**4. Results**

After the preparation of the specimens for investigation regarding creep behavior, the blends and the pure polymers were characterized by infrared ATR spectroscopy (Figure 1).

The pure PLA showed an infrared spectrum with main bands at 2996 and 2946 cm−<sup>1</sup> (stretching C–H), 1746 cm−<sup>1</sup> (stretching C=O), 1180 cm−<sup>1</sup> (stretching C–O–C), and 1082 cm−<sup>1</sup> (stretching O–C–C–) [52,53]. PBSA, being, like PLA, an aliphatic polyester, showed similar bands, but shifted at lower wavenumbers because of the higher macromolecular flexibility of PBSA [54], being 2943, 2867, 1725, 1161, and 1043 cm−1, respectively. The blends PLA/PBSA 85/15 and 80/20 blends showed a spectrum much more similar to pure PLA, but some differences could be noticed. In the blend, the main stretching C=O band at 1746 cm−<sup>1</sup> showed a shoulder at lower wavenumbers attributable to the presence of the C=O stretching band of PBSA. Moreover, a similar trend could be observed for the band at 1180 cm−1, showing a shoulder at lower wavenumbers due to the overlapping of the 1161 cm−<sup>1</sup> band of PBSA. The 955 cm−<sup>1</sup> was more intense in blends because PBSA showed this band, attributable to the –C–OH bending in the carboxylic acid groups of PBS [55] or vinyl esters [56] more intense than PLA. Moreover, the band at 806 cm−<sup>1</sup> (present in the spectrum of pure PBSA at 840 cm−1) was attributable to the presence of vinyl ester moieties [57]. In general, the ATR evidence resulted in good agreement with the selected compositions of PLA/PBSA blends.

**Figure 1.** (**a**) ATR spectra of pure PLA and PBSA; (**b**) ATR spectra of PLA-PBSA blends.

In particular, the infrared characterization did not show any evident chemical change in the two polymers; this is can be ascribed to the very short processing time. In fact, significant transesterification occurs in the presence of proper catalysts and for processing times longer than 10 min [58]. Moreover, in the literature, it has been observed by Ding et al. [59] that the formation of copolymers between PLA and PBSA during extrusion and injection molding can be considered negligible. In addition, any eventual degradation during processing can be excluded due to the processing temperature adopted that did not overcome 190 ◦C. For the pure polymers and their blends, in fact, it was observed that the onset temperature for pure PLA was 274 ◦C, and the temperature at which the

maximum degradation rate was reached (inflection point) was 354.5 ◦C; while for PBSA, these two temperatures were shifted at 301 and 401 ◦C, respectively, indicating a higher thermal stability for this polymer [60–62].

From the tensile tests (reported in Table 2), it can be observed that the addition of PBSA makes the material more ductile. Pure PLA is a fragile material characterized by an elevated modulus of 3.58 GPa, a high tensile strength of 62 MPa, and a low elongation at break (3.6%). The mechanical results are in accordance to what can be observed in the literature [28,62,63]; the introduction of the rubbery PBSA phase at 15 and 20 wt.% led to a decrement in the Young's modulus and tensile strength counterbalanced by an increment of the elongation at break. These trends are more marked where the PBSA content was higher (20 wt.%). With the addition of PBSA, the material yielded with the appearance of the neck that propagated along the gauge length.


Standard creep curves at different stress levels are reported in Figure 2.

At stresses of 10 and 20 MPa, all the specimens did not break during the test time interval (set at 8 h). However, when increasing the PBSA content, the creep resistance decreased. In fact, pure PLA did not break even at 30 MPa, while the 85-15 and 80-20 blends began to break (8 h before) already at 30 MPa. Moreover, at the same stress level applied, the time at which breakage occurred decreased with the increase in the rubber content.

The results of the volume variations recorded after that the specimens reached the set stress levels (reported in Figure 3) showed an almost linear variation in the volume change with time. The volume variation at time zero (ΔV/V0,i) corresponded to the intercept of the *y*-axis in the ΔV/V0 vs. the axial strain curve. ΔV/V0,i can be linked to the volume variation caused by the instantaneous elastic deformation of the specimen at the selected stress level applied. Increasing the stress level, the instantaneous volume change increased in accordance with the increment in the instantaneous elongation.

The volume change did not increase proportionally with the rubber content but, on the contrary, a greater volume variation was encountered for the 85-15 blend. This trend can be observed both from Figure 3 but also from the volume variation at time zero reported in Table 3 in which the slopes of the straight lines passing through the experimental data are also reported. For pure PLA, the slope remained almost constant with a slight increment with the stress level. For the 85-15 and 80-20 blends, at higher stresses, the volume strain increased more rapidly with higher slopes values. The introduction of PBSA made a substantial contribution to the dilatational processes that were more marked for the 85-15 composition.

In order to better understand the volume variation curves, it is necessary to investigate deeper the micro-mechanics of the deformation process and how the PBSA addition influences the micromechanical behavior of the binary blend.

For all compositions, it can be observed in Figure 4 that the linear region of the creep curves can be estimated, with a good fitting, by the Andrade equation (Equation (1)); consequently, for all the stress levels applied, the Andrade b parameter for pure PLA and its binary blends can be obtained, making the linear regression of the ε(t) against t1/3 plots.

**Figure 2.** Axial elongation vs. time (creep curves) curves of pure PLA (**a**), 85-15 (**b**), and 80-20 (**c**) for different applied stresses from 10 to 45 MPa.

**Figure 3.** Volume variation vs. axial strain curves (dilatometric curves) of pure PLA (**a**), 85-15 (**b**), and 80-20 (**c**).


**Table 3.** Volume variation at time zero and the slopes of the ΔV/V0 vs. the axial strain curve at different stress levels.

Applying the Eyring relationship (Equation (2)) between the applied stress, *σ*, and the *b* parameter, interesting results can be obtained plotting (Eyring plot) *log b* against the applied stress, *σ*. It can be observed from Figure 3, that the Eyring plot for pure PLA gives a straight line in accordance with the Eyring model for stress activated flow. This result is also in accordance to what was observed for pure PA66 in a rubber-toughened polyamide 6,6 system [48]. For pure matrix, due to the absence of rubber PBSA particles, the stress concentration factor, γ, can be assumed equal to 1, and the activation volume, *V,* for pure PLA can be easily obtained from the Eyring plot slope. An apparent activation volume of 0.25 nm<sup>3</sup> was obtained.

Under 20 MPa and 30 MPa for the 85-15 and 80-20 blends, respectively, the Eyring plots behaved in a similar manner to pure PLA (following Equation (2)) but with a slightly increased slope. An explanation for this behavior was found to be in accordance with the literature [48] and was ascribed to the same deformation mechanisms that operate in both pure PLA and PLA–PBSA blends having the same value of activation volume, *V*. The higher slope was ascribed to an increased value of the stress concentration factor, γ. On this basis, using the value of the activation volume found for pure PLA, the stress concentration factor for the PLA–PBSA binary blends can be easily obtained from the Eyring slope until 20 MPa and 30 MPa. For the two blends, 85-15 and 80-20, the stress concentration factor values obtained were very close to each other (γ = 1.15 for 85-15 and γ = 1.16 for 80-20); consequently, the γ seemed not to be affected directly by the difference in volume variation observed for the 85-15 and 80-20 blends.

However, a very interesting difference emerged from the Eyring plot (Figure 5) in which the slope changed for the binary blends occurring at two different stress levels. A small increase with no slope change could be detected over the lower range of applied stress (up to approximately 18 MPa), where, according to the literature [48], shear yielding is the predominant deformation mechanism. The stress level for which the change in slope took place can be easily identified using a simple geometric construction (the intersection point is highlighted by the green and orange arrows in Figure 5). The slope change occurred earlier (close to 20 MPa) for the 85-15 blend, while it occurred later (at approximately 25 MPa) for the 80-20 blend. The different intersection points, registered for the two binary blends investigated, were strictly connected to the volume variation differences encountered for the two types of blends. The intersection point allowed for

the identification of the stress level for which the micromechanical process of debonding and/or cavitation starts and contributes to deformation. Based on the results observable in Figure 5, the debonding and/or cavitation mechanism was apparently activated earlier (at a lower stress level) for the 85-15 blend.

**Figure 4.** Andrade equation fittings (axial elongation vs. cubic root time) of pure PLA (**a**), 85-15 (**b**), and 80-20 (**c**).

**Figure 5.** Andrade B parameter in which is highlighted the different slope changes between pure PLA and its blends (85-15 and 80-20).

In fact, it must be taken into account that the creep test was carried out at a constant rate of 10 MPa/min to achieve the set stress levels. Consequently, a slightly different strain rate (of approximately 1mm/min) caused by the higher deformability of the blend that increased with PBSA content was registered. Nevertheless, the application of this method was extremely useful, because it allowed to highlight the connection between the micromechanical processes and their volume variation caused by the introduction of a rubber dispersed phase into a polymeric matrix. The voids generated by the cavitation/debonding mechanism caused the increase in the apparent activation volume, since voids allow plastic flow of the matrix around it more easily than an intact rubber particle. In fact, the bulk modulus of a rubber particle is very high, while for a void it is zero. Thus, the increased local strain rate of the polymer matrix was due to the fact that the material was no longer a "real" continuum on a microscopic scale. The deformation involves a macroscopic volume increase due to the growth of the voids generated around or inside the rubber particles, which is favored by the high level of triaxiality, leading to a nonlinear yield curve. In contrast to the case of a "continuum" polymer, where high levels of triaxiality favor crazing and cleavage mechanisms over shear yielding, for a "porous" polymer, a high triaxiality considerably accelerates plastic flow.

In order to confirm the results obtained and to understand if PBSA particles undergo debonding or cavitation, SEM micrographs (at 8000 X, Figure 6) at the surface of the tensile specimen, cryo-fractured along the draw direction and at different stress levels, were carried out.

On the basis of the results obtained in Figure 5, the most significant stress levels were analyzed. In particular, three stress levels were chosen: (I) 10 MPa, which was the lowest value for which both the binary blends underwent the same deformation mechanism and no deviation in the Eyring's slopes occurred; (II) 30 MPa in which the slope variation occurred for both 85-15 and 80-20; (III) 45 MPa, which was the highest stress level tested and for which the micromechanical deformation process should be more marked.

**Figure 6.** SEM micrographs made at the surface of tensile specimen cryo-fractured along the draw direction at different creep stress levels.

From the SEM images shown in Figure 6, it emerges that the PBSA particle size increases with the PBSA content, in agreement to what is also observed in the literature [28]. At 10 MPa, for both blends, neither cavitation nor debonding of the PBSA particles appeared. At this stress level, only shear yielding took place. The fracture surfaces were characterized by well dispersed PBSA spherical particles that were well attached into the PLA matrix. The matrix showed signs of deformation caused by the creep test with

greater matrix deformation occurring for the 80-20 composition in which the rubber amount was higher.

At 30 MPa, it was evident that the micromechanical deformation that occurred was caused by the debonding of the PBSA particles. Clear visible voids around the PBSA particles could be distinguished. At 30 MPa, therefore, despite 80-20 having a higher PBSA content, the 85-15 blend had the greatest number of particles that underwent debonding. This result is in agreement to what emerged from the Eyring's plot; in fact, for the 85-15 composition debonding started at a lower stress level, generating at 30 MPa a greater number of debonded particles. By increasing the stress level of the creep test (up to 45 MPa), debonding continued to be the main micromechanical deformation mechanism for both the binary blends. Increasing the stress level, the quantity of particles that underwent debonding also increased.

#### **5. Conclusions**

Over the last years, there has been a growing interest toward biobased and biodegradable polymers that show a more environmentally friendly end of life and, at the same time, are able to improve carbon neutrality when compared to their fossil-based counterparts. In order to improve the physical and mechanical properties of these biopolymers, a blending technique is fundamental. For this purpose, a better knowledge of the micromechanical deformation processes of these new biopolymeric blends is fundamental in order to better exploit their peculiar properties, favoring the replacement of their fossil-based counterparts in several application sectors.

Despite different studies that can be found in the literature on biopolymeric blends, an in-depth investigation of micromechanical mechanisms has not yet been considered. In this study, poly(lactic acid) (PLA) and poly(butylene succinate-adipate) (PBSA) blends were investigated with a PBSA content of 15 and 20 wt.%, respectively (named 85-15 and 80-20). Measurements of the volume strain, using an optical extensometer, were carried out with a universal testing machine in creep configuration trying to determine, accompanied by SEM images, the micromechanical deformation processes involved in the biopolymeric blend systems. Analytical models were also applied to correlate the creep to the dilatation variation; in particular, the Andrade equation was applied and the *b* parameter for the polymeric systems was calculated. Plotting the *log b* against the applied stress, using an Eyring plot, a significant change in the curves was found. The binary blends, in fact, showed a sharp increase in *dlogb/dσ* where significant dilatation began. The point at which the slope change occurred reasonably coincided with the onset of the cavitation/debonding mechanism. For both binary blends, the SEM analysis evidenced that the starting micromechanical deformation mechanism responsible for the volume increment was due to the PBSA debonding.

Another interesting result obtained by applying the Eyring relationship was the calculation (knowing the data of the pure PLA matrix) of the stress concentration factor; it emerged that the second PBSA phase acts as a stress concentrator, probably due to the presence of weak interfaces between PLA and PBSA that are also responsible for the debonding mechanism.

**Author Contributions:** Conceptualization, A.L. and L.A.; methodology, L.A. and V.G.; validation, L.A. and V.G.; investigation, L.A. and V.G.; resources, A.L. and M.-B.C.; data curation, L.A., M.-B.C. and V.G.; writing—original draft preparation, L.A., V.G., and M.-B.C.; writing—review and editing, M.-B.C., and A.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** Centre for Instrumentation Sharing—University of Pisa (CISUP) is thanked for its support in the use of FEI Quanta 450 FEG scanning electron microscope. BIONTOP project (grant number: G.A 837761) is thanked for the helpful suggestions regarding the present manuscript.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Acknowledgments:** Centre for Instrumentation Sharing—University of Pisa (CISUP) is thanked for its support in the use of FEI Quanta 450 FEG scanning electron microscope. BIONTOP project (grant number: G.A 837761) is thanked for the helpful suggestions regarding the present manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.
