*3.3. BA Distribution*

To measure the distribution of BA particles in the PLA matrix, samples containing different BA concentrations were cryosectioned and viewed under TEM. Figure 5a illustrates that the particles of BA formed more agglomerates in the PLA matrix, and that the particles were not well distributed. Figure 5b shows that the addition of the initiator in the PLA matrix decreased the agglomerations of BA particles and distributed the particles better than in PLA/BA. This is attributed to the fact that the viscosity of the PLA matrix was increased by chain extension and/or branching, which assisted in breaking the BA agglomerates. On the other hand, better distribution of BA particles was observed in the presence of Joncryl, due to more chain branching that was created in the PLA structure. Furthermore, Figure 5d shows that the addition of all components at once produced a fair distribution of BA particles and strong intercomponent bonding. This observation correlates with the FTIR spectroscopy results in the next section, which show the interfacial bonding between all components.

**Figure 5.** TEM images of (**a**) PLA/BA, (**b**) PLA/DCP/BA2, (**c**) PLA/J/BA2, (**d**) PLA/DCP/J/BA2, (**e**) BA3, (**f**) BA4, (**g**) B5, and (**h**) BA6.

Furthermore, it is evident from Figure 5e that upon the addition of 3 wt.% BA to the PLA matrix, the finest distribution was observed, showing an optimal distribution amongst all composites, owing to better intercomponent bonding amongst the neat PLA and BA particles. Das et al. [2] reported similar results in PLA/BA, revealing that the 3 wt.% BA loading in the PLA matrix was the optimal distribution. This led to improved mechanical properties, which were affected by the distribution of the BA particles in the PLA composites. On the other hand,

the increase in BA concentration (i.e., 4, 5, 6, 10, and 20 wt.%) produced poorer distribution and more agglomerations of the BA particles in the PLA matrix.

#### *3.4. Non-Isothermal Crystallization of the Modified PLA Systems*

DSC was used to study the effects of BA, DCP, Joncryl, and the resultant structures on the crystallization and melting temperature of the PLA matrix. The DSC data are summarized in Table 3. The degree of crystallinity (*χm*) during cold crystallization, during heating (*χcc*), and total crystallinity (*χc*) were calculated using the following equations [2,23]:

$$X\_m = \frac{\Delta H\_m}{\mathcal{D}\_{PLA} \quad \Delta H\_m^\circ} \tag{1}$$

or

$$\mathbf{X\_{cc}} = \frac{\Delta H\_{\mathbf{cc}}}{\mathcal{D}\_{\text{PLA}} \quad \Delta H\_{\text{m}}^{\circ}}$$

$$\mathbf{X\_{c}} = \mathbf{X\_{m}} - \mathbf{X\_{cc}}\tag{2}$$

where ∆*H<sup>m</sup>* is the melting enthalpy, ∆*Hcc* is the enthalpy of cold crystallization, ∅*PLA* is the weight fraction of PLA, and ∆*H*◦ *<sup>m</sup>* is the enthalpy of fusion of 100% PLA, taken as 93.7 J/g [33]. Figure 6 shows the DSC thermograms from the second heating. PLA shows diverse transitions; the first transition is related to the PLA *T<sup>g</sup>* at (60 ◦C), the second transition is allied with the *Tcc* at (110.24 ◦C), and the last transition is linked with the *T<sup>m</sup>* at (169.11 ◦C). In addition, Figure 6a also shows various melting temperature peaks at 164.13 and 169.11 ◦C. Moreover, there was no *T<sup>c</sup>* detected for PLA during the cooling cycle, because PLA crystallizes very slowly [34]. Upon addition of BA, and chain extension/branching by DCP/Joncryl, the *T<sup>m</sup>* of the samples moved to the low side of the graph compared to neat PLA. This observation suggests that BA acted as a weak nucleating agent in the samples; meanwhile, 3 wt.% showed a profound shift to the lower side of the *T<sup>m</sup>* peak (165.5 ◦C), suggesting a strong nucleating effect. Similar nucleating effects were reported by Malwela et al. [10] and Das et al. [2]. On the other hand, the *Tcc* temperatures were also affected by the addition of BA, due to the nucleating effect. Upon addition of BA, the *T<sup>c</sup>* value of the composites decreased, moving towards the lower crystallization temperatures, indicating enhanced nucleation. Malwela et al. [10] reported a similar nucleating effect.

Additionally, BA limited the mobility of the PLA macromolecules, restricting their chain arrangement. When the molecular structure of PLA was altered by DCP and Joncryl, the *Tcc* was reduced, and the crystallinity increased. This phenomenon is related to PLA degradation [35–37]. However, when the BA content was increased, the crystallinity in 3 wt.% decreased due to the well-dispersed BA particles causing a physical barrier in the PLA matrix. The *T<sup>g</sup>* of the samples remained unchanged regardless of incorporating BA or alterations to the molecular structure of PLA. This observation is related to the DMA results that will be discussed later. Overall, the *T<sup>m</sup>* and *T<sup>c</sup>* of the composites decreased compared to the neat PLA, confirming that BA, DCP, and Joncryl are good nucleating agents. Small loading of the nucleating agent assisted in forming the polymer crystals; meanwhile, high loading of the nucleating agent restricted the ordered arrangement of the molecular chain, leading to low crystallinity. Moreover, during heating, more crystals were formed; as a result, the PLA crystallinity was improved.

The XRD patterns of neat PLA, BA powder, and composites are shown in Figure 7. The diffraction patterns on the as-received BA powder were observed at 2θ = 13.98◦ , 28.12◦ , 38.36◦ , 49.46◦ , 55.11◦ , and 64.60◦ , attributed to the (20, 120, 031, 200, 002, and 151 crystallographic planes, respectively. In the case of neat PLA, the broad amorphous peak at 16.50◦ was observed and ascribed to the 200/100 crystallographic plane of PLA crystal, consistent with the features of PLA [38–41]. Upon the addition of various BA concentrations to PLA, the features of BA at the peak of interest (2¦È = 13.98◦ ) were also recognized, signifying the presence of the filler in the composites. On the other hand, the XRD patterns of all of the composites show an intensive peak around 16.21◦ , and have

slightly moved to a higher angle, suggesting that the crystal size of the composites has decreased due to the interaction and distribution of BA in the PLA matrix [2]. Chain branching contributes to the higher crystallinity of PLA.


**Table 3.** DSC measurements of neat PLA and composites.

**Figure 6.** DSC traces of the second heating curve of neat PLA and samples containing DCP/Joncryl and BA.

**Figure 7.** XRD patterns of the neat PLA and composite.

Similarly, this tends to reduce the crystal sizes. Therefore, it was important to calculate the crystallite size of the samples, using the Scherrer equation shown below (Equation (3)) as a mathematical expression of the relationship between full width at half maximum FWHM and the crystallite size. The results of the crystalline size for all samples are listed in Table 4.

$$FWHM = \frac{K\gamma}{L\cos\theta} \tag{3}$$

where *FWHM* is the full width at half maximum attained from the instrument, λ is the wavelength of the X-ray that was used for the diffraction, *L* is the crystalline size, *θ* is the peak position (2θ/2) in radians obtained from the instrument, and *K* is a shape factor constant with a value of 0.9 [42]. Neat BA shows a crystalline size of 39.59, and PLA 12.8, with a *T<sup>m</sup>* of 169.11 ◦C from DSC curves. When a polymer is heated at the minimal *T<sup>m</sup>* and the equilibrium *T<sup>m</sup>* (*T<sup>m</sup>* 0 ), the remaining well-ordered structures in the melt will significantly influence the crystallinity [43]. Upon alteration of the PLA structure, the crystal size decreased, and the addition of various BA loading further increased the crystal size with 5 wt.% as the threshold. The small crystalline size is due to the crystal growth of the polymer, which is attained by the extra addition of folded polymer chain segments, meaning that the sample has a lower *T<sup>m</sup>* value [44]. Farid et al. [45] reported a similar observation. Further, the crystal size was not dependent on the BA concentration. In conclusion, we observed that the BA and the resultant branched structure acted as good nucleating agents, and this observation was consistent with the DSC analysis.

**Table 4.** Crystal sizes of neat PLA and composites.


#### *3.5. HDT*

The HDT results of neat PLA and composites are listed in Table 5. The HDT of neat PLA slightly improved from about 0.6 to 1.77 ◦C after the addition of BA and molecular structure alteration with DCP and Joncryl, which led to the enhancement of the mechanical properties of PLA. In the case of BA's inclusion in the PLA matrix, the results show an improvement of 1.7 ◦C; this increase is motivated by the higher degree of BA crystallinity, owing to its nucleating properties. This result is related to the DSC results shown in Figure 6. Upon the molecular structure alteration of PLA by DCP and Joncryl, the HDT increased by 2.0 and 2.7 ◦C, respectively, due to the chain extension and/or branching.

Further, upon the addition of different BA concentrations, the HDT increased with increasing loading. These results are consistent with the MFR test results reported in Table 2. Overall, the DSC and HDT results show that the incorporation of BA and the alteration of PLA's molecular structure enhance the distribution of BA particles and the mechanical properties of PLA.



#### *3.6. Thermomechanical Properties*

The effects of BA distribution on the thermomechanical properties of PLA were examined. Figure 8a shows the storage modulus (E0 ) of the samples as a function of temperature. The E0 of the samples is discussed at two different phases: glassy phase, below the *Tg*, where the polymer chains are highly restricted; and transition phase, at the *T<sup>g</sup>* of PLA (60 ◦C). The glassy phase shows that neat PLA has a very low E0 compared to the composites. The reinforcing effect of BA in increasing the E0 of PLA was noted. However, the addition of DCP further increased the E0 of PLA due to the enhanced distribution of BA and the possibility of branched chains and/or crosslinking, which contributed to the rigidity of the PLA matrix. On the other hand, when Joncryl was added to PLA/BA, it further increased the E0 higher than DCP in PLA/BA, because of the chain extender used to extend the PLA chains.

Furthermore, the concurrent addition of DCP and Joncryl to PLA/BA further increased the E0 . The presence of both DCP and Joncryl induces chemical bonding between PLA and BA, as shown in Figure 4; hence, better distribution of BA, as shown in Figure 5. This results in strong interfacial bonding between PLA and BA; hence, the E0 increases when both DCP and Joncryl are added. The increase in E0 can also be attributed to the increase in crystallinity, as shown in Figure 3. With an increase in temperature, the E0 decreased, as expected. Region 2 illustrates the *T<sup>g</sup>* of all of the samples, as listed in Table 6 and shown in Figure 7; the *T<sup>g</sup>* of all of the samples remained almost the same, indicating no effect of BA distribution on the *Tg*. The *T<sup>g</sup>* of all of the samples examined from the tan delta curve (Figure 7) clearly shows no effect on the *Tg*, although it was expected that the *T<sup>g</sup>* would move to higher temperatures due to the chain restriction in the presence of BA particles. Overall, it is evident that the storage modulus was dependent on the distribution of BA.

**Table 6.** Storage modulus and Tg of neat PLA and composites.


**Figure 8.** DMA plots of (**a**) storage modulus; (**a** 0 ) storage modulus and (**b**) tan delta curve of neat PLA and composites.

#### *3.7. Tensile*

Figure 9 displays the tensile modulus (E0 ) and elongation at break (εba) of the neat PLA, PLA/DCP, PLA/J, PLA/DCP/J/, and PLA/DCP/J/BA composites at various concentrations of BA. The neat rigid PLA shows a high E0 of 2040 MPa. Expectedly, PLA exhibits a low εba of 4.8%. Upon the structural modification of PLA using DCP and Joncryl individually, the εba did not change significantly. However, the E<sup>0</sup> of PLA/DCP was higher than that of PLA and PLA/J. The entanglement of crosslinked structures could have caused this increase.

On the other hand, the structural modification of PLA using both DCP and Joncryl concurrently did not change the εba, while the E<sup>0</sup> slightly increased with respect to neat PLA. Upon the addition of 2–4 wt.% BA to the matrix, the E0 slightly decreased, steadily increasing as the BA concentration increased. At the same time, the εba increased in low concentrations (2 and 3 wt.%), with a decrease at 4 and 5 wt.% loading. We believe that the good distribution of the filler and improved matrix interaction enhanced the stress transfer of the materials [2]. In this case, the better distribution observed in 3 wt.% loading, as shown in Figure 5, did not lead to the highest enhanced E<sup>0</sup> and εba. Overall, the structural modification of PLA and the incorporation of BA in all of the systems did not significantly affect the E<sup>0</sup> and εba of PLA.

#### *3.8. 3DP*

1

Figure 5e revealed that the BA3 sample showed a fair distribution of BA particles compared to other composites. It is noteworthy that the distribution of the filler particles and the polymer used play key roles in the 3DP process. Due to several extrusions involved

in the 3DP process, improved distribution is required in order for heat to be well dissipated in the polymer matrix. Based on that, Supplementary Figure S5 shows the demonstration of the 3D-printed components for neat PLA and BA3 samples. The BA3 sample was chosen from amongst the other composites due to its fair distribution of BA particles in the PLA matrix. The fused deposition modelling (FDM) process based on extrusion technology was used as the technical basis for successfully printing biodegradable PLA/BA composites. A desktop printer (Wanhao Dupilcator i3 plus, Odessa, FL, USA) with a 0.4 mm nozzle was used to produce the 3D-printed specimen. The printing conditions were as follows: nozzle temperature 210 ◦C; bed temperature 50 ◦C; print speed of 60 mm/s; 1 perimeter wall; 2 top and bottom layers; and a layer height of 0.2 mm. Supplementary Figure S4 shows different shapes printed by the 3D printer for neat PLA and BA3 samples. A detailed paper concerning this process will follow.

**Figure 9.** Tensile properties of neat PLA and BA composites with different loadings.

#### **4. Conclusions**

The present work investigated the effects of strategic modification with DCP and Joncryl on the PLA/BA composites, as well as the influence of BA distribution on the thermomechanical, mechanical, and flow properties. We found that the structural alteration of PLA by Joncryl and DCP had a significant effect on the flow properties and control of degradation. Based on the flow properties and FTIR spectroscopic analysis, the mechanism of stabilization is most likely chain extension. The chain extension leads to the long-chain branched structure in the sample containing Joncryl and DCP. The enhanced distribution of BA particles in the PLA matrix, and the interfacial bonding, were responsible for improving the PLA's properties. The mechanical properties of the PLA increased. The 3 wt.% BA showed the optimal distribution, and the PLA/DCP/J/BA3 system was chosen for further studies. The produced composite was indeed 3D printable. For future work, it will be interesting to investigate the influence of BA concentration on the PLA/DCP/J/BA system, in order to understand the effects of BA particle distribution on the thermal properties of the resulting system. In conclusion, the structural alteration with DCP and Joncryl successfully improved the BA particle distribution, leading to enhanced thermomechanical and HDT properties of PLA.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/polym13122019/s1, Figure S1: The mechanism of reaction between PLA and DCP, Figure S2: The mechanism of reaction between PLA and Joncryl, Figure S3: FTIR spectra for: (a) neat PLA, (b) PLA/DCP (c) PLA/J, (d) PLA/DCP/J, (e) PLA/BA, (f) PLA/BA/DCP, (g) PLA/BA, and (h) PLA/BA/DCP/J, Figure S4: FTIR spectra for neat materials of (a) BA, (b) DCP, and (c) Joncryl, Figure S5: The printability of: (a) neat PLA square shape, (b) BA3 square shape, (c) neat PLA CSIR logo, and (d) BA3 CSIR logo.

**Author Contributions:** Conceptualization, D.M.; writing—original draft preparation, D.M.; writing review and editing, V.O., J.B., and S.S.R. All authors have read and agreed to the published version of the manuscript.

**Funding:** The authors thank the Department of Science and Innovation and Council for Scientific and Industrial Research, South Africa, for their financial support.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

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

#### **References**

