Abstract
PLA/SiO2 nanocomposites (PLAS) were prepared by premixing in chloroform solution and then melt extrusion. SiO2 were surface grafted via ring-opening polymerisation of L-Lactide before blending. Crystallisation and rheological properties of PLAS were investigated as a function of SiO2 content. It was identified from differential scanning calorimetry cooling plots that the addition of SiO2 could improve crystallisation rate and relative crystallinity of PLA. Meanwhile, SiO2 had affected rheological properties of PLA. Pure PLA and PLAS were both typical shear-thinning fluid, and shear viscosity of PLAS was higher than pure PLA. In dynamic sweep test, the linear viscoelastic regions with critical strains of PLAS decreased with the increasing SiO2 content compared with pure PLA. The G′, G″ and tan δ versus ω of PLAS were all improved. The results of rheological tests indicated that the addition of SiO2 had effect on the structure of PLA molecular chain and had formed network structure in PLA matrix.
Introduction
Poly(lactic acid) (PLA), a biodegradable aliphatic polyester derived from 100% renewable resources such as corn and sugar beets, 1 is one of the researches focuses on the sustainable and environmental-friendly materials. Because of its renewability, biocompatibility, high strength and high modulus, thermal plasticity and processibility, it is widely used in biomedical, packaging, textile and automotive industries.2, 3 However, the limitations of PLA, such as the inherent brittleness, slow degradation rate and hydrophobicity, has restricted its further application, therefore, it needs to be improved to extend its engineering applications in many fields. A wealth of investigation has been performed to improve the properties, and one of the solutions is to add various inorganic fillers. Clay minerals,4, 5 aluminium hydroxide [Al(OH)3], 6 carbon nanotubes,7, 8 layered silicate, 9 calcium carbonate (CaCO3), 10 titanium dioxide (TiO2), 11 nanosilica (SiO2),12, 13 etc., have been added to PLA to overcome the drawbacks by means of solvent casting and melt-blending.
Nanosilica (SiO2) is the most common inorganic fillers, widely used in the thermoplastic such as polyamide-66, 14 polubutylene terephthalate, 15 polyethylene terephthalate, 16 polypropylene, 17 polystyrene, 18 etc. It is not only because of its extensive sources and easy modification, but also its diverse applications in material science, including biocatalysts, electrochemistry, and optical materials. 19 However, SiO2 nanoparticles incorporated into a polymeric matrix tend to form aggregates due to the small size effects, high surface energy and the relatively poor interaction between SiO2 nanoparticles and the polymer matrix. Therefore, SiO2 nanoparticles have been functionalised to overcome its poor dispersion and compatibility.
The research on the rheological properties of the polymer has important significance in the proper selection of process conditions and formula design, providing the basis for actual operation. The addition of inorganic fillers has an effect on the rheological properties of the polymer. However, the rheological behaviour of the composites, composed of biodegradable polymer and inorganic nanoparticles, have been rarely reported.
In this study, the well dispersed PLA/SiO2 nanocomposites (PLAS) were prepared via solution mixing and melt extrusion. The SiO2 nanoparticles were surface grafted by PLA before blending and characterised by Fourier transform infrared spectroscopy (FTIR) and thermal gravimetric analysis (TGA). Crystallisation behaviour of the PLAS was examined by using differential scanning calorimetry (DSC). The main objective of the study was to characterise the effect of grafted SiO2 nanoparticles on the rheological properties of PLA.
Experimental
Materials
Poly(lactic acid) was supplied by BrightChina Industrial Co., Ltd (China) and exhibited a viscosity-average molecular weight of 1·5×105. SiO2 nanoparticles, AEROSIL200, without any organic treatment was obtained from Degussa (Germany) with an average primary particle size of approximately 12 nm. L-Lactide was obtained from BrightChina Industrial Co., Ltd (China). Toluene, stannous octoate and chloroform were supplied by Shanghai Chemical Reagent Corporation, China Medicine Group. All reagents were identified as being an analytical grade and were used as received.
Sample preparation
SiO2 nanoparticles were first dried at 80°C for 12 h to get rid of the water on the surface. Then 2 g SiO2 nanoparticles were dispersed in the 30 mL toluene with ultrasonic oscillation to get solution A. L-Lactide (1 g) and 5 mL toluene were mixed to get solution B. Both of the A and B solutions were poured into a three-mouth flask with a condenser and the device was heated to 120°C. As the catalyst a few drops of stannous octoate were added, the surface hydroxyl groups of SiO2 as an initiator initiated the ring-opening polymerisation of L-Lactide on the surface of the nanoparticles. The system reacted under N2 with stirring at 120°C for 12 h. The grafted SiO2 nanoparticles were first separated by centrifugation, and were further washed with chloroform for several times. Then the grafted SiO2 nanoparticles were dried at 80°C for 24 h to remove the residual solvent. Finally the grafted SiO2 nanoparticles were obtained.
Poly(lactic acid) pellets were dried at 60°C for 18 h in the vacuum drying oven before blending. At first, the grafted SiO2 nanoparticles were mixed with PLA in chloroform solution and were stirred for well dispersion. The quality ratio of SiO2 and PLA was 1∶4. Then the mixture was precipitated out under the effect of ethanol and dried in the oven at 60°C for 24 h. The obtained mixture was smashed and was further diluted with PLA pellets on a SJSZ-10A conical double-screw extruder (Wuhan, China) (D = 15 mm, L/D = 12) at temperature of 200°C for 10 min with a screw speed of 25 rev min−1. The granulated and dried PLA/SiO2 nanocomposites were named PLAS-x in the following discussion, x representing the SiO2 nanoparticle contents (wt-%) in PLA matrix. The procedure was summarised in Fig. 1. The pure PLA was subjected to the same treatment as PLAS for comparison.

Procedure for processing nanocomposites
Characterisation
Fourier transform infrared spectroscopy (Nicolet is10) was used to analyse the chemical structure of the nanoparticles. A small amount of SiO2 nanoparticles was mixed and grinded with spectral grade KBr, and then was crushed into thin slice in a standard procedure. The FTIR spectrum was obtained through transmission method and recorded in the range of 400–4000 cm−1.
Thermal gravimetric analysis (TA Q500) was utilised to determine the amount of the grafting polymer on the surface of SiO2 nanoparticles. The actual content of SiO2 nanoparticle in PLA matrix was also needed to measure by TGA because there was the mass loss of SiO2 nanoparticles in the process of mixing with PLA on the twin-screw exturder. The range of test temperature was 25 to 600°C and the heating rate was 20°C/min.
Differential scanning calorimeter (TA Q200) was employed to study the crystallisation property of pure PLA and PLAS. N2 was used as the purge gas at a flow rate of 50 mL/min. The samples with a weight range of 5 to 8 mg were put into an aluminium pan and hermetically sealed. The sample was first heated to 200°C and kept for 3 min. Then the sample was cooled to 0°C at the rate of 4°C/min. The cooling plots of pure PLA and PLAS were used to study.
The rheological behaviour of pure PLA and PLAS were analysed in the melt state by a rheometer (Anton Paar MCR301) using a 25 mm parallel-plate fixture. Before testing, pure PLA and PLAS pellets were hot-pressed into diameter 25 mm, thickness 1·5 mm parallel plates by heating at 180°C under the pressure of 6 MPa for 3 min on a compression moulding machine. The sample was located inside of an environmental chamber, and was placed between the preheated fixtures with a constant gap setting of 1 mm. All the rheological tests were carried out at 200°C. Shear viscosity (η) test was performed in the shear rate (
) range of 0·1 to 100 s−1. Dynamic strain sweep was tested in the range of strain amplitude (γ) from 0·1 to 500% with an angular frequency (ω) of 10 rad s−1 to determine the linear viscoelastic (LVE) region. Dynamic frequency sweep was performed in the ω range of 0·1–500 rad s−1 with a strain value in the LVE region to study the storage modulus (G′) and loss modulus (G″) of pure PLA and PLAS.
Results and discussion
FTIR and TG analysis of PLA-grafted SiO2 nanoparticles
Fourier transform infrared spectrum is shown in Fig. 2 to confirm the organic groups onto the surface of SiO2 nanoparticles. The peaks at 1200–1050 and 809 cm−1 were characteristic peaks of SiO2. 20 The strong and wide band at 1200–1050 cm−1 referred to the anti-symmetric stretching vibration of Si–O–Si. The peak at 809 cm−1 represented the symmetric stretching vibration of Si–O. The peaks at 3433 and 1636 cm−1 corresponded to the anti-symmetric stretching vibration of constitutional water –OH and bending vibration of H–O–H, respectively. Compared with the original sample, the grafted SiO2 nanoparticles exhibited absorption at 1735 cm−1. The absorption at 1735 cm−1 was attributed to the stretching vibration of C = O of PLA chains. 21 From the analysis, it was valid to conclude that the surface of SiO2 nanoparticles was successfully grafted with PLA.

Spectra (FTIR) of original and grafted SiO2 nanoparticles
It could be further confirmed from the result of the TG test that there was the grafted polymer of PLA on the surface of SiO2 nanoparticles. It can be obtained from Fig. 3 that the ratio of weight loss for grafted SiO2 nanoparticles was about 13%, which was the amount of the graft polymer on the surface of SiO2 nanoparticles. It also can be seen from Fig. 3 that the thermal weight loss process of grafted SiO2 nanoparticles consisted of two stages, which may be due to the different molecular weight of the graft polymer on the surface of SiO2 nanoparticles.

TG plots of original and grafted SiO2 nanoparticles
Crystallisation property of PLA/SiO2 nanocomposites
The DSC cooling plots of pure PLA and PLAS are shown in Fig. 4. The actual content of SiO2 nanoparticle in PLA matrix was measured to be 1·23, 3·73 and 6·36 wt-%, respectively, through the TG test. Therefore, the PLAS could be designated as PLAS-1·23, PLAS-3·73 and PLAS-6·36, respectively.

Cooling plots (DSC) of pure PLA and PLAS
In Fig. 4, the melt-crystallisation peaks could be clearly observed from the plots of PLAS. However, the melt-crystallisation peak of pure PLA was not obvious due to the slow crystallisation rate. It can be seen that with the addition of SiO2 nanoparticles, the onset melt crystallisation temperature (Tonset) and the melt crystallisation peak temperature (Tc) both moved toward higher temperature. Meanwhile the melt crystallisation peaks became sharp and the peak width was narrowed with the increased content of SiO2 nanoparticles. In other words, PLA was improved to crystallise intensively at higher temperature and form better crystals after adding the SiO2 nanoparticles.
Molecular chains usually need a certain time to rearrange into the lattice, making the crystallisation process lag behind the cooling process. The slower the crystallisation rate, the greater the molecular chains into the crystal structure at low temperature. However, the mobility of molecular chains is poor at low temperature, which is easy to cause the imperfect crystal structure. The crystallisation peak of imperfect crystal structure was relatively wide and flat. By the results of the high Tonset and sharp crystallisation peak on the cooling plots of PLAS, it can be concluded that the addition of SiO2 nanoparticles could effectively improve the crystallisation rate of PLA.
The degree of relative crystallinity (Xc) of pure PLA and PLAS was calculated through the cooling plots, using the following equation
22
Characteristic temperatures and degree of relative crystallinity of pure PLA and PLA/SiO2 nanocomposites from DSC cooling plots
Flow curve of PLA/SiO2 nanocomposites
Figure 5 shows the flow curves of shear rate (
) versus shear viscosity (η) for pure PLA and PLAS. It can be seen from Fig. 4 that the flow curve can be roughly divided into two areas. When
was low, η tended to be constant and a viscosity platform appeared. This area is known as the linear flow area. When
exceeded a certain threshold, η decreased with the increasing
. This area is a typical pseudoplastic area. The flow curve of pure PLA, PLAS-1·23 and PLAS-3·73 had both the linear flow area and shear thinning area. While the flow curve of PLAS-6·36 showed the shear-thinning characteristic without a viscosity platform. It also can be seen that the η varied with SiO2 nanoparticles content. The higher the content of SiO2 nanoparticles, the greater the η, especially at low shear rate region.

Flow curves of η versus
for pure PLA and PLAS at temperature of 200°C
The flow curve of polymer reflects the results that the morphology of molecular chain changes in the flow field. In general, the polymer molecular chains could form the transient physical crosslinking because of mutual entanglement or interaction of van der Waals force. These physical crosslinking points are constantly in the dynamic balance of disintegration and reconstruction with the action of molecular thermal motion, which makes the whole fluid with a transient crosslinking network structure. At the low shear rate range, the physical crosslinking points are less destroyed and structure density could basically remain the same. Thus the shear viscosity can stay the same, and the polymer fluid is in the linear flow area. The melt flow property was similar to Newton fluid. When the shear rate increases to a certain value, the molecular chains extend and orient along the direction of the flow field, and the disintegration speed of physical crosslinking points is faster than the reconstruction speed. Therefore the shear viscosity begins to decrease and the fluid shows pseudoplastic behaviour.
Therefore it can be concluded that the pure PLA and PLAS are all the pseudoplastic fluid in the melt state, and have the characteristics of shear thinning under a certain condition of shear rate. For PLAS, the grafted PLA on the surface of SiO2 nanoparticles could form some new physical crosslinking points with the molecular chains of PLA matrix. With the increased content of SiO2 nanoparticle, the interaction between them also increased. The mobility of PLA molecular chains was thus restricted. Therefore η of PLAS was greater than the pure PLA. In addition, when the content was high, SiO2 nanoparticles may form a network structure in PLA matrix. The network structure would be destroyed with the increasing
, led to a decrease of η. This might be the reason for no viscosity platform appeared in the flow curve of PLAS-6·36.
Nonlinear viscoelasticity of PLA/SiO2 nanocomposites
An important step in performing dynamic rheological characterisation is to determine the LVE region of materials in which dynamic rheological parameters are independent of applied strains. Figure 6 shows the dependence of storage modulus (G′) versus the strain amplitude (γ) for pure PLA and PLAS.

Plots of G′ versus γ for pure PLA and PLAS at temperature of 200°C
It can be seen from Fig. 6 that G′ was less strain dependent at the low frequency region for pure PLA and PLAS. The LVE regions with critical strains (γc) gradually decreased with the addition of SiO2 nanoparticles. γc characterises the deviation from the linear to non-linear regimes, which was from about 100% of strain for pure PLA to less than 10% of strain for PLAS-6·36. The behaviour of G′ from linear to non-linear with increased γ associated with both mechanisms of chain disentanglements and filler breakdown depending of SiO2 nanoparticles content and amplitude deformation. 24 Therefore it can be concluded that the addition of SiO2 nanoparticles had effect on the molecule chain structure of PLA and SiO2 nanoparticles had formed network structure in PLA matrix.
Linear viscoelasticity of PLA/SiO2 nanocomposites
In this part, dynamic frequency sweep was performed and the relationships of storage modulus (G′), loss modulus (G″) and loss factor (tan δ) versus the angular frequency (ω) for pure PLA and PLAS were studied.
It can be obtained from Fig. 7a that the values of G′ for PLAS were higher than that of pure PLA and increased with the content of SiO2 nanoparticles at the low frequency region. It indicated that the addition of SiO2 nanoparticles enhanced the G′ of PLA. In addition, there was a G′ plateau appeared on the plot of PLAS-6·36 at the low frequency region in Fig. 7a. The G′ plateau usually meant the formation of a percolated network in polymer matrix and was considered to be the performance of solid-like behaviour.25, 26 That is, when the content of SiO2 nanoparticles was 6·36 wt-% a network was formed in PLA matrix. Thus the movement of PLA molecular chains was restricted, resulting in the solid-like behaviour to PLAS-6·36.

Plots of G′, G″ and tan δ versus ω for pure PLA and PLAS at temperature of 200°C
As observed from Fig. 7b, the plots of G″ versus ω for PLAS-1·23 and PLAS-3·73 had a little difference with pure PLA, except for the high SiO2 nanoparticle content of PLAS-6·36. It suggested that when the SiO2 nanoparticle content was low, the impact on G″ was small. When the SiO2 nanoparticle content was high, G″ could be increased.
Loss factor (tan δ), obtained by the ratio of G″ and G′, is a measure of the dissipation of heat energy in each deformation period. The big value of tan δ indicated that the elasticity of materials is small. The plots of tan δ versus ω for pure PLA and PLAS were showed in Fig. 7c. The value of tan δ was significantly decreased with the addition of SiO2 nanoparticles at the low frequency region. It illustrated that the elastic strain component of PLA increased after adding SiO2 nanoparticles. The addition of SiO2 nanoparticles made contributions to the improvement of elasticity for PLA.
Conclusions
PLA-grafted SiO2 nanoparticles were synthesised via the ring-opening polymerisation of L-Lactide on the surface of the nanoparticles, and the combination of solvent-blending and melt-extruding was applied for preparation of PLAS with various nanoparticle contents. The existence of C = O at 1735 cm−1 in the FTIR spectrum of the grafted SiO2 nanoparticles indicated that PLA chains were successfully grafted on the surface of SiO2 nanoparticles. The content of grafting PLA on the SiO2 surface was characterised to be about 13 wt-% by TGA. The actual contents of SiO2 nanoparticle in PLA matrix were determined to be 1·23, 3·73 and 6·36 wt-%, respectively, also by TGA. Through the DSC measurement, it could be concluded that the crystallisation property of PLA was improved by adding SiO2 nanoparticles. The crystallisation rate and relative crystallinity were both promoted. The rheological test proved the effect of grafted SiO2 nanoparticles on the molecular chains structure of PLA. All of pure PLA and PLAS were typical non-Newtonian pseudoplastic fluid with shear thinning. The shear viscosity of PLAS was increased with the increasing SiO2 content. This indicated that SiO2 nanoparticles had effect on the mobility of the molecular chains of PLA, led to the increase in shear viscosity. The LVE region was determined by the plots of G′ versus γ. The values of γc of PLAS were less than pure PLA and decreased with the increasing SiO2 content. Meanwhile the addition of SiO2 nanoparticles improved the G′ and G″ of PLA. The elasticity of PLA had been especially improved, proved by the results of tan δ. All the results of dynamic sweep measurement illustrated that the SiO2 nanoparticles had affected the molecular chain structure of PLA and formed network structure in PLA matrix.
Footnotes
Acknowledgement
This work was supported by the Fundamental Research Funds for the Doctoral Graduate (Grant No. JUDCF11035).
