Materials Science and Engineering C 49 (2015) 612–622

Contents lists available at ScienceDirect

Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec

Mechanical and thermal property characterization of poly-L-lactide (PLLA) scaffold developed using pressure-controllable green foaming technology Shen-Jun Sheng a,b, Xiao Hu c,d, Fang Wang a,c,⁎, Qing-Yu Ma e, Min-Fen Gu a a

Center of Analysis and Testing, Nanjing Normal University, Nanjing 210023, China School of Chemistry and Materials Science, Nanjing Normal University, Nanjing 210023, China c Department of Physics and Astronomy, Rowan University, Glassboro, NJ 08028, USA d Department of Biomedical and Translational Sciences, Rowan University, Glassboro, NJ 08028, USA e Key Laboratory of Optoelectronics of Jiangsu Province, School of Physics and Technology, Nanjing Normal University, Nanjing 210023, China b

a r t i c l e

i n f o

Article history: Received 10 September 2014 Received in revised form 5 December 2014 Accepted 6 January 2015 Available online 8 January 2015 Keywords: Poly-L-lactide Gas foaming Glass transition temperature Rigid amorphous Mesophase Biodegradation

a b s t r a c t Poly-L-lactide (PLLA) is one of the most promising biological materials used for tissue engineering scaffolds (TES) because of their excellent biodegradability and tenability. Here, microcellular PLLA foams were fabricated by pressure-controllable green foaming technology. Scanning electron microscopy (SEM), dynamic mechanical analysis (DMA), differential scanning calorimetry (DSC), wide angle X-ray diffraction measurement (WAXRD), thermogravimetric (TG) analysis, reflection-Fourier transform infrared (FTIR) analysis, enzymatic degradation study and MTT assay were used to analyze the scaffolds' morphologies, structures and crystallinities, mechanical and biodegradation properties, as well as their cytotoxicity. The results showed that PLLA foams with pore sizes from 8 to 103 μm diameters were produced when the saturation pressure decreased from 7.0 to 4.0 MPa. Through a combination of StepScan DSC (SSDSC) and WAXRD approaches, it was observed in PLLA foams that the crystallinity, highly-oriented metastable state and rigid amorphous phase increased with the increasing foaming pressure. It was also found that both the glass transition temperature and apparent enthalpy of PLLA significantly increased after the foaming process, which suggested that the changes of microcellular structure could provide PLLA scaffolds better thermal stability and elasticity. Moreover, MTT assessments suggested that the smaller pore size should benefit cell attachment and growth in the scaffold. The results of current work will give us better understanding of the mechanisms involved in structure and property changes of PLLA at the molecular level, which enables more possibilities for the design of PLLA scaffold to satisfy various requirements in biomedical and green chemical applications. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Recently, various biomaterials have been fabricated and developed to meet different biomedical applications in tissue engineering field. For example, scaffold has been used to replace diseased or damaged organs and to restore the functionality of these organs such as cartilage repair [1] and fibroblast adhesion [2]. The micro-porous and nano-sheet biomaterials can also be used in clinical application of drug delivery by loading growth factor [3], vaccines [4], enzyme entrapment [5], bioactive molecules [6], anticancer and antibiotic drug [7]. Poly-lactide (PLA), a linear aliphatic thermoplastic polyester originally obtained from natural crops [8], is one of the most promising biomaterials used for tissue engineering scaffolds (TES) because of their ⁎ Corresponding author at: Center of Analysis and Testing, Nanjing Normal University, Nanjing 210023, China. E-mail address: [email protected] (F. Wang).

http://dx.doi.org/10.1016/j.msec.2015.01.025 0928-4931/© 2015 Elsevier B.V. All rights reserved.

environment-friendly composition, tenability and biodegradability, etc. [9]. Lactic acid has two enantiomeric forms: L-lactide and D-lactide. Poly(L-lactide) (PLLA) and poly(D-lactide) (PDLA) are semicrystalline materials derived from L-lactide and D-lactide, respectively. In general, a desired foam scaffold suitable for biomedical applications should have tunable mechanical properties and excellent thermal stability [10,11]. Moreover, it is also critical that the pore size and the biodegradation rate of the foam scaffold can match up with the tissue regeneration process in vivo. Previous studies reported that the presence of small pores in the TES often significantly slowed down its tissue regeneration rate. However, the overlarge pore size not only reduced the mechanical stability and the strength of TES, but also affected the adhesion of regenerated cells on the scaffolds [11,12]. Therefore, controlling the pore size and mechanical properties of scaffolds is extremely important for fostering tissue regeneration in vivo [13–15]. By using a novel gas method, Shimbo et al. [16] produced polyethylene terephthalate (PET) sponges and Nadella et al. [17] fabricated

S.-J. Sheng et al. / Materials Science and Engineering C 49 (2015) 612–622

acrylonitrile butadiene styrene (ABS) foams for package and engineering applications. They found that the mechanical properties of microcellular foams have been significantly improved when compared with raw materials. In addition, many efforts were made to investigate the crystallization behavior [12,13], mechanical and thermal properties of polylactic acid (PLA) raw materials and mixed polymers [18,19], while little work was reported on the effects of microstructure on the crystallinity of fabricated PLA foams. Thus, in the present work, a solvent-free solid-state physical foaming technology was used to fabricate PLLA scaffolds with different microcellular sizes [20]. The scaffold morphologies such as averaged pore size, pore density and scaffold wall thickness were examined by scanning electron microscopy (SEM) [21]. The thermal stability of PLLA foams were measured by thermogravimetric (TG) analysis [22,23]. Reflection-Fourier transform (FTIR) [24] analysis was utilized to determine whether there was any functional group change in PLLA foams. The visco-elastic properties of PLLA foams were evaluated by measuring the storage modulus, the loss modulus and the loss angle using the dynamic mechanical analyzer (DMA) [18]. Furthermore, the crystalline structure and degree of the PLLA materials were determined by wide angle X-ray diffraction (WAXRD) measurements [25]. The changes of the glass transition temperature and melting process were observed by a differential scanning calorimeter (DSC) [26]. StepScan DSC (SSDSC) approach was used to characterize the heat capacity, phase contents and transitions of PLLA samples. Also, the porous scaffolding materials have been studied to ensure the localized and prolonged availability of genetic materials to cells [27,28] and to control the release of biomolecules (e.g., proteins) for proliferation and differentiation [29,30]. Here, PLLA scaffolds with a suitable degradation rate and good biocompatibility were designed to guide tissue cells to grow towards specific direction. Thus, the enzyme degradation and cytotoxicity assay of different PLLA scaffolds were also performed in the present work by using weight-loss method and MTT assay, respectively. The outcomes of above examinations not only provide new insights into all-green foaming methods for producing novel polymer materials, but also optimize the fabrication parameters to design the polymer scaffold with controllable microcellular structures, tunable mechanical properties and better thermal stability, which might improve new application for microcellular polymer materials in various areas in the future. 2. Experimental section 2.1. Materials Raw PLLA was obtained from Shenzhen Yisheng New Material Co. Ltd. (China). The density of the raw PLLA is 1.25 g·cm− 3. The radius and height of the cylindrical PLLA specimens are 2 mm and 3 mm, respectively. The melting and glass transition temperatures of the raw PLLA are in the ranges from 175 to 185 °C and from 60 to 65 °C, respectively. The carbon dioxide of medical purity was provided by the Nanjing Special Gas Co. Ltd. (China). 2.2. Preparation of microcellular foamed PLLA First of all, PLLA sheets were placed in an autoclave machine. CO2 was inflated for 15 min to drive away the air inside the autoclave. Then, the exhaust valve of the autoclave was closed and the samples were saturated with CO2 at the pre-selected pressures (e.g., 4.0 MPa, 5.0 MPa, 6.0 MPa and 7.0 MPa) at room temperature for 48 h. Once the PLLA samples moved from the autoclave to the atmospheric pressure, the desorption was carried out immediately. To produce microcellular foamed structures in PLLA, the CO2-saturated samples were foamed by immersing them in a water bath at the temperature of 90 °C. Details of the processing conditions can be found in our previous work [20]. Fig. 1 shows a schematic of foaming the PLLA foams and scaffolds by this solvent-free solid-state physical method.

613

Fig. 1. Schematic preparation of PLLA foams based on a solvent-free solid-state physical foaming method. (a): Raw PLLA materials and its molecular formula; (b): the preparation process of PLLA foam; (c): the formed PLLA foams. Significant size change of samples was observed due to the increase of the pore sizes in the PLLA scaffolds.

2.3. Methods 2.3.1. SEM observation SEM (JSM-5610LV, JEOL, Japan) was used to observe the cell microstructure of PLLA foams. Before the measurements, the samples were coated with gold to make them conductive. The pore size and the scaffold wall thickness were measured through SEM photographs by using the software of Smile View. The pore size of each sample was averaged for at least 100 pores. The average pore size, D, is calculated by the following equation: i X

di

i¼1

D¼

ð1Þ

i

where i is the number of pores with the pore size of di. The pore density, N (viz. the number of pores per unit volume of the PLLA foams), can be determined by the equation as follows: " N¼

nM A

#

2 3=2

Rv

ð2Þ

where n is the number of pores in the SEM micrograph, M is the magnification factor, A is the area of the micrograph (in cm2), and Rv is the volume expansion ratio of the PLLA foams, which can be obtained using the following equation: Rv ¼

ρo ρf

ð3Þ

614

S.-J. Sheng et al. / Materials Science and Engineering C 49 (2015) 612–622

where ρo and ρf are the densities of pure PLLA and PLLA foams. The density of pure PLLA foams ρf can be measured according to ASTMD792-00 [31].

with frequency ω; and K (ω) is a calibration factor with τ being a correction value used at the given conditions of the measurement, which was determined from the sapphire calibration scans.

2.3.2. FTIR analysis FTIR spectra were obtained using a FTIR spectrometer (NicoletNEXUS 670, USA). Spectra were recorded in the range of wave number from 4000 to 400 cm−1 with a resolution of 4000 cm−1, and 32 scans were applied for each measurement.

2.3.6. WAXRD analysis WAXRD analyses were performed by using a D/max 2500VL/PC diffractometer (Rigaku Corporation, Japan). Cu-Kα radiation was generated by the system with a graphite diffracted beam monochromator operated at 45 kV and 30 mA. Here, K = cos22θM, 2θM = 26.6°, where 2θM is the diffraction angle of monochromator. α-SiO2 was used as a standard sample to measure the width of samples in our experiment. Data were acquired in a 2θ scale from 2° to 40°. The relationship between crystallite size of the polymer material and the width of diffraction line is given by the Sherrer equation [34]:

2.3.3. DMA Samples with a dimension of 7.0 × 5.0 × 2.0 mm3 were subjected to a dynamic mechanical analysis using a DMA (Perkin-Elmer Diamond DMA, USA), which was calibrated by polymethyl methacrylate (PMMA) standards at the mode of compression. They were tested under a temperature range from 25 °C to 150 °C with a heating rate of 2 °C·min−1. The frequencies were fixed at 1 Hz. The glass transition temperature (Tg) of the tested samples was taken as the temperature at which the maximum of the tanδ peak was exhibited. 2.3.4. TG analysis Thermal properties of the microcellular foamed PLLA and pure PLLA were tested by using a Pyris 1 thermogravimetric analyzer (Pyris 1 TGA, Perkin-Elmer, USA), which was calibrated by aluminum, nickel and per allory standards. Samples of about 0.5 mg were heated from 25 °C to 550 °C at a rate of 10 °C·min−1 under nitrogen atmosphere (flow rate 50 mL·min−1).

Lhkl ¼ Kλ=β cosθ

where Lhkl is the crystallite size that is perpendicular to the lattice plane (hkl) (nm), λ is the wavelength of incident X-ray (nm), θ is the Bragg angle, β is the increment of pure diffraction ray (°), while K is often called the crystal shape factor, which is associated with the shape of mini-crystallite, coefficient of lattice plane, β and Lhkl. The value of K is 0.9 when β1/2 is defined as the full width half maximum. Therefore, Eq. (8b) can be written as L ¼ 0:9λ=β1=2 cosθ:

2.3.5. DSC analysis The microcellular foamed PLLA samples were sealed in Al-crucible pans and studied using a Diamond DSC (Perkin-Elmer, USA), which was calibrated by indium and sapphire standards. The samples were 10 mg–12 mg with an empty aluminum pan as the reference. The standard mode DSC scans were obtained by heating the samples from 25 °C to 190 °C at a heating rate of 10 °C·min−1 under nitrogen atmosphere (a flow rate of 30 mL·min−1). The SSDSC method was applied by a rate of 3 °C·step−1 at a heating rate of 5 °C·min− 1, with 1-min isothermal holding at each step. The glass transition temperature (Tg), melting temperature (Tm) and melting enthalpy (ΔHm) were calculated by the Perkin-Elmer Pyris software. The degree of crystallinity (XC) of semicrystalline polymer can be obtained from the following equation: X C ð%Þ ¼ ΔH m =ΔH f  100%

ð4Þ

where ΔHf is the enthalpy of fusion of a 100% crystalline sample melted from the full crystalline to the liquid state at the end of equilibrium melting temperature. Specifically, the melting enthalpy of 100% crystalline PLLA is reported to be 93 J·g−1 in previous literature [32]. While ΔHm is the measured heat of fusion of the semicrystalline polymer. The modulation of T S (t) with the amplitude ATs and the period p (ω = 2π / p) can be calculated as [33]: T S ðt Þ ¼ T 0 þ bqN½C S =K  þ ATs sinðωt−εÞ

ð5Þ

where ε is the phase shift related to the internal reference frequency, K is the Newton's law constant, C S is the heat capacity of the sample calorimeter, bqN is the underlying heating rate, and T0 is the starting temperature. So the apparent reversing heat capacity CP can be expressed by Eq. (6): h i C p ¼ bAϕ N=bATs Nω K ðωÞ

ð6Þ

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 þ τ 2 ω2

ð7Þ

K ðωÞ ¼

where bA ϕN is the amplitude of the heat-flow-rate in modulation cycle; bATsN is the modulation amplitude of the temperature (T S )

ð8aÞ

ð8bÞ

2.3.7. Enzyme degradation Individual PLLA scaffolds were placed in different vials containing 5 mL of Tris–HCl buffer (pH 8.6), 1 mg of proteinase K from Tritirachium album (Merck), and 1 mg of sodium azide (Fisher). Four replicated measurements were performed in different vials to determine the weight loss of each scaffold sample at a specified incubation time. The enzymatic degradation of the scaffolds was evaluated at 37 °C in a rotary shaker (100 rpm) for the time periods up to 5 days (5 d). The buffer/enzyme solution was replaced daily so that the enzyme activity remained at a desired level throughout the entire experiment. Then, the specimens were taken out and washed thoroughly with distilled water, and then dried in a vacuum at room temperature until the weights remained constancy. The samples without enzyme in Tris–HCl solution were also evaluated to provide the control data. 2.3.8. Cytotoxicity assessment Before cell culture, all PLLA scaffolds were sterilized by ultraviolet (UV) light for 1 h and then immersed in 70% ethanol for 10 min followed by washing the samples three times with sterile phosphate buffered saline (PBS) solution. A human cervical cancer cell line (HeLa cells) was seeded in Dulbecco's Modified Eagle Medium (DMEM, Sigma, USA) with the addition of 10% (v/v) fetal bovine serum (FBS). The cells were rinsed briefly with sterile PBS, trypsinized, and resuspended in the medium at a concentration of approximately 104 cells·cm−2. PLLA scaffolds fabricated at different pressures (e.g., from 4.0 MPa to 7.0 MPa) were placed in a 96-well tissue culture plate (TCP). Then, HeLa cells were plated in PLLA scaffolds, and incubated at 37 °C with 5% CO2 until reaching 80–90% cell confluency. Cell viability of different scaffolds was evaluated by MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) assay. After 24, 48, and 72 h of incubation, the media were replaced with PBS containing 5 mg·mL−1 MTT. Following another 4 h of incubation, the supernatant was removed and replaced by adding 150 μL dimethyl sulfoxide (DMSO). Then the plates were shaken for 10 min to ensure the complete dissolution of formazan before UV absorbance measurements at 490 nm on a BioTek

S.-J. Sheng et al. / Materials Science and Engineering C 49 (2015) 612–622

Immunoanalyzer (EL-x800). The measurements were repeated three times for each sample. Cell viability was calculated according to the following Eq. (9): D Viability % ¼ s  100% Dc

ð9Þ

where Ds is the absorbance of scaffold and Dc is the absorbance of control. Samples with culture medium only (viz., without scaffolds) were treated as controls to determine background absorbance. 3. Results and discussion 3.1. Morphological analysis SEM images of different PLLA foams are shown in Fig. 2. The surface morphologies of PLLA foams fabricated at 4.0 MPa and 5.0 MPa are demonstrated in Fig. 2a and c, respectively, and Fig. 2b and d shows their corresponding microstructures in the cross-section regions. It is clearly observed that the pore size of PLLA foam produced at 4.0 MPa is larger than the one made at 5.0 MPa. Some broken cavities can be also observed in the cross-section regions of all samples, which might result from the intensive shearing of the gas during the fast bubble growth [35]. Table 1 summarizes the structural characteristics of PLLA foams (e.g., the average pore size, the pore density, and the scaffold wall thickness) fabricated at different saturation pressures. It clearly shows that the scaffold wall thickness decreases with the increasing gas pressure, which suggests that the higher saturation pressure might help PLLA foam to stretch out much more easily. The average pore size and density are plotted as the function of the saturation pressure in Fig. 3. The error bars in Fig. 3 indicates the overall

615

Table 1 The characteristic parameters of PLLA foams fabricated at different saturation pressures (each data have an error bar less than 5%). Pressure (MPa) Pore size (μm) Pore density (pores · cm−3) Wall thickness (μm)

4.0 103 1.12 × 107 0.9

5.0 34 7.72 × 107 0.7

6.0 12 2.55 × 108 0.6

7.0 8 9.61 × 108 0.54

distribution of the measured data. It is observed in Fig. 3 that the average pore size of PLLA foam decreases significantly from 103 to 8 μm when the saturation pressure increases from 4.0 to 7.0 MPa. The relationship between the pore size D (μm) and the saturation pressure p (MPa) can be fitted by the following formula: D ¼ 13112 expð−p=0:82Þ þ 4:70

ð10Þ

which indicates that the pore size of the PLLA foam exponentially decreases with the increasing saturation pressure. On the other hand, the fitted curve between the pore density N (number of pores per cm3) and the saturation pressure p (MPa) can be represented as: LogðNÞ ¼ 0:64p þ 4:60:

ð11Þ

These results are consistent with previous observations [20], and the dependences of pore size and density on the saturation pressure suggest that the microcellular structures of PLLA foams can be manipulated by adjusting the saturation pressure at which the foams are fabricated. 3.2. FTIR analysis FTIR analyses were conducted to further determine if there was any structural change or molecular chain interaction during the PLLA

Fig. 2. SEM images of the PLLA foams. (a) and (c): the surface morphologies of PLLA foams at 4.0 MPa and 5.0 MPa, respectively; (b) and (d): the microstructures of PLLA foam cross-sections at 4.0 MPa and 5.0 MPa, respectively (the length of the scale bar represents 100 μm).

616

S.-J. Sheng et al. / Materials Science and Engineering C 49 (2015) 612–622

Fig. 3. Average pore size and pore density of PLLA foams plotted as the function of saturation pressure. The scatter markers represent the experimental data, while the dash and solid lines represent the fitted curves.

foaming process. As shown in Fig. 4A, all tested samples demonstrate significant absorption bands at 1087 cm− 1 (C\O antisymmetric stretching), 1454 cm− 1 (C\H deformation vibration), 1359 cm− 1, 1270 cm− 1, 1180 cm− 1 (C\O\C stretching), 1380 cm− 1 and 2940 cm− 1 (C\H stretching and bending of methyl groups), which agrees with previous reports [19,36,37]. However, compared with raw PLLA samples, the absorption band of C_O stretching of carbonyl group in PLLA foams shifts from 1750 cm− 1 to 1757 cm−1, and the corresponding absorption intensity is enhanced (Fig. 4B). Moreover, a new peak, which is not observed for raw PLLA material, appears at 1210.8 cm−1 for all PLLA foams (Fig. 4C). These results indicate that the foaming process might strengthen the C_O bond in PLLA material, and aggravate the molecular interactions in the C\O\C bond region as well [15,19]. In addition, after the foaming process, the intensity of FTIR spectra at some wave numbers, such as 1087 cm−1, 1359 cm− 1 and 1180 cm−1, gradually increases with the decreasing pore size of PLLA foams. These findings suggest that the molecular chain interactions in PLLA foams are significantly different from those in raw PLLA materials. 3.3. Mechanical properties E′ is the storage modulus of materials. Normally, materials with low storage modulus can be easily deformed when the load is applied on it. Thus, it is more desirable to fabricate materials with high storage modulus that exhibit great elasticity when loading at a specified temperature range. In the present work, the storage modulus of the PLLA foams is higher than that of the raw PLLA, and increases gradually with the increasing saturation pressure (see Table 2). It may be induced by the strengthened molecular chains after the foaming process, which enables molecular matrix of PLLA foams to exhibit greater elasticity [15, 38]. The E″ of visco-elastic solids measures the energy dissipates as heat representing the viscous portion. The ratio of energy lost to energy retained in the loading cycle, i.e. ratio of the E″/E′, is denoted as the lose factor (tanδ) to represent the visco-elastic properties of materials. Low value of tanδ means that, once the deformation is induced, the material will quickly recover to its original status. Fig. 5 shows the variation of tanδ of PLLA samples versus the change of temperature. There is no substantial difference between PLLA foams and raw PLLA materials at 37 °C, a simulated human body temperature. However, it is noticed that the peak height of tanδ tends to become smaller (from 0.54 to 0.22) along with the increasing of pressure, and the sharpest peak of tanδ is observed for raw PLLA materials. Lee et al. [38] pointed out that the peak height and the sharpness of tanδ were associated with the molecular chain motion. Therefore, the sharper and higher peak of tanδ

Fig. 4. The infrared spectra of (a) pure PLLA and PLLA foams from (b) 4.0 MPa, (c) 5.0 MPa, (d) 6.0 MPa, and (e) 7.0 MPa pressures. (A): Full FTIR spectra from 500 cm−1 to 3500 cm−1; (B): comparison of 1750 cm−1 peaks; (C): comparison of 1210.8 cm−1 peaks.

observed for raw PLLA indicates that there is almost no restriction to the chain motion in the raw PLLA, while the microcellular structures in PLLA foams may hinder the chain mobility and lead to reduced sharpness and peak height of tanδ. Thus, after the foaming process, the mechanical loss of PLLA foams might be reduced, while the elasticity of PLLA foams would be increased.

S.-J. Sheng et al. / Materials Science and Engineering C 49 (2015) 612–622 Table 2 The characteristic values of different PLLA samples by DMA (error bar is less than 5%).

617

Table 3 The characteristic temperatures of PLLA samples in TG curves.

PLLA

E′ (MPa)

tanδ

Tg (°C)

PLLA

Tonset (°C)

Tp (°C)

Raw 4.0 MPa 5.0 MPa 6.0 MPa 7.0 MPa

4.5 4.6 6.0 8.8 12.0

0.54 0.34 0.32 0.28 0.22

68.6 80.5 82.0 84.0 85.3

Raw 4.0 MPa 5.0 MPa 6.0 MPa 7.0 MPa

342.21 318.65 324.19 336.07 339.48

363.78 349.83 353.05 360.56 362.70

E′, tanδ are the values between the baseline to the maximum values (peak positions) of storage modulus and loss factor in the DMA curves, respectively; Tg is the corresponding peak temperatures in the tanδ curves.

Tonset is the initial decomposition temperature found in the thermogravimetric curve; Tp is the maximum temperature of the first derivative thermogravimetric curve. Each data have an error bar less than 5%.

In addition, the glass transition temperature (Tg), where molecules regain their mobility, can be determined based on DMA studies. Tg is directly related to the flexibility of the molecular chains. The greater the stiffness is, the higher the Tg is [39,40]. In the present work, the measured Tg of PLLA foams is higher than that of raw PLLA, and tends to shift to higher temperature with the increasing pressure. Similar phenomenon is observed for the rigidity of PLLA foams. DMA results demonstrate that the attraction force between the polymer molecules might be weakened with the increasing pore size, which would change the entanglement scheme of segment. Thus, the crosslink density, elasticity and toughness of foams could be improved by the increase in the saturation pressure, which is also consistent with previous reports [41,42].

3.5. Crystallinity and rigid amorphous structure

3.4. Thermal stability Only a one-stage weight loss was found during the thermal degradation of all the samples. The degradation regimes related to the decomposition of PLLA backbone are above 300 °C. Table 3 summarized the onset decomposition temperature (Tonset) and the temperature with the maximum mass loss rate (Tp) of PLLA samples, where Tonset and Tp are calculated by the first derivative of the TG curves. The result shows that, both the Tonset and Tp of PLLA foams increase with the increasing saturation pressure. For example, the percentages of increase in Tonset and Tp from 4.0 MPa to 7.0 MPa were approximately 6.54% and 3.68%, respectively. These results indicated that the increase in the fabrication pressure might provide greater strength and thermal stability to the microcellular PLLA foams. However, it should be noticed that, compared with microcellular foams, the raw PLLA material demonstrates the highest Tonset and Tp, which implies that the introduction of microcellular structure might weaken the thermal stability of the polymer by breaking the molecular bindings of PLLA chains during the foaming process. Some additional details will be discussed in Section of 3.5.

Fig. 5. DMA curves of PLLA samples. Tanδ curves of (a) pure PLLA, and PLLA foams produced from (b) 4.0 MPa, (c) 5.0 MPa, (d) 6.0 MPa, and (e) 7.0 MPa pressures.

During the foaming process, PLLA samples were produced under different gas pressures, while processing temperature was maintained at a constant value. Therefore, the pressure should play an important role in controlling the microcellular structure of PLLA samples. Since the polymer chains in the amorphous phase is highly disordered, when the polymer sample is treated under the conditions of high pressure and high temperature, the molecular network in the sample will extend, then partially crystallize, and finally form a foam material [10]. Zhang et al. [43] found that there was a mesophase existed in the PLA block domains when the copolymer of PLA-poly(ethylene glycol) was meltquenched. Ma et al. [44] investigated and calculated the rigid amorphous and mesophase fractions of electrospun PLA fiber by the wide angle X-ray scattering (WAXS) and DSC. They showed that the mesophase, which constituted oriented non-crystalline chains, behaved very similarly to the rigid amorphous phase. Therefore, the WAXRD method was also adopted in the current work to investigate the crystal structures of the raw PLLA material and PLLA foams, as well as their mesophase. By using the Peak Fit™ software, WAXRD intensity profiles of samples were deconvoluted based on the knowledge of the crystalline and amorphous scattering contributions. The typical peak fitted [44,45] and azimuthally-integrated WAXRD profiles are shown in Fig. 6A and B for raw PLLA samples and PLLA foams fabricated at 6.0 MPa, respectively. And the goodness of the fitted correlation, r2, ranges from 0.996 to 0.999. It is clearly shown in Fig. 6A that the raw PLLA demonstrates the maximum WAXRD intensity, which is approximately at 2θ = 16.5° and indicates that the amorphous phase is the dominated structure in the raw PLLA sample [46,47]. However, the foamed PLLA samples exhibit an evident crystallization with a sharp WAXRD peak around 16.5° (Fig. 6B), which can be contributed to the diffraction of (200) and (110) planes of the orthorhombic crystals of PLLA [48,49]. A minor diffraction can be observed at 2θ = 18.7° (as shown in Fig. 6B), which corresponds to the (203) plane [48,49]. These reflection indexes indicate that stable α-crystals were formed in the PLLA foam samples [50]. These three phases (crystal phase, mobile amorphous phase, and mesophase) can be found co-existed in all PLLA foams. And the contents of each component are calculated from the WAXRD data, and then listed in Table 4. When CO2 gas pressure increases from 4.0 to 7.0 MPa, the amorphous content of PLLA samples decreases from 0.66 to 0.50, while the crystallinity increases from 0.27 to 0.37 and the mesophase fraction increases from 0.07 to 0.13 (Table 4, XC-WAXRD, XMAF-WAXRD and XMES-WAXRD). It is believed that with the increase of the pressure, some polymer chains can be extended and acted as row nuclei (mesophase) during the foaming process [44]. Then the lamellar crystals would grow onto these row nuclei and the molecular orientation of crystals could be manipulated during the folding of polymer chains. Therefore, after the raw PLLA sample was treated under the foaming process, some parts of the amorphous structures in the raw PLLA changed to crystals, and its mesophase fraction was also raised [45]. Besides, the increment of crystallinity also makes the molecular chains of PLLA foams become more ordered and rigid (E′ increased from 4.1 MPa to 12.0 MPa when pressure increased from 4.0 MPa to

618

S.-J. Sheng et al. / Materials Science and Engineering C 49 (2015) 612–622

researchers pointed out that the thermal stability of polymer sample could be affected by other factors, such as the conformations of segment groups [52,53], the physical structures and the strength of molecular chains [54], as well as the residues of small molecules and their molecular weights [52,55]. Therefore, it is speculated that the mesophase fraction (0.03) existing in PLLA raw samples has restricted the movement of polymer chain and might in turn affect the decomposition temperature of PLLA samples. Moreover, standard DSC scans of all PLLA samples are shown in Fig. 7A. In Fig. 7A, the inserted graph (I) shows the enlarged glass transition regions of PLLA foams between 60 °C and 70 °C, and the inserted graph (II) shows the enlarged melting region of pure PLLA between 100 °C and 180 °C. Three thermal events can be observed for the raw PLLA during the DSC analyses: a step change, a broad exothermic peak, and an endothermic peak, which are associated with the glass transition (Tg) region, the cold crystallization region, and the melting process (Tm) of PLLA, respectively. The values of Tg, Tm, and ΔHm of the melting samples are reported in Table 5. Furthermore, an endothermic peak is observed in the Tg region of the raw PLLA sample (curve a),

Fig. 6. Deconvolution of WAXRD peaks of azimuthally-integrated PLLA samples. (A): raw PLLA film; (B): PLLA foam samples formed in 6.0 MPa pressure. The solid curves in (a) and (b) are the experimental data (black) and the best fitted line (red). The fitted individual Gaussian peaks can be assigned into the crystal phase (c), the mesophase (d), and the amorphous phase (e). Miller indices for the crystallines were obtained from literature [48–50].

7.0 MPa). Simultaneously, TGA results show that the thermal decomposition of PLLA foams becomes more difficult. Hence, it suggests that the increase of both crystal and mesophase fractions might benefit the maintenance of the thermal stability of PLLA foams. However, some Table 4 The heat capacities at Tg, crystallinities, mobile amorphous fractions, and rigid amorphous (mesophase) fractions of PLLA foams fabricated at different gas pressures (with error bars less than 5%). Pressure (MPa) ΔCPa (J·g−1 °C−1) XC-DSCb XMAF-DSCb XRAF-DSCb XC-WAXRDc XMAF-WAXRDc XMES-WAXRDc a

0 0.59 0 0.97 0.03 0 0.97 0.03

4.0 0.40 0.26 0.66 0.08 0.27 0.66 0.07

5.0 0.36 0.31 0.59 0.10 0.32 0.59 0.09

6.0 0.33 0.34 0.54 0.12 0.33 0.56 0.11

7.0 0.29 0.39 0.47 0.14 0.37 0.50 0.13

The heat capacity increment of 100% amorphous PLLA at Tg is 0.61 [44,51]. b XC-DSC, XMAF-DSC and XRAF-DSC are the three phase fractions which are crystallinities (C), mobile amorphous fractions (MAF), and rigid amorphous fractions (RAF) obtained from DSC and Eqs. (4), (14a), (14b) and (14c), individually [13,44]. c XC-WAXRD, XMAF-WAXRD and XMES-WAXRD are the three phase fractions which are crystallinities (C), mobile amorphous fractions (MAF), and mesophase fractions (MES) obtained from WAXRD.

Fig. 7. DSC analysis of PLLA samples. (A): standard DSC heat flows of (a) pure PLLA and PLLA foams produced from (b) 4.0 MPa, (c) 5.0 MPa, (d) 6.0 MPa, and (e) 7.0 MPa pressures, the exothermic direction is downward. The inserted graphic (I) shows an enlarged glass transition region of PLLA foams between 60 °C and 70 °C (top left), and the inserted graphic (II) shows the enlarged melting region of pure PLLA between 100 °C and 180 °C (top right); (B): calculated specific reversing heat capacity curve of pure PLLA sample between 30 °C and 170 °C by SSDSC. The theoretical heat capacity baselines used for Cpliquid (blue dashed line) and Csolid (red dashed line) were obtained from ATHAS databank [56]. p

S.-J. Sheng et al. / Materials Science and Engineering C 49 (2015) 612–622 Table 5 The thermodynamic parameters of PLLA samples obtained by DSC.

CP

PLLA

Tg (°C)

Tm (°C)

ΔHm (J·g−1)

RAW 4.0 MPa 5.0 MPa 6.0 MPa 7.0 MPa

60.00 62.24 62.80 63.37 63.98

149.01 150.02 150.59 150.93 151.57

1.71 24.18 28.83 31.62 36.27

Tg, Tm, ΔHm are the glass transition temperature, melting temperature, and melting enthalpy measured by DSC, respectively.

which might be induced by the physical relaxation process of polymer chains when the sample was heated. However, the physical relaxation process does not exist in any PLLA foam samples, which might indicate that the foaming process would reduce the stress between the molecular chains in PLLA samples. It is also found that the Tg and Tm of PLLA foams increase slightly with the increasing foaming pressure, and the ΔHm values also become larger (see Table 5). Moreover, the shifting trend of Tg observed in DSC experiments is consistent with the one demonstrated in DMA experiments, which indicates that the foaming process might affect the chain mobility of PLLA samples in both low (DSC) and high (DMA) measuring frequencies. A three-phase polymer model at glass transition region, which contains crystalline (C), mobile amorphous fraction (MAF) and rigid amorphous fraction (RAF), has been used by many researchers [44,57,57] to observe other amorphous and semicrystalline polymers, such as PET films and electrospun PLA fibers. In this model, the molar fractions of these three phases can be determined by SSDSC [57,58], and a phase fraction equation can be written as: X C þ X MA F þ X RA F ¼ 1

ð12Þ

where XC, XMAF and XRAF are the crystal, mobile amorphous and rigid amorphous fractions of PLLA samples, respectively. Ma et al. [44] investigated the electrospun PLA fibers and studied their structural developments based on time-resolved WAXS. They proved that a similar three-phase model could be applied to electrospun PLA fibers. According to their studies, the chain conformation and chain packing in mesophase (rigid amorphous phase) should be originally disordered but rigid, and it would partially “melt” and become partially ordered crystal after being heated subsequently. Wunderlich [59] systematically studied the RAF in semicrystalline polymers, and revealed that RAF should be an amorphous phase with a degree of orientation in polymer samples. He also found that the rigid amorphous phase could not undergo relaxation during the glass transition, and could not contribute to the heat fusion of crystal melting. However, the glass transition of the MAF and the stability of molecular chains in polymer samples would be affected by RAF. In general, it is considered that the RAF exists between the crystal and amorphous phase due to the immobilization of a polymer chain [58–60]. And their fraction can be quantitatively determined by the heat capacity increment: X solid ¼ X C þ X RA F ¼ 1−ΔC P =ΔC P0

ð13Þ

where Xsolid is the fraction of solid phase in polymers which includes crystal and rigid amorphous, ΔCP0 is the heat capacity increments at the glass transition for 100% amorphous polymer, while ΔCP is the measured heat capacity change of mobile amorphous polymer at Tg. The ratio ΔCP/ΔCP0 is the amorphous chain fraction (XMAF) which contributes to the glass transition process. Then the temperature-dependence of PLLA heat capacity above Tg can be written as: CP

cal

ðT Þ ¼ X solid C P

solid

ðT Þ þ X MA F C P

liquid

CP

cal

ðT Þ ¼ X solid C P

solid

ðT Þ þ ð1−X solid ÞC P

ðT Þ liquid

ð14aÞ ðT Þ

ð14bÞ

cal

ðT Þ ¼ ðX C þ X RA F ÞC P

solid

619

ðT Þ þ ð1−X C −X RA F ÞC P

liquid

ðT Þ:

ð14cÞ

The theoretical data of solid heat capacity and liquid heat capacity of PLLA can be obtained from the ATHAS data bank [56]. The heat capacity increment of 100% amorphous PLLA at Tg is 0.61 [60]. SSDSC method not only can be used to precisely determine the heat capacity of PLLA samples [57,58], but also can be used to identify the devitrification of RAF in the sample. Schick et al. [58] investigated the RAF in nanocomposites of polymethyl methacrylate (PMMA) with silicon oxide nanoparticles by using the SSDSC method. They pointed out that the devitrification of RAF should be associated with the chain mobility of crystals and the restricted mobility must disappear before the RAF was devitrified, which indicated that the crystals of semicrystalline polymers should melt before the relaxation of RAF. Following this phase model, Fig. 7B shows the specific reversing heat capacity of the raw PLLA sample with the theoretical baselines Cliquid p (blue dashed line) and Csolid (red dashed line), which are obtained p from ATHAS databank [56]. When the heating temperature is raised from 30 °C and 170 °C, an endothermic shoulder firstly appears in the glass transition (Tg) region (50–65 °C). The endothermic shoulder results from the reversing heat capacity increment of the raw PLLA sample at the glass transition region, while the following exothermic peak is associated with the cold crystallization process of raw PLLA. Since physical relaxation of polymer chains at Tg region is a non-reversing process [60], the ΔCP measured in the reversing heat capacity curve is only associated with the amorphous structure fraction (XMAF = ΔCP/ΔCp0) of the PLLA polymer. Here, the ΔCp0 of PLLA can be calculated from the heat capacity difference between the theoretical baselines Cliquid (blue dashed line) p and Csolid (red dashed line) at Tg, which is around 0.61 [45,60]. Since p the raw PLLA has almost no crystalline initially (XC = 0, proved by WAXS), the heat capacity difference between Cliquid line and the top of p the endothermic shoulder at Tg is directly associated with the rigid amorphous structure (XRAF), which is close to 1 − XC − ΔCP/ΔCP0 = 3% for the raw PLLA sample. The heat capacities of PLLA samples measured by SSDSC are listed in Table 4, as well as the calculated phase fractions. As the CO2 gas pressure increases from 4.0 to 7.0 MPa, ΔCP decreases gradually from 0.40 to 0.29 and the MAF decreases from 0.66 to 0.47, while the crystallinity of tested samples increases from 0.26 to 0.39. Meanwhile, the RAF content increases from 0.09 to 0.14 with the increasing pressure (Table 4, XC-DSC, XMAF-DSC and XRAF-DSC). Consistent with previous WAXRD analyses, it is observed by SSDSC that the foams treated at higher pressure have higher crystallinities, more rigid amorphous contents (RAF) and higher decomposition temperatures (measured by TG), which indicates the improvement of stability of the PLLA samples. Therefore, it is clear that the foaming procedure can significantly affect the structure of PLLA by increasing both the crystallinities and the contents of rigid amorphous (mesophase) structure. 3.6. Enzyme degradation analysis Enzyme degradation of different PLLA samples was also studied in the present work. Fig. 8A illustrates the weight losses of the raw material and PLLA foams fabricated at 4.0, 5.0, 6.0, and 7.0 MPa. The degradation profiles show that more crystallinities are formed in the samples with higher saturation pressure, which results in a slower biodegradation rate. After 1 day, the total mass losses of PLLA foams fabricated at various pressures (e.g., 4.0, 5.0, 6.0, and 7.0 MPa) are measured to be about 47%, 32%, 25%, and 20%, respectively. The result suggests that the 1-day degradation rate of PLLA foam is quite fast, perhaps because the degradation is initially targeted at amorphous structures of the scaffold. As reported by Reeve et al. [61], the enzymatic degradation preferentially occurred in the amorphous region of PLLA, and then the degradation tended to be slowed down due to the slow degradation of crystal structures in samples. After 5 days, the remaining masses of

620

S.-J. Sheng et al. / Materials Science and Engineering C 49 (2015) 612–622

Fig. 8. (A): Five-day proteinase K enzymatic degradation study of PLLA scaffolds prepared by 4.0 MPa, 5.0 MPa, 6.0 MPa, and 7.0 MPa. (n N 5, bars represent standard deviation, p b 0.001). (B): typical SEM morphologies of PLLA scaffolds prepared at different pressures: a. 4.0 MPa; b. 5.0 MPa; c. 6.0 MPa and d. 7.0 MPa after 5 d enzymatic degradation (bar length: 20 μm).

scaffolds treated at 4.0 MPa, 5.0 MPa, 6.0 MPa, and 7.0 MPa reduce to 22%, 30%, 42%, and 57% of the original mass, respectively, while the remaining mass of raw PLLA is as high as 64%. The degradation analyses also show that samples with higher porosity tend to have slower degradations. As previously demonstrated in the SSDSC and XRD studies, compared to the samples fabricated at lower pressures, PLLA foams fabricated at higher pressures have more crystalline, more rigid amorphous fraction and smaller pore size. To further understand the enzymatic degradation kinetics, the surface morphologies of different PLLA samples were examined by using SEM after 5 days enzymatic degradation, and the results are shown in Fig. 8B. After the proteinase K degradation, the scaffolds demonstrate different cracked or interconnected porous structures, and the pore sizes of all samples are enlarged. Scaffolds with bigger pores at the initial stage will degrade further, and in turn result in the enhancement of pore sizes and depths. Therefore, the combined results confirmed that the pressure, which can affect the crystallinity and porosity of scaffolds, should also be a crucial factor to control the degradation rate of PLLA scaffolds, which makes it possible to design desired scaffolds that can meet specific requirements in tissue regeneration.

3.7. Cytotoxicity assessments based on MTT assay In order to evaluate cell biocompatibility and viability in different porous PLLA scaffolds (4.0 MPa, 5.0 MPa, 6.0 MPa, 7.0 MPa), HeLa cells which are human cervical cancer cells were utilized in this study. Cell viability results were determined by MTT assay after 24, 48, and 72 h of incubation in DMEM at 37 °C and 5% CO2, respectively. Fig. 9 indicates that more than 90% of cells are alive in all PLLA scaffolds at any culture time, while the overall cell viability has an increase along with the culture time (24, 48, and 72 h). This result indicates that the PLLA scaffolds can provide a micro-environment for HeLa cells to attach and proliferate [62,63]. Fig. 9 also showed that the viability of cells increases with the decreasing pore size. It is observed that the percentage of HeLa cells successfully grow in the scaffold fabricated at a pressure of 7.0 MPa is ~4.2%, 4.1% and 5.0% higher than that on 4.0 MPa scaffold at 24, 48 and 72 h, respectively. It indicates that the cell growth can be elevated drastically with the decreased pore size of scaffold, which provides more porous topology among the foams and might facilitate the cell attachment and growth [64].

S.-J. Sheng et al. / Materials Science and Engineering C 49 (2015) 612–622

621

this study to fabricate PLLA scaffold which had suitable stability, mechanical property and tissue cell viability for TES applications. In our future work, more efforts will be made to investigate the application of this structure in the area of controlled release system (e.g., proteins and drug delivery).

Acknowledgments

Fig. 9. Assessments of HeLa cell viability in 4.0 MPa, 5.0 MPa, 6.0 MPa, and 7.0 MPa PLLA scaffolds for 24, 48, and 72 h by MTT assay. Assays of equivalent cell number (1 × 104/well) on well plates were used as control values for total available cells (**p b 0.01, *p b 0.05).

This work was supported by the National Natural Science Foundation of China (11274176 and 11474166), the Analysis Method and Technology Guide Project of Science & Technology Department of Jiangsu Province (JKYB201401), and the Priority Academic Program Development of Jiangsu Higher Education Institutions (JSYSXK2014), as well as the Nanjing Laboratory Platform Foundation (1640703064) and Nanjing Normal University Visiting Scholarship Foundation of China (2013–2014) (NJNU20121217). XH also thanks the Rowan University 2013–2014 Seed Funding Program (10110-60930-7460-12) for support of this research.

References 4. Conclusions In this study, PLLA foams with different pore sizes were fabricated with a solvent free solid-state gas foaming method. The structure and the properties of the PLLA samples, such as their morphological characteristics, the crystallinity, thermal stability, structural properties and biodegradation properties have been investigated. SEM analysis demonstrates that the pore size and the pore density of foams can be controlled by the saturation pressure. An increase in the saturation pressure will enhance the pore density of PLLA foams, while in turn reduces the pore size and wall thickness. After the foaming process, both the storage modulus and the loss factors will decrease, and the elasticity of material will be strengthened. Moreover, the loss factor, glass transition temperature and material's enthalpy of PLLA foam will also be raised with the increasing pressure. FTIR results suggest that the structures and the molecular chain interactions in PLLA matrix also changed during the foaming process. TG analysis illustrates that the thermal stability of scaffolds increases with the decrease of pore size and the increase of foaming pressure. And the enzymatic degradation study showed the same trend. The combination of SSDSC and WAXRD studies proves that the crystallinity of PLLA polymer increased after the foaming process; the content of rigid amorphous phase in PLLA, a mesophase that can affect the rigidity and other properties of the polymer samples, is also showed to gradually increase with the increasing foaming pressure. According to the experimental results, it is reasonable to believe that the saturation foaming pressure can significantly affect the nucleation mechanism, crystalline structure, and stability of PLLA scaffolds. Meanwhile, the foaming pressure can initiate the partial ordering of some amorphous polymer chains, and form additional rigid amorphous structures in PLLA, which cannot have sufficient time to crystallize during the quick foaming process. On the other hand, the foam process can also reorganize many amorphous chains into crystals, so that a crystallized PLLA matrix can be fabricated. Also, the MTT assay demonstrates lower cytotoxicity for PLLA scaffolds with smaller pore size in vitro, which is likely beneficial for cell attachment and growth. In summary, this study might provide a better understanding of the influence of foaming pressure effects on the morphology, mechanical property, thermal stability, enthalpy value, crystallinity and rigid amorphous structure of microcellular PLLA materials. These findings would enable precise design and fabrication of PLLA scaffolds with controllable material properties, which would significantly benefit the applications of engineered scaffold materials in various biomedical fields in the future. The green foaming technique has been used and investigated in

[1] D.D. Tu, Y.G. Chung, E.S. Gil, A. Seth, D. Franck, Bladder tissue regeneration using acellular bi-layer silk scaffolds in a large animal model of augmentation cystoplasty, Biomaterials 34 (2013) 8681–8689. [2] X. Hu, X.L. Wang, J. Rnjak, A.S. Weiss, D.L. Kaplan, Biomaterials derived from silktropoelastin protein systems, Biomaterials 31 (2010) 8121–8131. [3] I. Pallotta, J.A. Kluge, J. Moreau, R. Calabrese, D.L. Kaplan, A. Balduini, Characteristics of platelet gels combined with silk, Biomaterials 35 (2014) 3678–3687. [4] X.Q. Wang, T. Yucel, Q. Lu, X. Hu, D.L. Kaplan, Silk nanospheres and microspheres from silk/pva blend films for drug delivery, Biomaterials 31 (2010) 1025–1035. [5] J.F. Xie, H.W. Zhang, X. Li, Y.L. Shi, Entrapment of methyl parathion hydrolase in cross-linked poly(gamma-glutamic acid)/gelatin hydrogel, Biomacromolecules 15 (2014) 690–697. [6] K. Numata, S. Yamazaki, N. Naga, Biocompatible and biodegradable dual-drug release system based on silk hydrogel containing silk nanoparticles, Biomacromolecules 13 (2012) 1383–1389. [7] J.W. Qin, Y.Y. Jiang, J.J. Fu, Y.Q. Wan, R.H. Yang, W.D. Gao, H.B. Wang, Evaluation of drug release property and blood compatibility of aspirin-loaded electrospun PLA/ RSF composite nanofibers, Iran. Polym. J. 22 (2013) 729–737. [8] Y.X. Duan, J. Liu, Molecular weight dependence of the poly(L-lactide)/poly(Dlactide) stereocomplex at the air–water interface, Biomacromolecules 7 (2006) 2728–2735. [9] J.F. Mano, Structural evolution of the amorphous phase during crystallization of poly(L-lactide): a synchrotron wide-angle X-ray scattering study, J. Non-Cryst. Solids 353 (2007) 2567–2572. [10] D. Hermida-Merino, M. Belal, B.W. Greenland, P. Woodward, A.T. Slark, F.J. Davis, G.R. Mitchell, I.W. Hamley, W. Hayes, Electrospun supramolecular polymer fibers, Eur. Polym. J. 48 (2012) 1249–1255. [11] M.S. Huang, J. Coudane, S. Li, M. Vert, Methylated and pegylated PLA–PCL–PLA block copolymers via the chemical modification of di-hydroxy PCL combined with the ring opening polymerization of lactide, J. Polym. Sci. Polym. Chem. 43 (2005) 4196–4205. [12] M. Day, A.V. Nawaby, X. Liao, A DSC study of the crystallization behaviour of polylactic acid and its nanocomposites, J. Therm. Anal. Calorim. 86 (2006) 623–629. [13] H. Chen, M. Pyda, P. Cebe, Non-isothermal crystallization of PET/PLA blends, Thermochim. Acta 492 (2009) 61–66. [14] H.I. Kim, K. Ishihara, S. Lee, J.H. Seo, H.Y. Kim, D. Suh, M.U. Kim, T. Konno, M. Takai, J.S. Seo, Tissue response to poly(L-lactic acid)-based blend with phospholipid polymer for biodegradable cardiovascular stents, Biomaterials 32 (2011) 2241–2247. [15] H.L. Zhu, X.X. Feng, H.P. Zhang, Y.H. Guo, J.Z. Zhang, J.Y. Chen, Structural characteristics and properties of silk fibroin/poly(lactic acid) blend films, J. Biomater. Sci. 20 (2009) 1259–1274. [16] M. Shimbo, I. Higashitani, Y. Miyano, Mechanism of strength improvement of foamed plastics having fine cell, J. Cell. Plast. 43 (2007) 157–167. [17] K. Nadella, V. Kumar, in: E.E. Gdoutos (Ed.), Experimental Analysis of Nano and Engineering Materials and Structures, Alexandroupolis Publishing, Greece, 2007, pp. 765–766. [18] A. Shakoor, R. Muhammad, N.L. Thomas, V.V. Silberschmidt, Mechanical and thermal characterisation of poly(L-lactide) composites reinforced with hemp fibres, J. Phys. Conf. Ser. 451 (2013) 1–10. [19] H.T. Zou, C.H. Yi, L.X. Wang, H.T. Liu, W.L. Xu, Thermal degradation of poly(lactic acid) measured by thermogravimetry coupled to Fourier transform infrared spectroscopy, J. Therm. Anal. Calorim. 97 (2009) 929–935. [20] F. Wang, G.P. Guo, Q.Y. Ma, M.F. Gu, X.Y. Wu, S.J. Sheng, X.S. Wang, Investigation on the thermo-mechanical properties and thermal stability of polylactic acid tissue engineering scaffold material, J. Therm. Anal. Calorim. 113 (2013) 1113–1121. [21] S.Z. Guo, Scanning Electron Microscope Technology and Its Applications, second ed. Xiamen University Press, Xiamen, China, 2006.

622

S.-J. Sheng et al. / Materials Science and Engineering C 49 (2015) 612–622

[22] F. Carrasco, J. Gamez-Perez, O.O. Santana, M.L. Maspoch, Processing of poly(lactic acid)/organomontmorillonite nanocomposites: microstructure, thermal stability and kinetics of the thermal decomposition, J. Chem. Eng. 178 (2011) 451–460. [23] H. Liu, S.D. Wang, N. Qi, Controllable structure, properties, and degradation of the electrospun PLGA/PLA-blended nanofibrous scaffolds, J. Appl. Polym. Sci. 125 (2012) E468–E476. [24] H.Y. Mi, M.R. Scalick, X. Jing, B.R. Jacques, W.C. Crone, X.F. Peng, L.S. Turng, Characterization of thermoplastic polyurethane/polylactic acid (TPU/PLA) tissue engineering scaffolds fabricated by microcellular injection molding, Mater. Sci. Eng. C Mater. Biol. Appl. 33 (2013) 4767–4776. [25] R. Pantani, G. Gorrasi, G. Vigliotta, M. Murariu, P. Dubois, PLA–ZnO nanocomposite films: water vapor barrier properties and specific end-use characteristics, J. Eur. Polym. 49 (2013) 3471–3482. [26] H.L. Zhang, N.A. Liu, X.H. Ran, C.Y. Han, L.J. Han, Y.G. Zhuang, L.S. Dong, Toughening of polylactide by melt blending with methyl methacrylate–butadiene–styrene copolymer, J. Appl. Polym. Sci. 125 (2012) E550–E561. [27] I.V. Yannas, J.F. Burke, Design of an artificial skin. I. Basic design principles, J. Biomed. Mater. Res. 14 (1980) 65–81. [28] G. Galassi, P. Brun, M. Radice, R. Cortivo, G.F. Zanon, P. Genovese, G. Abatangelo, In vitro reconstructed dermis implanted in human wounds: degradation studies of the HA-based supporting scaffold, Biomaterials 21 (2000) 2183–2191. [29] A. Martins, A.R.C. Duarte, S. Faria, A.P. Marques, R.L. Reis, N.M. Neves, Osteogenic induction of hBMSCs by electrospun scaffolds with dexamethasone release functionality, Biomaterials 31 (2010) 5875–5885. [30] F. Wang, Z.Q. Li, M. Khan, K. Tamama, P. Kuppusamy, W.R. Wagner, C.K. Sen, J.J. Guan, Injectable, rapid gelling and highly flexible hydrogel composites as growth factor and cell carriers, Acta Biomater. 6 (2010) 1978–1991. [31] Designation: D792-00, Standard Test Methods for Density and Specific Gravity (Relative Density) of Plastics by DisplacementASTM Committee D20. Approved Dec. 10, 2000 ASTM, March 2001. [32] M. Murariu, A. Doumbia, L. Bonnaud, A.L. Dechief, Y. Paint, M. Ferreira, C. Campagne, E. Devaux, P. Dubois, High-performance polylactide/ZnO nanocomposites designed for films and fibers with special end-use properties, Biomacromolecules 12 (2011) 1762–1771. [33] M. Pyda, B. Wunderlich, Reversing and nonreversing heat capacity of poly(lactic acid) in the glass transition region by TMDSC, Macromolecules 38 (2005) 10472–10479. [34] H.P. Klug, L.E. Alexander, X-ray Diffraction Procedures, John Wiley & Sons Inc., New York, 1962. 452–453. [35] D.C. Li, T. Liu, L. Zhao, X.S. Lian, W.K. Yuan, Foaming of poly(lactic acid) based on its nonisothermal crystallization behavior under compressed carbon dioxide, Ind. Eng. Chem. Res. 50 (2011) 1997–2007. [36] S. Chapple, R. Anandjiwala, R.S. Sinha, Mechanical, thermal, and fire properties of polylactide/starch blend/clay composites, J. Therm. Anal. Calorim. 113 (2013) 703–712. [37] J.D. Shao, Y.J. Wang, X.F. Chen, X.M. Hu, C. Du, Nanomechanical properties of poly(Llactide) nanofibers after deformation, Colloid. Surface. B. 120 (2014) 97–101. [38] S.M. Lee, D. Cho, W.H. Park, S.G. Lee, Novel silk/poly(butylene succinate) biocomposites: the effect of short fibre content on their mechanical and thermal properties, Compos. Sci. Technol. 65 (2005) 647–654. [39] M. Dustin, K. Vipin, Microcellular and nanocellular solid-state polyetherimide (PEI) foams using sub-critical carbon dioxide II. Tensile and impact properties, Polymer 52 (2011) 2910–2919. [40] Y. Grohens, J. Sacristan, L. Hamon, H. Reinecke, C. Mijangos, J.M. Guenet, Glass transition of ultra-thin films of modified PVC, Polymer 42 (2001) 6419–6423. [41] N.M. Alves, J.F. Mano, J.L.G. Ribelles, Molecular mobility in a thermoset as seen by TSR and DMA near Tg, Mater. Res. Innov. 4 (2001) 170–178. [42] R.S. Shih, S.W. Kuo, F.C. Chang, Thermal and mechanical properties of microcellular thermoplastic SBS/PS/SBR blend: effect of crosslinking, Polymer 52 (2011) 752–759.

[43] J.M. Zhang, Y. Duan, A.J. Domb, Y. Ozaki, PLLA mesophase and its phase transition behavior in the PLLA–PEG–PLLA copolymer as revealed by infrared spectroscopy, Macromolecules 43 (2010) 4240–4246. [44] Q. Ma, M. Pyda, B. Mao, P. Cebe, Relationship between the rigid amorphous phase and mesophase in electrospun fibers, Polymer 54 (2013) 2544–2554. [45] Y. Chen, J. Lin, Y.Q. Wan, Y.N. Fei, H.B. Wang, W.D. Gao, Preparation and blood compatibility of electrospun PLA/curcumin composite membranes, Fiber Polym. 13 (2012) 1254–1258. [46] M. Pluta, Morphology and properties of polylactide modified by thermal treatment, filling with layered silicates and plasticization, Polymer 45 (2004) 8239–8251. [47] R.M. Felfel, I. Ahmed, A.J. Parsons, P. Haque, G.S. Walker, C.D. Rudd, Investigation of crystallinity, molecular weight change, and mechanical properties of PLA/PBG bioresorbable composites as bone fracture fixation plate, J. Biomater. Appl. 26 (2012) 765–787. [48] T.M. Wu, C.Y. Wu, Biodegradable poly(lactic acid)/chitosan-modified montmorillonite nanocomposites: preparation and characterization, Polym. Degrad. Stab. 91 (2006) 2198–2204. [49] S. Iannace, A. Maffezzoli, G. Leo, L. Nicolais, Influence of crystal and amorphous phase morphology on hydrolytic degradation of PLLA subjected to different processing conditions, Polymer 42 (2001) 3799–3807. [50] J. Zhang, D.X. Yan, J.Z. Xu, H.D. Huang, J. Lei, Z.M. Li, Highly crystallized poly(lactic acid) under high pressure, AIP Adv. 2 (2012) 42159–42165. [51] Y. Azuma, N. Yoshie, M. Sakurai, Y. Inoue, R. Chujo, Thermal behaviour and miscibility of poly(3-hydroxybutyrate)/poly(vinyl alcohol) blends, Polymer 33 (1992) 4763–4767. [52] K. Jamshidi, S.H. Hyon, Y. Ikada, Thermal characterization of polylactides, Polymer 29 (1998) 2229–2234. [53] H. Nishida, T. Mori, S. Hoshihara, Y.J. Fan, Y. Shirai, T. Endo, Effect of tin on poly (L-lactic acid) pyrolysis, Polym. Degrad. Stab. 81 (2003) 515–523. [54] H.X. Yu, N.X. Huang, C.S. Wang, Z.L. Tang, Modeling of poly(L-lactide) thermal degradation: theoretical prediction of molecular weight and polydispersity index, J. Appl. Polym. Sci. 88 (2003) 2557–2562. [55] X.Q. Zhang, M. Espiritua, A. Bilyka, L. Kurniawan, Morphological behaviour of poly(lactic acid) during hydrolytic degradation, Polym. Degrad. Stab. 93 (2008) 1964–1970. [56] M. Pyda, ATHAS Databank, http://athas.prz.rzeszow.pl/. [57] G. Stoclet, R. Seguela, J.M. Lefebvre, S. Elkoun, C. Vanmansart, Strain-Induced molecular ordering in polylactide upon uniaxial stretching, Macromolecules 43 (2010) 1488–1498. [58] C. Schick, A. Wurm, A. Mohammed, Formation and disappearance of the rigid amorphous fraction in semicrystalline polymers revealed from frequency dependent heat capacity, Thermochim. Acta 396 (2003) 119–132. [59] B. Wunderlich, Reversible crystallization and the rigid–amorphous phase in semicrystalline macromolecules, Prog. Polym. Sci. 28 (2003) 383–450. [60] M. Pyda, R.C. Bopp, B. Wunderlich, Heat capacity of poly(lactic acid), J. Chem. Thermodyn. 36 (2004) 731–742. [61] M.S. Reeve, S.P. McCarthy, M.J. Downey, R.A. Gross, Polylactide stereochemistry: effect on enzymatic degradability, Macromolecules 27 (1994) 825–831. [62] H.L. Zhu, X.X. Feng, H.P. Zhang, Y.H. Guo, J.Z. Zhang, Structural characteristics and properties of silk fibroin/poly(lactic acid) blend films, J. Biomater. Sci. 20 (2009) 1259–1274. [63] K. Hu, Q. Lv, F.Z. Cui, L. Xu, Y.P. Jiao, Y. Wang, Q.L. Feng, H.L. Wang, L.Y. Huang, A novel poly(L-lactide) (PLLA)/fibroin hybrid scaffold to promote hepatocyte viability and decrease macrophage responses, J. Bioact. Compat. Polym. 22 (2007) 395–410. [64] R. Dorati, C. Colonna, C. Tomasi, I. Genta, G. Bruni, B. Conti, Design of 3D scaffolds for tissue engineering testing a tough polylactide-based graft copolymer, Mater. Sci. Eng. C 34 (2014) 130–139.