Poly (lactic acid) (PLA) is a thermoplastic polyester produced by fermentation of renewable agricultural raw materials followed by polymerization. In this study, it has been used the PLLA (PLA4032D), which contained ~2 wt% D-lactic acid and the average molecular weight was 190,000. It was purchased by Nature Works LLC. The density, melting point (Tm) and glass transition temperature of the PLA4032D were 1.24 g/cm³, 170˚C, and 59.2˚C, respectively. The natural filler used was the Typha stem powder. It was collected on the Senegal River banks, in Bango in the region of Saint Louis. The plant was been cut and dried in the air then it undergoes mechanical grinding to transform the stems into powders.

The Typha stem powder was oven-dried at 105˚C and the PLA at 60˚C for 24 hours to remove moisture. The composites were elaborated in a HAAKE Minilab twin-screw extruder (Thermo Scientific), operated at a screw speed of 100 rpm, at a temperature of 180˚C for 5 min. Four formulations of the composites were prepared with different content of filler equal to 0%, 25%, 35%, 45% in mass of TPS. The composites were then compression molded using a Carver laboratory press at 180˚C and 200 bar for 1 minute to obtain a film of about 150 µm. The film is used for the mechanical test.

A HIROX SH4000M scanning electron microscope (SEM) was used. The samples were fractured in liquid nitrogen and after sputter-coated with gold under an argon gas flow for 15 min. The coated samples were left to dry at room temperature before SEM could be performed. The SEM analyses were done at 20 kV and at a range of magnification equal at ×80 to ×150.

The viscoelastic behavior of molten polymers can be determined using oscillatory rheological experiments such as dynamic mechanical testing, which offers a convenient way to assess the frequency dependence of mechanical properties of polymers. A useful representation is to plot the experimental frequency sweep data points in the complex plane. That means that imaginary part (of the complex viscosity) η' values are reported along the abscissa (X axis) and the real ones (η'') on the ordinate (Y axis) as schematized in Figure 1; where η' is the loss viscosity, η'' is the storage viscosity and ω is the pulsation of the frequency: ω = 2πN. Usually, the experimental points are located on an arc of circle characteristic for a Cole- Cole distribution. The extrapolation of this arc of circle to the zero ordinate value gives the Newtonian viscosity which is related to the average molecular weight Mw of the considered polymer through a power law: (η₀ = K (Mw)^3.4) [17].

Melt rheological measurements of neat PLA and PLA composites were carried out with an ARES Rheometer Scientific mechanical spectrometer in oscillatory frequency sweep mode with a parallel-plates measuring cell. The diameter of the plates was 8 mm and the gap was 1 - 1.5 mm. The frequency was ranged from 0.1 rad/s to 100 rad/s. Deformation was kept constant at 5% over the entire frequency range to ensure linearity. This strain was selected from a dynamic strain sweep test, in which, within 1% to 10% of strains, at a fixed frequency of 10 rad/s, the drift deviation from linearity was followed; the frequency sweep test was then carried out at a constant temperature at 180˚C.

The apparatus used is a Lloyd EZ 50 traction bench. The results are analyzed using the software: NEXYGEN. The tensile strength and elongation at break of the composite were measured according to the ASTM D638-10 method at a crosshead speed of 30 mm/min. Each sample was subjected to at least 5 repetitive tests to obtain an average.

The cross-sectional topography of PLA/TSP composites and TSP were observed by scanning electron microscopy. It can be observed from Figure 2(a) the differences in shape and size of the particles which constitute the TSP. The shape of powder ranges from particles of approximately spherical geometry lower than 50 microns, going up to filaments more than 1 mm. The large distribution of the particles, in terms of size and shape, reflects on the dispersion in the PLA matrix. As shown in Figures 2(b)-(d) a randomly orientation in the three directions of space and a phase separation between the TSP and matrix was rough with grooves and holes. In fact, in the tensile test, some TSP fibers could be extracted directly from the PLA matrix, indicating that the poor mechanical properties mainly resulted from the poor interfacial adhesion (mechanical test-Section 4.4). In addition, the presence of gaps between the pores and the two phases indicated that the motion of PLA molecules was restricted by poor interfacial compatibility [19].

Figure 5 represents the complex plane variations for the neat PLA and the PLA composites. As introduced above, an extrapolation of the circle arc constructed on the complex graph allows us to determine the Newtonian viscosity. Table 2 reports the Newtonian values determined for all the samples. These values show an increase in the viscosity of PLA 25TSP that is 76% higher than virgin PLA. Though, an increase in the quantity of the charge causes a reduction of the viscosity, reaching a Newtonian viscosity value, for the sample PLA 45TSP, comparable with that measured for PLA Pure with a difference of about 6% more. The changing of the viscosity is caused by the perturbation of the normal polymer flow since the filler particles hinder the mobility of chain segments. Moreover, great changes are observed for the samples loaded regarding the reference

PLA Pure. All the samples with the natural fillers exhibit a deviation to the circular behavior promoting the shift of the Newtonian plateau to the low frequencies. This is characteristic of a gel behavior, indicating that at different concentrations of the powder, the level of interaction between the matrix and the filler is different at different frequencies applied [23]. In fact, at low frequencies, the viscosity strongly increases with the Typha powder content, as evidenced by the storage module of composites (G') (Figure 6) since G' is higher than pure matrix due to the intrinsic rigidity of TSP. Shear-thinning behavior of the molten mass was observed, the viscosity of the composites is strongly influenced by the shear rate. The storage modulus, especially for the highest powder content, shows that the matrix transfers great stress to the filler and based on G' data, serving as the measure of elastic behavior, we can conclude that with the addition of the TSP the solid-like behavior for all the samples clearly increases [21].

The mechanical tensile properties (Young’s modulus, maximum strain, ultimate tensile strength) of the biocomposites are presented in Table 3 and Figure 7. It

is noted a decrease in Young’s modulus of biocomposites containing 25% and 35% of TSP compared to virgin polymer. However at 45%, there is a clear increase of the module due to the intrinsic rigidity of TSP and also with the high wood content (the latter having high tensile properties), these tensile properties of the composite were improved. Decreases in the modulus observed for samples containing 25% and 35% could be caused by the inhomogeneity of the powder consisting of particles as observed in the morphological characterization section. This lowering is certainly related to the formation of aggregates and the piling up of the fibers against each other, phenomena that favor the formation of defects. At high fiber proportions, the molten mixture is very viscous and appears as “dry” and this promotes friction between fibers and contributes to the creation of pores and cavities at the interface. As a result, the fiber/matrix contact area is reduced resulting in poorer stress transfer.