMC length,it was found that the PEO-containing PECUU had a lowerglass transition temperature (�47.7 �C) than the PEO–

1000150020002500300035004000

Wavelength (cm-1)

-OCO-O

-C-O-C-

-NHCO-

-NHCNH-

O

O

-NHCO-

-NHCNH-

O

O

(1120)

(3340)

(1760)

(1550)

Fig. 2. FTIR spectrum of synthesized PECUUs. The wavenumbers forcharacteristic groups are presented in the parentheses.

PPO–PEO-containing PECUU. All of the polymers exceptfor PECUU–PPP8 exhibited a melting temperature of 18 �C.

The thermal behaviors for PECUU–PPP8 and PECUU–PEO8 were further characterized in the wet state over a tem-perature range of �10 to 90 �C (Fig. 4). This was to investi-gate whether PECUU–PPP8 inherits thermosensitivity fromthe thermosensitive PEO–PPO–PEO. The peaks at 0 �C wereattributed to the ice melting resulting from the water that hadbeen absorbed in the polymer bulk. No other transition wasseen for the PEO-containing PECUU–PEO8. In contrast,there was a transition at 37 �C for the PEO–PPO–PEO-con-taining PECUU–PPP8, which is similar to the reversiblethermal gelling temperature of PEO–PPO–PEO [32].

3.3. Swelling properties

The swelling properties characterized by water absorp-tion were measured for PECUUs as they reflect polymerbulk hydrophilicity and are related to degradation proper-ties [5,27]. Fig. 5 exhibits the water absorption of differentPECUUs at 4 and 37 �C. All of the PEO–PPO–PEO-con-

PECUU-PEO8

37 C

-10 0 10 20 30 40 50 60 70 80 90

Temperature ( oC)

PECUU-PPP8

End

othe

rmal

o

Fig. 4. DSC heating curves of water-impregnated PECUUs. The PEC-UUs were immersed in water overnight before being tested. The polymerswere tested in a temperature range of �10 to 90 �C, with a heating rate of10 �C min�1.

0

20

100

120

140

Wat

er c

onte

nt (

%)

4 oC 37 oC

PECUU-PPP8 PECUU-PEO8PECUU-PPP4

110

130

10

Fig. 5. Water content of PECUUs at different temperatures (4 and 37 �C).All values are means ± SD, n = 7.

22 oC, air

37 oC, water

0

4

8

12

16

20

0 200 400 600 800

Tensile strain (%)

Tens

ile s

tres

s (M

Pa)

Fig. 6. Typical stress–strain curves of PECUU–PPP8 tested at roomtemperature in air and 37 �C in water. For testing in the water, sampleswere soaked in 37 �C water overnight before the test.

F. Wang et al. / Acta Biomaterialia 5 (2009) 2901–2912 2907

taining PECUUs demonstrated significantly higher waterabsorption at 4 �C than at 37 �C (p < 0.01), presumablyresulting from thermosensitivity of the PEO–PPO–PEOin the soft segment. In contrast, the PECUU containingPEO presented higher water absorption at 37 �C than at4 �C (p < 0.01). In comparing polymers with the samePTMC length, the PEO-containing PECUU had more than10 times the water absorption of the PEO–PPO–PEO-con-taining PECUU (p < 0.001). For the PEO–PPO–PEO-con-taining PECUUs, the increase in PTMC lengthsignificantly decreased the water content at 4 �C(p < 0.05), but there was no significant difference at 37 �C(p > 0.05).

3.4. Mechanical properties

The mechanical properties of PECUUs tested at roomtemperature in air and at 37 �C in water are presented inTable 2. The typical stress–strain curves of PECUU–PPP8 at these two testing conditions are shown in Fig. 6.All of the polymers were found to possess higher mechan-ical properties at room temperature in air than at 37 �C inwater (p < 0.05). At room temperature in air, the PEO-con-taining PECUUs had significantly lower tensile strength,breaking strain and modulus than the PEO–PPO–PEO-containing PECUUs (p < 0.05) when they had the samePTMC length. The PEO–PPO–PEO-containing PECUUswere highly flexible and soft, with breaking strains rangingfrom 363 to 711%, tensile stresses ranging from 8.1 to

Table 2Mechanical properties of PECUUs tested at different conditions.

Breaking strain (%) Tensi

22 �C, air 37 �C, water 22 �C

PECUU–PPP4 363 ± 61 61 ± 9 8.1 ±PECUU–PEO8 53 ± 3 – 1.8 ±PECUU–PPP8 711 ± 75 127 ± 19 17.9 ±

17.9 MPa and moduli ranging from 5.5 to 7.4 MPa. Theincrease in PTMC length significantly increased the tensilestrength, breaking strain and modulus (p < 0.05).

When the mechanical testing was performed in water at37 �C, the PEO-containing PECUU could not be manipu-lated due to its high water content and extremely lowmechanical properties. For the PEO–PPO–PEO-containingPECUUs, the mechanical properties also significantlydecreased, with breaking strains ranging from 61 to127%, tensile stresses ranging from 1.7 to 3.1 MPa andmoduli ranging from 3.4 to 4.0 MPa. As was found inair, the PECUU with higher PTMC length possessedhigher mechanical properties than that with shorter PTMClength (Table 2).

3.5. Polyurethane degradation

The in vitro degradation of PECUUs in PBS is pre-sented in Fig. 7. All of the PECUUs demonstrated gradualweight loss during the 6 week degradation period, with theweight remaining ranging from 98 to 68%, depending onthe PTMC length and type of polyether (PEO or PEO–PPO–PEO) used. The PEO-containing PECUU had aweight loss of 32% at day 42, which was much higher thanfor the PEO–PPO–PEO-containing PECUUs (p < 0.05),the weight losses of which ranged from 2 to 5%. For thePEO–PPO–PEO-containing PECUUs, the increase inPTMC length was found to slightly decrease mass loss over6 weeks (p < 0.05).

le stress (MPa) Initial modulus (MPa)

, air 37 �C, water 22 �C, air 37 �C, water

0.2 1.7 ± 0.2 5.5 ± 0.9 3.4 ± 0.20.6 – 3.5 ± 1.4 –2.9 3.1 ± 0.4 7.4 ± 1.2 4.0 ± 0.3

60

65

70

75

80

85

90

95

100

0 7 14 21 28 35 42

Degradation time (days)

Wei

ght

rem

aini

ng (

%)

PECUU-PEO8PECUU-PPP4

PECUU-PPP8

Fig. 7. Weight remaining of PECUUs after degradation in PBS (pH 7.4)at 37 �C. All values are means ± SD, n = 4.

0

100

200

300

400

0 2-week 4-week 8-week

Bre

akin

g st

rain

(%

)

PECUU-PPP4

PECUU-PPP8

0

1

2

3

4

0 2-week 4-week 8-week

Tens

ile s

tres

s (M

Pa)

PECUU-PPP4

PECUU-PPP8

4

5PECUU-PPP4

PECUU-PPP8

(a)

(b)

(c)

2908 F. Wang et al. / Acta Biomaterialia 5 (2009) 2901–2912

The mechanical properties after degradation were char-acterized for the PEO–PPO–PEO-containing PECUUs asshown in Figs. 8 and 9. The mechanical testing was con-ducted in a 37 �C water bath. The stress–strain curves dem-onstrated that the polymers retained flexibility during the8 week degradation period. For both polymers, the break-ing strains increased significantly at week 8 compared withthe corresponding original polymers (p < 0.05, Fig. 9a).Conversely, the tensile stresses did not change significantly(p > 0.1, Fig. 9b). Moreover, the polymers retained theirmoduli after 8 weeks of degradation (Fig. 9c).

PECUU–PPP8 and PECUU–PEO8 were chosen for fur-ther characterizing their inherent viscosity change duringthe degradation (Fig. 10). For PECUU–PPP8, the viscosityat 0, 4 and 8 weeks did not change significantly (p > 0.05).The viscosity of PECUU–PEO8 after 4 weeks of degrada-tion remained the same as the starting polymer

0

1

2

3

4

0 50 100 150 200 250 300 350

Tensile strain (%)

Tens

ile s

tres

s (M

Pa) original

2-week

4-week8-week

Fig. 8. Typical stress–strain curves for PECUU–PPP4 and PECUU–PPP8after 0, 2, 4 and 8 weeks of degradation. Samples were tested in a 37 �Cwater bath. All of the samples were immersed in the 37 �C water overnightbefore being tested.

0

1

2

3

0 2-week 4-week 8-week

Init

ial m

odul

us (

MP

a)

Fig. 9. Mechanical properties of PECUU–PPP4 and PECUU–PPP8 afterdegradation. (a) Breaking strain; (b) tensile stress; and (c) initial modulus.Samples were tested in a 37 �C water bath. All values are means ± SD,n = 4.

(0.98 dl g�1), whereas it decreased to 0.52 dl g�1 after8 weeks of degradation (p < 0.05 between 0 and 8 weeks,and between 4 and 8 weeks).

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0 2 4 6 8

Degradation time (week)

Inhe

rent

vis

cosi

ty (

dL/g

)

PECUU-PPP8

PECUU-PEO8

Fig. 10. Inherent viscosities of PECUUs after degradation. All values aremeans ± SD, n = 4.

402 400 398 396

Bonding energy (eV)

RGDS modified film

Plasma treated film

Untreated film

Fig. 12. XPS spectra (N1s) of PECUU films before and after surfacemodification.

F. Wang et al. / Acta Biomaterialia 5 (2009) 2901–2912 2909

3.6. Cytotoxicity of degradation products

SMCs were cultured in culture medium supplementedwith the 8 week PECUU–PPP4 degradation solution toassess cytotoxicity of the degradation products. Fig. 11shows that the viability of the cells cultured in the mediumcontaining degradation products was not significantly dif-ferent to that of cells cultured in the control medium after2 and 4 days of culture (p > 0.1 for each time point). Fur-thermore, the cells cultured in the medium containing deg-radation products exhibited a similar morphology to thosecultured in the control medium (not shown).

3.7. Surface modification

Surface modification was conducted to immobilize thecell adhesive peptide RGDS on the polymer surface in

0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

Control medium Medium containingdegradation products

MT

T a

bsor

ptio

n

2 days

4 days

Fig. 11. Cytotoxicity of degradation products. Cell viability after 2 and4 days of culture in regular medium and degradation products containingmedium. All values are means ± SD, n = 8.

order to improve cell adhesion. XPS was used to indirectlyconfirm RGDS immobilization [33]. Fig. 12 demonstratesthat the nitrogen intensity of the RGDS-modified surfacewas apparently higher than that of untreated and O2

plasma-treated surfaces. The ratio of N1s/C1s was usedto characterize the relative nitrogen content. The untreatedand O2 plasma-treated surfaces had N1s/C1s ratios of 1.2and 1.3%, respectively. RGDS modification increased theratio to 2.2%.

3.8. Smooth muscle cell culture

SMC adhesion on the RGDS-modified and unmodifiedPECUU surfaces after 24 h is shown in Fig. 13a. Consider-ing the adherent cell density on the TCPS to be 100%, thecell adhesions on PECUU–PPP4 and PECUU–PPP8 were109 and 116%, respectively, and did not significantly varyfrom TCPS. The surface O2 plasma treatment did notimprove cell adhesion significantly (p > 0.05) as comparedwith unmodified counterparts. Surface RGDS immobiliza-tion markedly improved cell adhesion to over 180% onPECUU–PPP8, which was significantly higher than theO2-treated (p < 0.05), untreated (p < 0.05) and TCPS(p < 0.01) surfaces.

SMC viability change during the culture on RGDS-modified and unmodified PECUU–PPP8 surfaces is shownin Fig. 13b. SMCs were found to increase viability during a7 day culture period. At days 3 and 7, cell viability on theunmodified PECUU–PPP8 surface was similar to that onthe TCPS surface. The RGDS-modified surface demon-strated greater cell viability than the unmodified surfaceat day 3 (p < 0.05), but not at day 7 (p > 0.05). Cells formedconfluent layers on both the RGDS-modified and unmod-ified surfaces at day 7 (not shown).

4. Discussion

The purpose of this work was to develop low moduliand flexible polymers that could be used to fabricate scaf-

0

50

100

150

200

250

TC

PS

PE

CU

U-P

PP

4

PE

CU

U-P

PP

8

PE

CU

U-P

PP

8 M

odif

ied

wit

h O

2 pl

asm

a

PE

CU

U-P

PP

8-R

GD

S

Cel

l adh

esio

n (%

TC

PS)

0

0.4

0.8

1.2

1.6

2.0

TCPS PECUU-PPP8-RGDS

MT

T a

bsor

ptio

n

1 day 3 days 7 days

PECUU-PPP8

(a)

(b)

Fig. 13. Smooth muscle cell adhesion and viability change on the PECUUsurfaces before and after RGDS surface modification. (a) Cell adhesion(cell adhesion was normalized by the MTT absorption on TCPS surface);(b) cell viability change. All values are meanss ± SD, n = 4.

2910 F. Wang et al. / Acta Biomaterialia 5 (2009) 2901–2912

folds for engineering low moduli soft tissues, such as bloodvessels and cardiac muscle. A group of polyurethanes weresynthesized to meet these criteria by using block copoly-mers of either PEO–PPO–PEO or PEO flanked with lowmodulus PTMC as a soft segment, with BDI and putres-cine as a hard segment and a chain extender, respectively.The chemical and physical properties of the polymers werelargely dependent upon the soft segment chemistry, includ-ing the types of polyether (PEO–PPO–PEO or PEO) andlength of PTMC.

The generated polymers had glass transition tempera-tures lower than �46 �C (Fig. 3), allowing them to be ina rubbery state at ambient and physiological temperatures.These glass transitions were attributed to the soft segments.None of the polymers showed any evidence of hard seg-ment transitions, suggesting that the degree of phase sepa-ration between the soft and hard segments in the PECUUs

was low [29]. The glass transition temperatures were foundto increase with PTMC length. This increase may be due toincreased phase separation, leading the polymers to exhibitcharacteristics more like those of pure PTMC. This is con-sistent with a previous study which demonstrated that theincrease in PTMC length in PTMC–PEO–PTMC copoly-mers led to an increase in glass transition temperatures[34]. When the PTMC length was the same, the PEO–PPO–PEO-containing PECUU had a lower glass transitiontemperature than the PEO-containing PECUU, indicatingthat the former had a higher degree of phase separation.This may have resulted from the steric hindrance by themethyl groups in PPO decreasing hydrogen-bond forma-tion between the soft and hard segments [35]. OnlyPECUU–PPP8 showed a melting temperature in the drystate with a small peak at 18 �C. There was no such peakfor PECUU–PEO8, which had the same PTMC length(Fig. 3). This melting peak may be attributed to PTMC,and might be explained by more flexible PEO–PPO–PEOblocks providing conformational freedom for neighboringPTMC blocks to relax and crystallize [27]. High-molecu-lar-weight PTMC is generally considered to be amorphous;however, low-molecular-weight PTMC tends to be crystal-line [36].

The thermal properties were further characterized inwater as this better reflects the physiological environment.Compared to the PEO-containing PECUU–PEO8, thePEO–PPO–PEO-containing PECUU–PPP8 was found tohave a transition at 37 �C (Fig. 4). Furthermore, therewas no melting temperature of 18 �C as was seen in thedry state. The lack of this melting temperature is presum-ably due to water penetration into the hydrophilic PEO–PPO–PEO hindering PTMC crystallization. Since PEO–PPO–PEO is thermosensitive and PECUU–PEO8 dis-played no transition at 37 �C, the transition at 37 �C maybe attributed to the thermosensitive PEO–PPO–PEO,which has a gelling transition temperature around 37 �C[32]. These data demonstrate that the synthesized PEO–PPO–PEO-containing PECUUs were thermosensitive,whereas the PEO-containing PECUUs were not. Thiswas further confirmed by the water contents of the poly-mers, where the PEO–PPO–PEO-containing PECUUsshowed higher water contents at 4 �C than at 37 �C(Fig. 5). In contrast, the PEO-containing PECUUs hadhigher water contents at 37 �C than at 4 �C. These thermo-sensitive PECUUs may find potential applications in drugdelivery, where hydrophilic and low-molecular-weightdrugs could be absorbed into the polymer bulk at low tem-perature (for example 4 �C) and be released at 37 �C. Anexample of such an application has been demonstrated byLoh et al. [37–39], who used thermosensitive polyurethanesfor programmed drug delivery controlled by temperature.

The PECUUs were flexible and soft at room tempera-ture in air, with moduli ranging from 3.5 to 7.4 MPa(Fig. 6 and Table 2). These moduli were much lower thanthose of the polycaprolactone-based polyurethanes onwhich we have reported previously, where most of the poly-

F. Wang et al. / Acta Biomaterialia 5 (2009) 2901–2912 2911

mers exhibited moduli greater than 17 MPa [29]. The PEO-containing PECUUs were much weaker and less flexiblethan the PEO–PPO–PEO-containing PECUUs. This isattributed to strain crystallization of the soft segment asthe PEO–PPO–PEO led to a higher degree of phase separa-tion relative to PEO. The decreased hydrogen-bond inter-actions between the soft and hard segments wouldprovide more conformational freedom for the PTMCblocks to crystallize during stretching. PTMC-based poly-urethanes were found to undergo strain crystallization,with the polymers displaying similar tensile stress–straincurves to what is shown in Fig. 6 [25]. For PEO–PPO–PEO-containing PECUUs, the tensile strength and modu-lus increased with PTMC length. This increase may bethe result of increased soft segment interaction andincreased strain crystallization as the longer PTMC canbe more readily to crystallize [36].

All of the PECUUs were found to have lower mechanicalproperties (tensile stress, strain and modulus) at 37 �C inwater than at room temperature in air (Table 2). This wasexpected as water entering the PECUUs decreased interac-tions among the polymer chains, and polymer chain disen-tanglement became easier at elevated temperatures. At37 �C in water, the PEO-containing PECUU–PEO8 lost flex-ibility and was not manipulable mainly due to its high watercontent. The PEO–PPO–PEO-containing PECUUsremained flexible and soft, exhibiting breaking strains of61–127%, maximum tensile stresses of 1.7–3.1 MPa andmoduli ranging from 3.4 to 4.0 MPa. These low moduli poly-mers would be attractive for engineering low moduli soft tis-sues such as cardiac muscle (<0.5 MPa) and blood vessels(�1.4 MPa) when processed into porous scaffolds, as thescaffolds are expected to have lower stiffness than the originalpolymers [21].

In vitro biodegradation studies demonstrated that PEO-containing PECUUs degraded significantly faster thanPEO–PPO–PEO-containing PECUUs (Fig. 7). This canbe explained by the higher water content in the PEO-con-taining PECUUs and is consistent with earlier studieswhere the degradation rates were demonstrated to increasewith water content in polyurethanes [5,12,27]. Interestingly,the tensile stresses and moduli of the PEO–PPO–PEO-con-taining PECUUs did not change significantly during degra-dation (Figs. 8 and 9). Furthermore, viscosities at differenttime points remained the same (Fig. 10). The degradationproperties of these polyurethanes are different from thoseof the polycaprolactone-based polyurethanes, the mechan-ical properties and molecular weights of which were foundto decrease during degradation [5,12,40]. These polyure-thanes would be attractive for engineering mechano-activesoft tissues as they would provide prolonged mechanicalsupport to the developing tissue and allow extended trans-fer of mechanical stimuli during tissue development, ulti-mately forming mechanically functional tissues [5,12].

A common concern in the design of biodegradable poly-urethanes is cytotoxicity. The BDI- and putrescine-derivedpolymer segments in this work should ultimately degrade

into putrescine, an essential polyamine associated with cellgrowth and differentiation [5]. The use of PEO or PEO–PPO–PEO is not expected to introduce cytotoxicity as theyhave established biocompatibility [32]. PTMC degradesinto biocompatible 1,3-trimethylene glycol. Supportingthese expectations, cytotoxicity tests using the 8 week deg-radation solution from PECUU demonstrated the absenceof any detectable cytotoxic effect on SMCs in a basic viabil-ity test (Fig. 11). SMC adhesion on the unmodified PECU-Us was comparable to that on the TCPS (Fig. 13a).Tailored PTMC length did not change cell adhesion. Sur-face modification with O2 plasma treatment followed byRGDS immobilization significantly improved cell adhe-sion. RGDS modification was also found to improve cellcoverage on the PECUU surface, and the cell viability ateach time point was significantly higher than that ofunmodified surfaces (Fig. 13b).

In summary, we have synthesized a family of flexibleand low moduli PECUUs. Polymer chemical and physicalproperties could be altered by manipulating soft segmentchemistry, and those polymers containing PEO–PPO–PEO were found to be thermosensitive. Polymers withPEO–PPO–PEO soft segments appeared to undergo sur-face erosion and maintained their mechanical propertiesduring degradation. Degradation products were shown tobe non-cytotoxic and the surfaces could be modified withRGDS peptide to improve cell adhesion. These polyure-thanes may find applications in engineering low moduli,mechano-active cardiovascular or other soft tissues.

References

[1] Levental I, Georges PC, Janmey PA. Soft biological materials andtheir impact on cell function. Soft Matter 2007;3:299–306.

[2] Gomillion CT, Burg KJL. Stem cells and adipose tissue engineering.Biomaterials 2006;27:6052–63.

[3] Langer R, Vacanti JP. Tissue engineering. Science 1993;260:920–6.[4] Nasseri BA, Ogawa K, Vacanti JP. Tissue engineering: an evolving

21st-century science to provide biologic replacement for reconstruc-tion and transplantation. Surgery 2001;130:781–4.

[5] Guan JJ, Sacks MS, Beckman EJ, Wagner WR. Biodegradablepoly(ether ester urethane)urea elastomers based on poly(ether ester)triblock copolymers and putrescine: synthesis, characterization andcytocompatibility. Biomaterials 2004;25:85–96.

[6] Guan JJ, Stankus JJ, Wagner WR. Biodegradable elastomericscaffolds with basic fibroblast growth factor release. J ControlRelease 2007;120:70–8.

[7] Webb AR, Yang J, Ameer GA. Biodegradable polyester elastomers intissue engineering. Expert Opin Biol Ther 2004;4:801–12.

[8] Amsden B. Curable, biodegradable elastomers: emerging biomaterialsfor drug delivery and tissue engineering. Soft Matter 2007;3:1335–48.

[9] Tiwari A, Salacinski H, Seifalian AM. New prostheses for use inbypass grafts with special emphasis on polyurethanes. CardiovascSurg 2002;10:191–7.

[10] Buma P, Ramrattan NN, van Tienen TG, Veth RPH. Tissueengineering of the meniscus. Biomaterials 2004;25:1523–32.

[11] Giraud MN, Armbruster C, Carrel T, Tevaearai HT. Current state ofthe art in myocardial tissue engineering. Tissue eng 2007;13:1825–36.

[12] Guan JJ, Fujimoto KL, Sacks MS, Wagner WR. Preparation andcharacterization of highly porous, biodegradable polyurethane scaf-folds for soft tissue applications. Biomaterials 2005;26:3961–71.

2912 F. Wang et al. / Acta Biomaterialia 5 (2009) 2901–2912

[13] Cohn D, Hotovely-Salomon A. Designing biodegradable multiblockPCL/PLA thermoplastic elastomers. Biomaterials 2005;26:2297–305.

[14] Tseng SJ, Tang SC. Synthesis and characterization of the noveltransfection reagent poly(amino ester glycol urethane). Biomacro-molecules 2007;8:50–8.

[15] Kavlock KD, Pechar TW, Hollinger JO, Guelcher SA, Goldstein AS.Synthesis and characterization of segmented poly(esterurethane urea)elastomers for bone tissue engineering. Acta Biomater2007;3(4):475–84.

[16] Skarja GA, Woodhouse KA. Structure–property relationships ofdegradable polyurethane elastomers containing an amino acid-basedchain extender. J Appl Polym Sci 2000;75(12):1522–34.

[17] Stankus JJ, Soletti L, Fujimoto K, Hong Y, Vorp DA, Wagner WR.Fabrication of cell microintegrated blood vessel constructs throughelectrohydrodynamic atomization. Biomaterials 2007;28:2738–46.

[18] Fujimoto KL, Tobita K, Merryman DW, Guan JJ, Momoi N, StolzDB, et al. Biodegradable cardiac patch induces contractile smoothmuscle and prevents cardiac remodeling in sub-acute myocardialinfarction. J Am Coll Cardiol 2007;49:2292–300.

[19] Courtney T, Sacks MS, Stankus J, Guan JJ, Wagner WR. Analysisand design of tissue engineered scaffolds that mimic soft tissuemechanical anisotropy. Biomaterials 2006;27(19):3631–8.

[20] Guelcher SA. Biodegradable polyurethanes: synthesis and applica-tions in regenerative medicine. Tissue Eng Part B Rev 2008;14:3–17.

[21] Chen QZ, Bismarck A, Hansen U, Junaid S, Tran MQ, Harding SE,et al. Characterisation of a soft elastomer poly(glycerol sebacate)designed to match the mechanical properties of myocardial tissue.Biomaterials 2008;29:47–57.

[22] Zhu KJ, Hendren RW, Jensen K, Pitt CG. Synthesis, properties, andbiodegradation of poly (1, 3-trimethylene carbonate). Macromole-cules 1991;24:1736–40.

[23] Pego AP, van Luyn MJA, Brouwer LA, van Wachem PB, Poot AA,Grijpma DW, et al. In vivo behavior of poly(1, 3-trimethylenecarbonate) and copolymers of 1, 3-trimethylene carbonate with D, L-lactide or epsilon-caprolactone: degradation and tissue response. JBiomed Mater Res A 2003;67A:1044–54.

[24] Storey RF, Hickey TP. Degradable polyurethane networks based onD,L-lactide, glycolide, epsilon-caprolactone, and trimethylene carbon-ate homopolyester and copolyester triols. Polymer 1994;35:830–8.

[25] Asplund B, Bowden T, Mathisen T, Hilborn J. Synthesis of highlyelastic biodegradable poly(urethane urea). Biomacromolecules2007;8:905–11.

[26] Asplund B, Aulin C, Bowden T, Eriksson N, Mathisen T, BjurstenLM, et al. In vitro degradation and in vivo biocompatibility study ofa new linear poly(urethane urea). J Biomed Mater Res B2008;86B:45–55.

[27] Guan JJ, Wagner WR. Synthesis, characterization and cytocompat-ibility of polyurethaneurea elastomers with designed elastase sensi-tivity. Biomacromolecules 2005;6:2833–42.

[28] Tanzi MC, Verderio P, Lampugnani MG, Resnati M, Dejana E,Sturani E. Cytotoxicity of some catalysts commonly used in thesynthesis of copolymers for biomedical use. J Mater Sci Mater Med1994;5:393–6.

[29] Guan JJ, Sacks MS, Beckman EJ, Wagner WR. Synthesis, charac-terization, and cytocompatibility of elastomeric, biodegradablepoly(ester-urethane)ureas based on poly(caprolactone) and putres-cine. J Biomed Mater Res 2002;61:493–503.

[30] Ray JL, Leach R, Herbert JM, Benson M. Isolation of vascularsmooth muscle cells from a single murine aorta. Methods Cell Sci2002;23:185–8.

[31] Ruckenstein E, Yuan YM. Molten ring-open copolymerization of L-lactide and cyclic trimethylene carbonate. J Appl Polym Sci1998;69:1429–34.

[32] Alexandridis P, Hatton TA. Poly(ethylene oxide)–poly(propyleneoxide)–poly(ethylene oxide) block copolymer surfactants in aqueoussolutions and at interfaces: thermodynamics, structure, dynamics, andmodeling. Colloid Surf A 1995;96:1–46.

[33] Santiago LY, Nowak RW, Rubin JP, Marra KG. Peptide-surfacemodification of poly(caprolactone) with laminin-derived sequencesfor adipose-derived stem cell applications. Biomaterials2006;27:2962–9.

[34] Wang H, Dong JH, Qiu KY. Synthesis and characterization of ABA-type block copolymer of poly(trimethylene carbonate) with poly(eth-ylene glycol): bioerodible copolymer. J Polym Sci A: Polym Chem1998;36:695–702.

[35] Gorna K, Gogolewski S. In vitro degradation of novel medicalbiodegradable aliphatic polyurethanes based on epsilon-caprolactoneand Pluronics� with various hydrophilicities. Polym Degrad Stab2002;75:113–22.

[36] Zhang Z, Grijpma DW, Feijen J. Thermo-sensitive transition ofmonomethoxy poly(ethylene glycol)-block-poly(trimethylene carbon-ate) films to micellar-like nanoparticles. J Control Release2006;112:57–63.

[37] Loh XJ, Sng KBC, Li J. Synthesis and water-swelling of thermo-responsive poly(ester urethane)s containing poly(epsilon-caprolac-tone), poly(ethylene glycol) and poly(propylene glycol). Biomaterials2008;29:3185–94.

[38] Loh XJ, Tan YX, Li ZY, Teo LS, Goh SH, Li J. Biodegradablethermogelling poly(ester urethane)s consisting of poly(lactic acid) –thermodynamics of micellization and hydrolytic degradation. Bio-materials 2008;29:2164–72.

[39] Loh XJ, Goh SH, Li J. Hydrolytic degradation and protein releasestudies of thermogelling polyurethane copolymers consisting ofpoly[(R)-3-hydroxybutyrate], poly(ethylene glycol), and poly(propyl-ene glycol). Biomaterials 2007;28:4113–23.

[40] Rothstein SN, Federspiel WJ, Little SR. A unified mathematicalmodel for the prediction of controlled release from surface and bulkeroding polymer matrices. Biomaterials 2009;30(8):1657–64.