**1. Introduction**

Interest in polylactide (PLA) has grown both industrially and academically due to its biocompatibility, biodegradability, and transparency [1]. Consequently, it has found application in the automotive, packaging, and medical industries, amongst others [2]. Regardless of its excellent properties, it also has limitations, such as poor thermal stability, brittleness, and slow crystallization during processing [3–6].

Generally, there are ways of improving the properties of PLA, including the addition of fillers or blending with other polymers [3,7–10]. For instance, Nieddu et al. [11] reported that 5 wt.% nanoclay increased the modulus of neat PLA. Other fillers, such as nano-silicon dioxide (SiO2) [12] and boehmite (BA) [10], have been found to improve the thermal stability, crystallizability, and mechanical and electrical conductivity of the polymers. Amongst these, BA has attracted more research interest due to its low cost, high flammability, high surface area, noble dispensability, and thermal stability [10,13]. Moreover, BA particles in nanometric form have greater reactivity than in micrometric form, owing to the increase of their surface area in nanometric dimensions [2]. A known challenge is the distribution of the hydrophilic BA in the hydrophobic PLA matrix, which leads to poor interfacial bonding between the polymer and the filler.

**Citation:** Makwakwa, D.; Ojijo, V.; Bandyopadhyay, J.; Ray, S.S. Flow Characteristics, Mechanical, Thermal, and Thermomechanical Properties, and 3D Printability of Biodegradable Polylactide Containing Boehmite at Different Loadings. *Polymers* **2021**, *13*, 2019. https://doi.org/10.3390/ polym13122019

Academic Editors: Antonella Patti and Domenico Acierno

Received: 7 May 2021 Accepted: 2 June 2021 Published: 21 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Several studies have been conducted on the modification of BA to improve the interfacial bonding between the polymer and the filler [10,13–18]. For instance, Malwela et al. [10] reported the impact of BA surface modification with *p*-toluene sulfonic acid on the mechanical, thermal, and rheological properties of PP/PS blends. Their results revealed that the addition of 1, 3, 5, and 7 wt.% modified BA particles in the PP/PS blend improved the modulus by 5.3, 13.5, 20.3, and 27.5%, respectively, due to the uniformly dispersed agglomerations of BA particles within the PS phase, whereas in the case of untreated BA, slight improvements of 0.2, 11.4, 17.8, and 25.7%, respectively, were observed, attributed to the agglomerations within the edges of the PS phase. On the other hand, the inclusion of 1 wt.% untreated BA particles had no effect on the crystallization temperature, while the crystallization temperature peaks moved from 116.90 to 119.0, 126.6, and 127.7 ◦C for the 3, 5 and 7 wt.%, respectively. Khumalo et al. [17] incorporated BA particles in low-density and high-density polyethylene (LDPE and HDPE). The authors revealed that 2 wt.% BA particles improved the resistance of the resulting composite to thermo-oxidative degradation from 0.5 to 2%. In 10 wt.% BA the improvement was from 1.5 to 3%, due to the presence of BA being reported for polyoxymethylene. Das et al. [2] reported the mechanical, thermal, and fire properties of biodegradable PLA/BA composites. Their results revealed that the incorporation of 3 wt.% BA nanoparticles increased the tensile strength of PLA by 57%, and the cold crystallization was observed in the range of 120–125 ◦C. However, this kind of modification includes a solvent, which has drawbacks, such as high cost and environmental problems.

One of the most common approaches to improving the distribution of the filler in the polymer matrix is through polymer modification [14,16]. Several studies on the interfacial modification of different polymers have been reported [1,19,20]. These modifications include Joncryl as a chain extender and DCP as an initiator. Joncryl is a polymeric chain extender with a low epoxy equivalent weight that reacts with the chain ends of polycondensates and effectively increases their melt viscosity. Additionally, it is a multifunctional oligomer chain extender that was intended to reverse the degradation in PLA; it has epoxy functional groups, which would react with the carboxyl and hydroxyl groups [1,20–24]. Joncryl has shown improvement in the interaction between carboxyl groups of PLA and the reactive functional epoxy group of the chain extender, and consequently, the enhancement of the crystallizability and mechanical properties of PLA [25]. Generally, DCP is a free radical generator in a polymer system, with the possibility of crosslinking or chain branching. Similarly, Joncryl helps with chain branching and extension in polyesters, such as PLA.

This study focuses on producing the PLA/BA composite in order to demonstrate the 3D printing of the samples. This study aimed to use the novel system to study the mechanical, flow, and thermal properties of PLA. We used DCP as a free radical generating agent in the reactive extrusion in order to enhance the amount of macroradicals that could introduce long-chain branching in the PLA chain. Joncryl, a patented, multifunctional, reactive polymer with improved thermal stability/chain extenders for specific food-contact applications, and polycondensation polymers including poly(ethylene terephthalate), was used as a crosslinking agent to extend the PLA chain and to improve the crystallizability of the PLA. Then, we demonstrated the 3D-orientability of the sample, with good distribution. This study aims to improve the BA particles' distribution in the PLA matrix using chain extension and branching, which will lead to enhanced mechanical properties.

#### **2. Materials and Methods**

#### *2.1. Materials*

The PLA used in this work was a commercial grade (PLA 4032D) purchased from NatureWorks LLC (Minnetonka, MN, USA), with a melt flow index (MFI) of 6 g/10 min (2.16 kg load) at 190 ◦C and a density of 1.23 g cm−<sup>3</sup> , while the BA powder was of a commercial grade manufactured by SASOL, under the trade name Dispersal 40, containing 80% Al2O3, donated by SASOL Germany. DCP was obtained from Sigma-Aldrich (Johannesburg, South Africa), with a molecular weight of 270.37 g/mol, density of 1056 g/mL, vapor

pressure of 15.4 mmHg, and a melting point between 39 and 41 ◦C, the chain extender Joncryl ADR 4368 CS was donated by BASF South Africa. The chemical structures of DCP and Joncryl are shown in Figure 1.

**Figure 1.** Chemical structure of dicumyl peroxide (DCP) and Joncryl.

#### *2.2. Preparation of the Samples*

Prior to processing, PLA was dried at 80 ◦C under vacuum for 12 h. The samples with different compositions (Table 1) were melt compounded in a co-rotating twin-screw extruder from Thermo Scientific, Waltham, MA, USA, with an L/D of 40. The extruder conditions were as follows: feeding rate 5.6 g/min; screw speed 202 rpm; and barrel temperatures from the hopper to the die were 140, 160, 180, 180, 180, 180, 180, 180, and 190 ◦C, respectively. The composites were then compression molded into different specimens using a Carver compression molder (Carver laboratory Model 973214A, Wabash, IN, USA) at a temperature of 190 ◦C and pressure of 1 MPa for 6 min, then cooled to room temperature.

**Table 1.** Sample names and compositions.


#### *2.3. Characterization*

A Fourier-transform infrared (FTIR) spectroscope from Perkin Elmer (Model: Spectrum 100, Branford, CT, USA) was used to verify the chemical interactions between PLA, DCP, Joncryl, and BA within the wavelength range of 500–4000 cm−<sup>1</sup> . For all of the spectra, 32 scans were collected, with a resolution of 4 cm−<sup>1</sup> .

The compression-molded disc specimens were used for XRD measurements using an X'pert PRO diffractometer from PANalytical (EA Almelo, The Netherlands). The operating voltage was 45 KV, and the current was 40 mA. The exposure time and the scanning rate used were 29 min 45 s and 0.011◦/min, respectively.

The distribution of BA particles in the PLA matrix was investigated using transmission electron microscopy (TEM) (JOEL, JEM 2100, Tokyo, Japan) with an acceleration voltage of 200 kV. The samples were prepared using a Leica (Austria) EM FC6 cryo-ultramicrotome at −100 ◦C, a cutting speed of 3 mm, and a feed rate of 80 nm. The samples were sliced using a diamond knife.

Differential scanning calorimetry (DSC) measurements were studied using a DSC-Q2000 instrument from TA Instruments, New Castle, DE, USA. Pellets with a mass of approximately 4–5 mg were heated from −20 ◦C to 190 ◦C at a rate of 10 ◦C/min, and then maintained at that temperature for 5 min. Samples were cooled to −20 ◦C at a rate of 10 ◦C/min and kept constant for 5 min, then heated to 190 ◦C at a rate of 10 ◦C/min. The heating and cooling cycles were conducted under nitrogen as the purge gas, with a flow rate of 25 mL/min for all samples. The glass transition temperature (*Tg*), melting temperature (*Tm*), enthalpy of fusion (∆*Hm*), crystallization temperature (*Tc*), cold crystallization temperature (*Tcc*), and enthalpy of cold crystallization (∆*Hcc*) were obtained.

The dynamic mechanical analysis (DMA) was conducted using a PerkinElmer DMA (Model 8000, Branford, CT, USA) analyzer in dual cantilever-bending mode. The temperature was measured at a frequency of 1 Hz, strain amplitude of 0.01%, and heating rate of 2 ◦C/min in the temperature range of −80 to +115 ◦C.

Heat distortion temperatures (HDTs) were measured using a CEAST HDT-VICAT instrument, and the measurements were recorded using the following conditions: Oil bath preheated to 30 ◦C; Tstart = 30 ◦C; heating rate (φ) = 120 ◦C/h; Tmax = 100 ◦C; end = 0.34 mm; span = 64 mm; and stress = 450 kPa.

The melt state rheological properties of neat PLA and composites were investigated using an Anton Paar stress/strain-controlled rheometer Physica MCR501 (Garz, Austria) with parallel plates of 25 mm in diameter. The injection-molded disc samples were used for this test. Frequency sweep tests were carried out from 0.1 to 100 rad/s. Each sample was melted in a parallel plate at 190 ◦C for 5 min to remove the remaining thermal history, and a dynamic strain sweep was then performed in order to determine the common linear region. The melt flow rate (MFR) properties of the neat PLA and composites were investigated using a melt flow meter (multiweight). The pelletized samples were used for this test, weighing 4 g per sample.

Tensile tests were performed in order to determine the modulus, yield strength, and elongation at break of each material, using an Instron 5966 tester (Instron Engineering Corp., Norwood, MA USA) with a load cell of 10 kN, in accordance with ASTM 638D standards. The test was carried out under tension mode at a single strain rate of 5 mm/min at room temperature. The dog-bone-shaped specimens were analyzed.

#### **3. Results**

#### *3.1. Flow Properties*

The melt flow rates (MFRs) of the prepared samples are shown in Table 2. PLA shows a higher MFR, implying lower viscosity and ease of processability. The addition of DCP and Joncryl individually decreased the MFR of PLA by 32.3 and 50.2%, respectively. A decrease in the MFR of PLA/DCP could be attributed to chain–chain coupling, which results from the interaction of radicals generated by DCP on PLA chains, as shown in Supplementary Figure S1. On the other hand, Joncryl as a chain extender could have increased the molecular weight and viscosity of PLA due to the reaction between the epoxide groups of Joncryl and the carboxylic groups of PLA (see Supplementary Figure S2) [25]. The FTIR spectroscopy shows the reactions between the chain extenders, and will be discussed in the FTIR section. The simultaneous addition of both Joncryl and DCP did not significantly affect the MFR of PLA/DCP/J compared to the PLA/J system. Upon the addition of BA at 2 and 3 wt.% to PLA/DCP/J, the MFR decreased slightly, by 12.7 and 8.5%, respectively, suggesting the immobilization of PLA chains by the nanoparticles. However, with further

increase in BA particles, the MFR started to increase, probably due to the separation and weakening of PLA chains by the rigid BA particles.


**Table 2.** The MFR of all samples.

DCP and Joncryl have a significant influence on the structural properties of polymers. In particular, the molecular weight (MW) and distribution (MWD) can be affected due to chain scission or chain–chain coupling in the presence of DCP. In addition, chain extenders such as Joncryl can also modify the structural properties of a polymer. Rheology is a powerful tool to elucidate the changes in the molecular structures of polymers in melt states. Figure 2 shows the plots of G0 and G" against angular frequency from 0.1 to 100 rad/s. In these plots, the crossover point between G0 and G" provides information about the changes in the M<sup>W</sup> and MWD. The horizontal shift of the crossover frequency (Gω) to lower values is related to an increase in MW. In contrast, the vertical shift of the crossover modulus (Gm) indicates an increased broadening of the MWD. It is worth mentioning that the crossover point is determined within the tested range (0.1 to 100 rad/s). For neat PLA (Figure 2a), the crossover point could not be determined within the tested range. However, looking at the distance between G0 and G" at 100 rad/s, it can be noticed that the G0 and G" curves are closer, suggesting lower angular frequency for the PLA/DCP (Figure 2b) system compared to neat PLA.

Moreover, a dramatic decrease in G<sup>ω</sup> was noticed in the PLA/J system (Figure 2c), indicating an increase in the M<sup>W</sup> of PLA. A horizontal shift to higher values could be noticed in the PLA/DCP/J system (Figure 2d); this suggests that there was a reduction in the M<sup>W</sup> of PLA, which could be attributed to chain scission when both DCP and Joncryl were added. The incorporation of BA did not significantly change the M<sup>W</sup> of the PLA/DCP/J system (Figure 2e–i), as can be seen from the distance between G0 and G" at 100 rad/s.

Figure 3a shows the complex viscosity of the prepared PLA/DCP/J-based composites against angular frequency. As evident from Figure 3, the flow behavior of PLA, PLA/DCP, PLA/J, and PLA/DCP/J followed a similar trend to that noted from the MFR analysis. However, PLA/J showed the highest viscosity due to an increase in the M<sup>W</sup> when Joncryl was introduced to PLA. This observation is attributable to the branches formed by the Joncryl and the introduction of long-chain branching (LCB) in the PLA structure, conveying pseudosolid-like behavior. The viscosity of the composites was solely dependent on the distribution of BA particles in the PLA matrix. The BA3 system showed better distribution, and exhibited a higher viscosity than the other composites (Figure 3b). With the increase in filler concentration, the viscosity decreased, possibly due to the poor distribution and agglomeration of BA particles, forming the weak points in the matrix. Further increase in BA concentrations resulted in an increase in viscosity, due to the reinforcing effect of the particles.

**Figure 2.** Storage and loss modulus curves of the PLA and BA composites at different contents.

**Figure 3.** (**a**) Complex viscosity curves of neat PLA and BA composites at different loadings, and (**b**) viscosity at 1 frequency (rad/s) BA composites at different loadings.

#### *3.2. Chain Extension*

FTIR spectroscopy was used to compare the processed PLA/BA chain extender and crosslinking systems, in order to identify the reaction that may have transpired between the PLA/BA and chain extender/crosslinking. Figure 4a shows the FTIR spectra of neat PLA and composites. In the case of neat PLA, the peaks at 1754, 1455, and 1369 cm−<sup>1</sup> are due to C=O stretching, C–H deformation, and C–O–H bands, respectively. Meanwhile, 1186 and 1080 cm−<sup>1</sup> are assigned to –C–O stretching, 868 cm−<sup>1</sup> is attributed to –C–C stretching, and 755 cm−<sup>1</sup> is attributed to C–H bending. The H–O–H peak at 1630 cm−<sup>1</sup> was not obtainable from the processed PLA, due to the existence of thermal chain scission at the C–O bond [26]. Supplementary Figure S3 shows the vibrations at 2852 and 3000 cm−<sup>1</sup> assigned to the O–H stretching, and 2922 cm−<sup>1</sup> due to the axial C–H stretching bond [27]. The FTIR spectra of neat BA (Supplementary Figure S4a) reveal that the vibrations at 3309 and 3095 cm−<sup>1</sup> relate to the O–H stretching of BA [28]. The vibration at 1397 cm−<sup>1</sup> is attributed to the amorphous surface structure that exists in crystalline BA [29]. In the case of neat DCP, as shown in Supplementary Figure S4b, vibrations are observed at 1727 cm−<sup>1</sup> due to the C=O stretching, at 910 cm−<sup>1</sup> due to C=C stretching, and at 762 cm−<sup>1</sup> and 698 cm−<sup>1</sup> due to C–H bending. Supplementary Figure S4c shows that in the FTIR spectra for neat Joncryl, vibrations are observed at 907 and 843 cm−<sup>1</sup> , attributed to the symmetric and asymmetric ring deformation of cyclic epoxide [30,31].

**Figure 4.** 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 samples.

Figure 4e shows that there is no significant change to the PLA matrix upon addition of BA, due to the low concentration of BA embedded in the PLA matrix. Further, upon addition of DCP to the PLA/BA composite, as shown in Figure 4f, DCP undergoes homolytic cleavage when heat is applied, breaking down into free radicals and assisting in the removal of H's from the PLA chains in order to create free radicals on the backbone structure of PLA, as shown in Supplementary Figure S1. This phenomenon is attributed to the propagation of the radical reaction to form a crosslinked/branched structure of PLA [13]. Therefore, PLA produces free radicals on the tertiary C atoms, which become

stabilized in reactive extrusion [4]. In the case of Joncryl, the 907 and 843 cm−<sup>1</sup> peaks disappear, owing to the interaction of the epoxy groups with carboxyl groups on the PLA, suggesting that the reaction occurrs between the Joncryl epoxy and the PLA terminal functional group [32]. Surprisingly, there is a synergistic effect of peak decreases for Joncryl at 2922 cm−<sup>1</sup> . These results relate to the decrease observed in the MFR results and the peak decrease shown in Supplementary Figure S2. Upon addition of all the components, the peak at 2922 cm−<sup>1</sup> increases, suggesting that the initiator has created macroradicals after the chain extender has truly extended the PLA chain for the BA attachment. In all of the composites, vibrations at 1752, 1455, 1186, 1080, and 868 cm−<sup>1</sup> remain unchanged. Inata and Matsumura [26] reported that the epoxides might react with carboxyl and hydroxyl end groups of polyesters, and the electrophilic group with the carboxyl end groups. It can be concluded that the chain extension/crosslinked/branched structures in the polymer composites play an important role in improving the properties of the reactive composites in a controlled way.
