Mechanical properties of PLA, PLA + GO and PLA + GNP films were evaluated using tensile tests. The films were dried in two different situations: under room conditions and in a vacuum oven at 40 °C. In the first case, a solvent content of about 3 wt% is still present, having a significant plasticizing effect. The resulting stress–strain curves show a well-defined yield point followed by some strain hardening, as seen in the representative curves shown in Fig. 7. Since the films were applied by blade spreading, preliminary traction tests were performed in the direction of spreading and in the perpendicular direction in order to check the existence of anisotropy. Identical results are obtained, therefore confirming that the films are mechanically isotropic.
Figure 7. (A) Representative stress–strain curves for PLA, PLA + GO 0.4 wt% and PLA + GNP 0.4 wt% plasticized films. (B) Close-up view of the initial portion of (A).
Download figure to PowerPoint
For a GO content of 0.3 wt%, the Young's modulus increases by 115% (Fig. 8(A)) and the yield strength by 95% (Fig. 8(B)), compared to pristine PLA film. For larger GO contents, the performance decreases in terms of yield strength and Young's modulus, probably due to less homogeneous distribution and agglomeration of GO particles within the PLA matrix. The existence of an optimum loading can also be seen for a GNP content of 0.4 wt%, for which Young's modulus increases by 156% (Fig. 8(A)) and yield strength by 129% (Fig. 8(B)). The results are very similar for GO and GNP, the latter seeming to yield slightly better results, but the differences are within experimental error. The ultimate strength (Fig. 8(C)) and the elongation at break (Fig. 8(D)) measurements do not show a well-defined dependence on loading, but seem not to be considerably affected by filler addition. This is not a determinant fact, since for many applications only the properties at the yield point, and not at the fracture limit, are relevant for the performance of the material.
Figure 8. Effect of increasing nanofiller load on mechanical properties of plasticized PLA films. Error bars represent the standard deviation computed from five measurements.
Download figure to PowerPoint
Films dried under vacuum are completely solvent free and are therefore not plasticized. These present stress–strain curves typical of a glassy polymer, as shown in Fig. 9. Yielding behaviour is not present, and elongation at break is considerably reduced: from values higher than 200% to 3–4%. These films also exhibit higher values of Young's modulus (2–4 GPa) and tensile strength (ca 50–60 MPa). Rhim and co-workers already reported the significant solvent-induced plasticization effect when solvent-cast PLA films are produced with incomplete drying.
The existence of an optimum loading for the unplasticized films can again be seen, as occurs with the plasticized material. Films with 0.4 wt% GNP content show an increase of 85% in Young's modulus (Fig. 10(A)) and of 15% in tensile strength (Fig. 10(B)). Results for GO and GNPs are again very similar. Nanofiller addition again does not have a considerably effect on elongation at break (values not shown here).
Figure 10. Effect of increasing nanofiller content on mechanical properties of PLA films after drying in a vacuum oven. Error bars represent the standard deviation computed from five measurements.
Download figure to PowerPoint
In principle, it would be expected that incorporation of GO would lead to a stronger reinforcement effect than incorporation of GNPs, considering that the presence of oxidized groups over the whole surface would favour interactions with hydrophilic groups of PLA. However, the wrinkled morphology of GO is less favourable than the planar geometry of GNPs for interaction with the polymer matrix, which may explain the similarities in the results.
The results presented here show very significant improvements in mechanical properties with the addition of very small amounts of both GO and GNPs. An optimum loading is identified of about 0.4 wt%, indicating that for higher filler additions agglomeration effects overcome the reinforcement benefits. The range of improvement in mechanical performance on addition of GO and GNP is larger for solvent-plasticized films. It would be expected a priori that filler addition would have different impacts on the performance of plasticized and unplasticized films, considering the completely distinct mechanical behaviour presented by each material. However, it may also be considered that the presence of solvent/plasticizer has an effect on the level of interaction between polymer and filler.
The relevance of these results is more evident when comparing to other works reporting PLA reinforcement with carbon-based materials. Murariu and co-workers observed a 30% increase in Young's modulus on incorporation of expanded graphite at 3 wt% loading in PLA, for 3.1 mm thick test specimens. The results obtained by Wu and Liao showed improvements in tensile strength of about 20% in PLA films after incorporation at 1–3 wt% loadings of chemically treated multi-walled carbon nanotubes in an also chemically modified PLA matrix. Cao and co-workers incorporated 0.2 wt% of reduced GO in PLA, obtaining an 18% increase in Young's modulus for 0.4–0.45 mm thick specimens.
DSC was performed for various filler loadings, with vacuum-dried films. Figure 11 shows representative calorimetric curves obtained for PLA and PLA with 0.4 wt% GO and GNP, after drying in a vacuum oven. In all curves Tg is ca 50–60 °C, immediately followed by a small hysteresis peak, associated with physical relaxation. Above 100 °C, a cold crystallization exothermic peak is visible. Finally, melting takes place at ca 150 °C.
Table 2 gives the values of Tg and Tm for all samples tested. Firstly, it must be noted that Tg of unprocessed PLA is about 7 °C higher than that of solvent-cast PLA. This is due to the formation of free volume within the polymer matrix when solvent is evaporated at 40 °C. It is therefore noted that unprocessed PLA may present different properties from the films studied here, but this study concerns the effect of GO and GNP incorporation taking as reference solvent-cast PLA films.
Table 2. Glass transition temperature and melting temperature for unprocessed PLA and PLA films. Cold crystallization and melting enthalpy are shown only for unfilled PLA (see text)
|Sample||Tg ( °C)||Tm ( °C)||ΔHm (J g−1)||ΔHc (J g−1)||χc (%)|
|PLA (Natureworks)||59.6||152.0||16.2||− 2.7||14.5|
|GO 0.2 wt%||55.5||151.8||—||—||—|
|GO 0.4 wt%||57.1||152.5||—||—||—|
|GO 0.6 wt%||55.8||150.1||—||—||—|
|GNP 0.2 wt%||58.5||151.0||—||—||—|
|GNP 0.4 wt%||59.0||151.5||—||—||—|
|GNP 0.6 wt%||58.0||150.9||—||—||—|
As the films are loaded with nanofillers, Tg increases significantly in relation to the unloaded PLA films, denoting restricted molecular mobility associated with good filler–matrix interaction. A maximum is observed for 0.4 wt% loadings, coinciding with the optimum loading observed for mechanical properties. This effect is more pronounced with GNP (Tg is 2 °C higher at the same loading), probably due to the more planar geometry of the filler yielding more effective confinement of chain segment mobility. Wu and Liao also observed increases in Tg for incorporation of carbon nanotubes in PLA: a 4 °C increase in Tg, but for higher loadings (3 wt%). Tm shows no defined changes. A decrease in Tm would be observed if phase separation occurred, as reported in several works., 
Gas permeability properties
Permeability of the films towards oxygen and nitrogen was measured using the time-lag method. Results are given in Table 3. PLA permeability towards oxygen and nitrogen is close to that previously reported by Komatsuka and co-workers, although permeability can vary depending on PLA type and on membrane manufacturing process. The largest decreases in permeability are obtained for 0.4 wt% of GO and GNP, corresponding to an about threefold decrease in permeability towards oxygen and a fourfold decrease towards nitrogen. For comparison, a twofold decrease in oxygen permeability was reported by Chang and co-workers for the incorporation of 10 wt% of various organic nanoclays in PLA films.
Table 3. Effect of nanofiller incorporation on permeability of PLA films towards O2 and N2
| ||O2 (×10−18 m2 s−1 Pa−1)||N2 (×10−18 m2 s−1 Pa−1)|
|Sample||Permeability||Standard deviation (n = 3)||Permeability||Standard deviation (n = 3)|
|GO 0.2 wt%||1.34||0.0212||0.271||—a|
|GO 0.4 wt%||1.23||0.0141||0.306||—a|
|GO 0.6 wt%||1.49||0.0033||0.502||—a|
|GNP 0.2 wt%||1.34||0.0265||0.270||—a|
|GNP 0.4 wt%||1.20||0.1556||0.250||—a|
|GNP 0.6 wt%||1.30||0.0555||0.237||0.1533|
This gas permeability reduction is probably associated with a barrier effect created by the nanofillers. It can be expected that GNP, having a more planar configuration, would be more efficient in creating a tortuous path for permeation than GO particles. This is not observed, and both fillers show similar effects. This may be related to an absence of orientation of the GNP platelets along the film plane, which does not contribute to increasing the tortuosity in the direction perpendicular to the film plane. A different manufacturing method, e.g. extrusion, may induce an effective difference in the permeation of GNP and GO.