**Development and Characterization of Polyester and Acrylate-Based Composites with Hydroxyapatite and Halloysite Nanotubes for Medical Applications**

#### **Elena Torres 1, Ivan Dominguez-Candela 2, Sergio Castello-Palacios 3, Anna Vallés-Lluch <sup>3</sup> and Vicent Fombuena 2,\***


Received: 14 July 2020; Accepted: 28 July 2020; Published: 29 July 2020

**Abstract:** We aimed to study the distribution of hydroxyapatite (HA) and halloysite nanotubes (HNTs) as fillers and their influence on the hydrophobic character of conventional polymers used in the biomedical field. The hydrophobic polyester poly (ε-caprolactone) (PCL) was blended with its more hydrophilic counterpart poly (lactic acid) (PLA) and the hydrophilic acrylate poly (2-hydroxyethyl methacrylate) (PHEMA) was analogously compared to poly (ethyl methacrylate) (PEMA) and its copolymer. The addition of HA and HNTs clearly improve surface wettability in neat samples (PCL and PHEMA), but not that of the corresponding binary blends. Energy-dispersive X-ray spectroscopy mapping analyses show a homogenous distribution of HA with appropriate Ca/P ratios between 1.3 and 2, even on samples that were incubated for seven days in simulated body fluid, with the exception of PHEMA, which is excessively hydrophilic to promote the deposition of salts on its surface. HNTs promote large aggregates on more hydrophilic polymers. The degradation process of the biodegradable polyester PCL blended with PLA, and the addition of HA and HNTs, provide hydrophilic units and decrease the overall crystallinity of PCL. Consequently, after 12 weeks of incubation in phosphate buffered saline the mass loss increases up to 48% and mechanical properties decrease above 60% compared with the PCL/PLA blend.

**Keywords:** biomedical polymers; hydroxyapatite; halloysite; mechanical properties

#### **1. Introduction**

Tissue engineering has been exploring new methods to replace missing human tissues through biomaterials-based scaffolds, usually engineered to drive cell growth and provide shape to the creation of the new tissue. However, we should consider alternatives when selecting materials for bone fracture remodeling since conventional materials used for these applications include metallic prosthesis, bone grafts, or polymers. Currently, biopolymers are being intensively studied to replace both metal prostheses and autologous bone grafts because metal prostheses induce poor bone regeneration with formation of fragile porous bone [1] and, although autologous bone grafts induce the growth of strong bone, donor bone is needed, requiring additional chirurgical interventions, eventually causing infections [2]. Thus, polymers having easy processability to obtain desired geometries and special functionalities to accelerate bone growth will be the best option to treat bone fracture remodeling. Accordingly, biopolymers used as scaffolds for tissue engineering applications need to overcome two significant challenges. First, the biodegradation process should be controlled with non-toxic

degradation by-products eliminated through natural pathways. Secondly, the material should maintain its structural and mechanical properties to avoid malformation of the new regenerated bone while healing [3].

Among all the biopolymers used for tissue engineering, bioabsorbable aliphatic polyesters are the dominant scaffolding materials because of their biodegradability properties. Biodegradable aliphatic polyesters, containing the ester functional group in their main chain, undergo hydrolytic cleavage generating oligomers, which will be subsequently assimilated into the surrounding environment. Poly (ε-caprolactone) (PCL), poly (lactic acid) (PLA) and poly (hydroxybutyrate) (PHB), approved by the U.S. Food and Drug Administration, are the most studied polyesters due to their easy processability and their tunability regarding crystallinity, thermal transition and mechanical strength properties [4,5]. The degradation rate and mechanical properties of biopolymers will be affected by the hydrophobicity, crystallinity and acidity of the selected polymer.

The hydrophobic biopolymer PCL is extensively used in drug delivery devices showing excellent biological activity. Accordingly, studies focused on its use as a scaffold and internal fixation system, although its low mechanical properties cannot meet the structural requirements of the host tissue. Consequently, an appropriate addition of fillers or blends could provide an adequate mechanical stiffness to resist in vivo stresses, preventing new tissue deformation [6–8]. As examples, Lowry et al. [9] tested PCL composites as internal fixation devices, observing a higher strength when using a PCL/bone complex compared with bony humerus healed with a stainless-steel implant. These observations are in concordance with the studies developed in the last decade by Rudd and co-workers [10–14].

Introduction of specific bioceramics can also confer new functions, such as higher biological activity. Thereupon, hydroxyapatite (HA) is broadly use as an inexpensive filler in tissue engineering [15–17] because of its osteoconductive properties, low inflammatory response and low toxicity in humans [18,19], based on the mineral phase of the human bone being mainly composed (around 60 wt %) of HA [20]. Therefore, introduction of HA into a polymer induces the formation of an apatite layer with similar characteristics to those of the bone mineral phase [21], inasmuch as HA improves cell attachment [22,23], inducing the differentiation of mesenchymal cells into osteoblasts, which accelerates bone formation [8]. Different authors observed both mechanical and biological improvement of biopolymer matrices with the addition of HA [24,25].

The PCL low stiffness can also be attributed to using halloysite nanotubes (HNTs) [26,27] which are an inexpensive biocompatible clay extensively used in biomedicine for drug delivery due to their tubular shape. HNTs also support cell adhesion, ascribed to HNTs surface nano-roughness, which acts as an anchor frame [28], and the interaction between silanol groups present on the HNTs surface [29] with hydroxyl and amino groups present on proteins.

In a previous study [30], mechanical and thermal properties of PCL were studied by modifying the additive percentage of the bioactive fillers HA and HNTs. Accordingly, the additive threshold was stablished in 7.5 wt % of HNTs and 20 wt % of HA achieving a noticeable improvement in mechanical properties with the simultaneous addition of the two fillers. As a result, the flexural modulus improved up to 112.3% reaching values of 886.8 MPa (standard deviation = 42.1), and Young's modulus increased to 109.3% with its greatest value at 449.6 MPa (standard deviation = 17.12). Knowing that HA promotes the formation of a layer of new bone, and that HA and HNTs alter hydrophobicity behavior, in a second study, [23], biological properties such as cell viability, proliferation and morphology supplied by both fillers were studied and compared on different pairs of polymers with similar chemical nature but different hydrophobicity. Accordingly, the hydrophobic polyester PCL was modified when it blended with poly (lactic acid (PLA) and combined with HA nanoparticles and HNTs. However, the hydrophilic poly (2-hydroxyethyl methacrylate) (PHEMA) was copolymerized as monomer with ethyl methacrylate (EMA) and also combined with HA and HNTs. These polymers, although dissimilar to PCL and PLA in terms of chemical nature and biodegradability, were chosen for comparison purposes because they are used for hard tissue applications. Initially, the in vitro biological development of polymers with different hydrophobicity showed that cells preferably proliferate on moderately hydrophobic surfaces (PCL/PLA). However, over longer culture periods, cell proliferation increased on more hydrophilic materials (P(HEMA-co-EMA)). Inorganic nanoparticles (HA and HNTs) improve cell viability and proliferation compared to the raw materials. We assumed that reduced cell spreading on hydrophobic surfaces at long culture times might occur as a consequence of two effects: protein absorption competition and the steric hindrance effect (solvation).

We acknowledge that the contributions of the previous studies [23,30] need to be accomplished by monitoring the degradation rate of biodegradable polyesters modified with HA and HNT. Biomedical polymers, after implantation, undergo significant changes regarding mechanical properties influenced by their degradation process. Considering that 75% of the human body is composed of water, hydrolytic degradation of aliphatic polyesters is an interesting feature for tissue engineering materials. Bone remodeling implies time-limited applications, which requires the elimination or degradation of the biopolymer after use to restore the surrounding living medium.

For all the above-mentioned reasons, we studied the bioactivity of polyester (PCL, PCL/PLA and PLA) and acrylates (PHEMA, P(HEMA-co-EMA) sets and their HA- and HNT-based nanocomposites, as well as the degradability of the polyester-based set. To determine the correct distribution of the fillers, a study was conducted using SEM-EDS and an evaluation of their wettability by measuring the contact angle. Finally, to demonstrate if the loads introduced in the nanocomposites diffuse to the environment, we evaluated the mechanical properties of the nanocomposites using tensile and flexion tests.

#### **2. Experimental**

#### *2.1. Materials*

Poly (ε-caprolactone (PCL), with trade name CAPA 6500, was provided by Solvay Interox (Solvay Interox, Warrington, UK). CAPA 6500 is a high-molecular-weight thermoplastic linear polyester derived from its own lactone monomer. PLA Ingeo™ biopolymer 6201D is a thermoplastic available in pellet form with a glass transition temperature of 55–60 ◦C and a melting point of 155–170 ◦C. NatureWorks LLC (Nature Works LLC, Minnetonka, MN, USA). Hydroxyapatite (HA) with chemical formula (HCa5O13P3), halloysite nanotubes (HNTs) (Al2Si2O5 (OH)4 2H2O), ethyl methacrylate (EMA) with 99% purity, and hydroxyl-2-ethyl methacrylate (HEMA), with a minimum of purity of 96%, were supplied by Sigma-Aldrich (Madrid, Spain). Benzoin and ethylene glycol dimethacrylate 98% (EGDMA) were used as ultraviolet initiator and crosslinking agent during the preparation of HEMA/EMA compounds. Both were also supplied by Sigma Aldrich.

#### *2.2. Preparation of the Polymer-Based Hybrids*

The first set of materials based on PCL and PLA were received in pellet form and dried prior to their preparation in an air oven at 50 and 60 ◦C, respectively, to remove humidity. In parallel, HA and HNTs were dried separately in a vacuum oven for 48 h at 200 and 80 ◦C. The proportions detailed in Table 1 were weighed and pre-mixed in a zipper bag. By a twin screw co-rotating extruder with different temperature profiles, the mixtures were mechanically homogenized. Specifically, for PCL-based compounds, the temperature profile was 65/75/85/90 ◦C, and for PLA based-compounds, we used temperatures of 170/173/17/180 ◦C. After the extrusion process, the samples were cooled to room temperature and pelletized. Again, prior to the injection process, the different compounds were dried under the same conditions as mentioned above. The injection was carried out in a Meteor 270/75 injection molding machine (Mateu and Solé, Barcelona, Spain) using as temperature profiles of the extruder: 80/80/85/85/90 ◦C for PCL compounds and 170/173/175/180 ◦C for PLA compounds. Next, 13 mm diameter samples were punched out.


**Table 1.** Composition and coding of PCL, PCL/PLA, PHEMA, and P(HEMA-co-EMA) composites.

The second set of compounds, based on ethyl methacrylate (EMA) and hydroxyl-2-ethyl methacrylate (HEMA), was obtained by simultaneous polymerization of the monomers, as summarized in Table 1. Using a 1:1 monomer ratio between HEMA and EMA to obtain the copolymer, the mixtures were stirred with 1 wt % benzoin and 0.5 wt % EGDMA. The corresponding ratios of HA and HNTs were added and stirred for 15 min. Each mixture was injected into a glass template for polymerization in an ultraviolet oven for 24 h and subsequently a 24 h post-polymerization process in an oven at 90 ◦C was required. The samples were immersed in boiling ethanol and cut into 13 mm diameter samples.

#### *2.3. Contact Angle Measurements*

The water contact angles (WCAs) of the nanocomposites were measured on the surface of the dry samples in the sessile drop mode. An Easy Drop Standard goniometer model FM140 (110/220 V, 50/60 Hz) supplied by Krüss GmbH (Hamburg, Germany) was used for this purpose. To determine the water contact angle, we used the Drop Shape Analysis SW21 (DSA1) software. A minimum of five replicates of each sample were analyzed, yielding a standard deviation of less than 5%.

#### *2.4. Mechanical Properties*

Tensile properties of PCL/PLA blends loaded with HA and HNTs were obtained using a universal test machine (Ibertest ELIB 30, SAE Ibertest, Madrid, Spain) according to ISO 527. Assays were carried out with a 5 kN load cell and a crosshead speed of 10 mm·min<sup>−</sup>1. Moreover, to determine the Young's modulus more accurately, an axial extensometer IB/MFQ-R2 from Ibertest (Madrid, Spain) coupled to the universal test machine was used. The Young's modulus was calculated in each case from the stress–strain initial slope and averaged from five replicates.

#### *2.5. Hydroxyapatite Nucleation*

Hydroxyapatite nucleation was followed on three replicates per sample and time point. First, a simulated body fluid (SBF) solution with an ion concentration close to that of human blood plasma was prepared by the method proposed by Kokubo and coworkers [31,32]. To obtain the SBF, we prepared two solutions. Solution 1 consisted of 1.599 g of NaCl (Scharlau, 99% pure), 0.045 g of KCl (Scharlau 99% pure, Barcelona, Spain), 0.110 g of CaCl2·6H2O (Fluka 99% pure, Madrid, Spain) and 0.061 g of MgCl2·6H2O (Fluka) in deionized ultrapure water (Scharlau) up to 100 mL. Solution 2 was prepared by dissolving 0.032 g of Na2SO4·10H2O (Fluka), 0.071 g of NaHCO3 (Fluka) and 0.046 g of K2HPO4·3H2O (Aldrich, 99% pure) in water up to 100 mL. Both solutions were buffered at pH 7.4 by

adding the necessary amounts of aqueous 1 Mtris-hydroxymethyl aminomethane, (CH2OH)3CNH2 (Aldrich), and 1 M hydrochloric acid (HCl, Aldrich, 37% pure). Next, both solutions were mixed to obtain SBF with the following molar ion concentrations: 142 Na+, 5.0 K<sup>+</sup>, 1.5 Mg2<sup>+</sup>, 2.5 Ca2<sup>+</sup>, 148.8 Cl−, 4.2 HCO3 <sup>−</sup>, 1.0 HPO4 <sup>2</sup><sup>−</sup> and 0.5 SO4 <sup>2</sup><sup>−</sup> mM. Samples were immersed in individual vials containing 10 mL of SBF solution with hydrazine (NaH2) to prevent bacterial proliferation. The vials were placed in an incubator at 37 ◦C and 5% CO2. A set of samples were withdrawn after 7 and 14 days.

#### *2.6. Cell Seeding*

NIH 3T3 fibroblast cells were expanded in the presence of 4.5 g L−<sup>1</sup> glucose supplemented with 10% fetal bovine serum (Thermo Fisher, Gibco, Waltham, MS, USA) and 1% penicillin/streptomycin (P/S; Thermo Fisher, Gibco) in Dulbecco's modified Eagle medium (DMEM; Thermo Fisher, Gibco) at 37 ◦C in a 5% CO2 incubator until confluence. After reaching confluence (3 days), cells were withdrawn from the culture flask. To proceed, 5 mL of versene solution (0.48 mM) formulated in 0.2 g ethyldiaminotetraacetic acid (EDTA) per liter of phosphate buffered saline (PBS) supplied by ThermoFisher (Gibco), were added for 5 min at 37 ◦C, and then removed. After, to neutralize the versene solution, 10 mL of DMEM was added, and the suspensions were centrifugated at 1000 rpm for 5 min. Then, the cells were resuspended in 1 mL medium, counted, diluted and seeded on the samples at a density of 2 <sup>×</sup> 104 cells cm<sup>−</sup>2.

#### *2.7. Morphological Analysis*

A ZEISS FESEM ULTRATM 55 scanning electron microscopy (SEM) device was used to analyze the morphology of the HA coatings and the NIH 3T3 fibroblast cells and their layout on the surfaces. The morphology of the HA coatings was studied by SEM and energy-dispersive X-ray spectroscopy (EDS) images obtained to validate the formation of a hydroxyapatite layer and the Ca/P ratio. To this end, the samples were sputter-coated with carbon under vacuum through a BALL-TEC/SCD 005 sputter coater. The mapping spectra were taken at 15 kV of acceleration voltage and 5 mm working distance; a secondary electron detector was used. Silicon was used as optimization standard. The mappings were taken at a magnification of 5000×.

In the study of NIH 3T3 fibroblast cells, arrangements were analyzed after 1 and 14 days of incubation. After each period, the culture medium was removed to rinse the samples in phosphate buffer (PB; Affymetrix, Santa Clara, CF, USA) and samples fixed with 4% paraformaldehyde solution during 30 h at 37 ◦C. A vacuum system was used to remove the water and to avoid any deformations on cell morphology. For this purpose, samples were rinsed in PBS twice and carefully frozen in liquid nitrogen and transferred to a freezer-dryer for drying.

#### *2.8. Degradation of PCL and PCL*/*PLA Based Hybrids*

Degradation of PCL and PCL/PLA loaded with HA and HNTs was followed in vitro at 37 ◦C using PBS (0.01 M (NaCl 0.138 M; KCl 0.0027 M) with a pH 7.4, at 25 ◦C was supplied by Sigma Aldrich). Due to the stability of thermostable compounds based on PHEMA, this study was only carried out on compounds based on PCL. With the aim of accelerating the process, samples were previously immersed in a 2M NaOH solution for 24 h. Three replicates of each composition were immersed in individual tubes with 10 mL of Dulbecco's Phosphate Buffered Saline (DPBS, Sigma Aldrich) (pH 7.4) with screw caps and maintained at 37 ◦C in an incubator. Each sample was removed after 4, 8 and 12 weeks, rinsed thoroughly with deionized water, and dried in an oven at 35 ◦C for 12 h.

The weight loss and the mechanical integrity of the materials were evaluated. An electronic balance with a resolution of 0.1 mg was used to determine the mass loss as follows:

$$\% \text{ Mass loss} = \frac{Mi - Mf}{Mi} \times 100\tag{1}$$

where *Mi* is the initial mass and *Mf* is the final mass of the dry sample.

#### **3. Results and Discussion**

The water contact angle formed in the range of 40◦–70◦ on a polymeric surface is known to influence cell attachment, since chemical surface interactions are a key factor during the bio-adhesion process [33]. Polymers with contact angles in this range, with a different chemical nature, were thus selected for this analysis, and their hydrophilicity was slightly modified by blending or copolymerizing with others of the same family. Therefore, polymer chemical surfaces were modified by blending hydrophilic/hydrophobic polymers and/or filling the polymer matrix with HA or HNT to analyze their role in wettability.

From the results in Figure 1, we can discern that the most hydrophobic sample was PCL with a contact angle of 105◦. Blending PCL with the more hydrophilic PLA, the surface wettability improved 25.3%, with a value of 83.7◦. Conversely, PHEMA was at the hydrophilic end, with a contact angle of 59.7◦, and the copolymerization with the more hydrophobic EMA decreased surface wettability up to 73.7◦. The addition of HA and HNTs clearly improved surface wettability on neat samples (PCL and PHEMA), but this effect was scarcely observed in mixed samples (PCL/PLA and P(HEMA-co-EMA)), considering the standard deviation. This effect could be attributed to the intermediate wettability of these samples, with contact angles in the vicinities of 80◦, together with their heterogeneous composition at the nanoscale.

**Figure 1.** Variation of contact angle in compounds based on (**a**) PCL/PLA and (**b**) PHEMA/EMA.

After determining the wettability of the samples and the influence of the addition of HA and HNTs fillers, a SEM-EDS analysis was conducted to firstly determine if the loads were homogeneously dispersed and, secondly, to assess the formation of a hydroxyapatite layer resulting from the incubation in SBF at 37 ◦C. In addition, the EDS analysis (EDS spectra not shown) allowed the quantification of the Ca/P ratio and the comparison with that of the stoichiometric HA (Ca10(PO4)6(OH)2), Ca/P = 1.67 [34].

Figure 2 shows the images obtained after 7 and 14 days in SBF on the samples based on PCL/PLA. After seven days, the PCL-based hybrids did not efficiently induce apatite growth. Precipitation on PCL and PCL/PLA compounds without needle conformation corresponded to the salt dissolved in SBF medium, usually NaCl. The samples with HA filler provided nucleation sites, and the silanol groups (Si–OH) present in HNTs provide favorable locations for apatite nucleation. We speculated that the electrostatic interaction drives the formation of calcium silicate [35], since comparing pure polymers (especially PCL and PCL/PLA blends) with HA- and HNTs-modified materials, the nucleation efficiency increases with the filler. We already showed that both polar carboxyl groups and hydroxyl groups induce apatite nucleation [36].

**Figure 2.** SEM images (5000×) taken for hydroxyapatite nucleation analysis on samples based on PCL/PLA.

However, Figure 3 summarizes SEM images of samples based on PHEMA/EMA. The polymethacrylate-based hybrids induced efficient apatite growth, with the exception of pure PHEMA, on which only scattered precipitated salts were observed on the surface. On the contrary, the P(HEMA-co-EMA) surface showed plenty of precipitates forming large and clear cauliflowers with intricate needle-shaped crystals [36]. As observed, once apatite nucleates on a location, it grows radially outward [37], creating cauliflower or hemispherical structures combined to form a continuous layer. Due to the weak hydrophilicity of P(HEMA-co-EMA), the biological activity is higher than PHEMA, where the number of polar groups available for nucleation per unit volume on the surface is greater. Therefore, the P(HEMA-co-EMA) surface adsorbs Ca2<sup>+</sup> ions from the SBF solution more efficiently, thereby increasing the concentration of Ca2<sup>+</sup> ions on the surface, and also forming Ca–P nucleation sites [38]. The first layer of apatite molecules generates the cauliflower aggregates from the secondary nucleation, observed especially in P(HEMA-co-EMA). This process induces a spherical growth perpendicular to the surface structure, which leads to the formation of clusters or grape-like structures [21].

**Figure 3.** SEM images (5000×) taken for hydroxyapatite nucleation analysis on samples based on PHEMA/EMA.

Figure 4 reveals the assessment of the Ca/P ratio to verify the formation of hydroxyapatite layers. In most determinations, the Ca/P atomic ratio remained in acceptable values between 1.3 and 2, which pointed to the formation of calcium and phosphate deposits resembling physiological apatite structures [34]. Particularly, the highest Ca/P ratios were obtained in the PCL, PHEMA and PHEMA 20HA samples after an incubation period of seven days, all with values higher than 1.8. After 14 days of incubation, the Ca/P ratio tended to the physiological ratio of 1.67 in most samples. However,

PHEMA did not show calcium on the surface after 14 days of incubation. The hydrophilic character of this polymer probably hinders the deposition of salts on its surface and their evolution toward HA. This observation coincides with that of [36], where its closely-related poly (hydroxyethyl acrylate) (PHEA) did not induce an efficient apatite growth, whereas its copolymer with EMA did.

**Figure 4.** Ca/P ratio after 7 and 14 days of incubation in SBF obtained by EDS on nanocomposites based on: (**a**) PCL and PCL/PLA, (**b**) PHEMA and P(HEMA-co-EMA).

In the SEM morphological images (Figure 5), we see the in vitro biological development of polymers with different hydrophobicity. At initial stages (day 1), cells preferably proliferated and colonized moderately on the hydrophobic surface (PCL/PLA, with contact angle of 83.7◦, as summarized in Figure 1). Thus, cells appear round, where interactions occur primarily between them or with the extracellular matrix, thus resulting in a monolayer of cells with few bonding sites with the polymer surface. When longer incubation periods were analyzed (14 days), cell proliferation was favored in more hydrophilic polymers, as is the case of P(HEMA-co-EMA), with a contact angle of 73.7◦. In these samples, cells exhibited a flatter morphology, establishing contact with the polymer surface. With the addition of HA and HNTs inorganic fillers, in general terms, we observed an increase in proliferation compared with the raw materials. As different authors have concluded, this improvement can be attributed to the generation of new reactive sites with Ca2<sup>+</sup> and PO4 <sup>3</sup><sup>−</sup> groups present in HA that bind with negative carboxylate and positive amino groups in proteins, respectively [39–42]. However, due to the presence of silanol groups (Si–OH) located at the surface of HNTs, the formation of hydrogen bonds between HNTs and proteins is allowed [29]. The results showed a greater proliferation at initial stages on moderately hydrophobic polymers (PCL/PLA), whereas over longer culture periods, more hydrophilic polymers P(HEMA-co-EMA) seem to improve cell proliferation, in concordance with the results by Zhou et al. [42]. Using polyvinyl alcohol (PVA), a highly hydrophilic polymer, the authors concluded that highly polar OH groups on neat PVA films might account for the delayed attachment of bone cells. Thus, we can possibly assume that reduced cell spreading on hydrophobic surfaces at long culture times might occur as a consequence of the protein absorption competition and the steric hindrance effect (solvation).

Regarding the mechanical properties, PLA is a hard polymer with good mechanical properties, its brittleness being its main disadvantage. However, PCL is a very soft polymer with a slow degradation rate. After mixing PLA and PCL, the synergy results in retaining the advantages of each polymer. Compared to pure PLA, the mixture has greater flexibility, hydrophobicity and crystallinity, which translate into a slower degradation rate (several months to years) [43]. To study if the PCL crosslinking factor (crystallinity) limits the mobility of the chains and, therefore, the degradation rate, an analysis of the mass loss and mechanical properties after 4, 8 and 12-weeks of incubation in PBS was conducted. The PCL crystallinity tends to reduce the chain mobility and thus its hydrolysis by means of the hindered access of enzymes to the polymer matrix [23].

**Figure 5.** SEM images (400×) taken for cell proliferation at 1 and 14 days on samples based on PCL/PLA and PHEMA/EMA.

From the results shown in Figure 6, the PCL-based material did not show significant mass loss after degradation in PBS at 37 ◦C for 12 weeks. Although the mass loss of all PCL-based samples gradually increased, compared with samples containing inorganic fillers, the mass loss of pure PCL increased rapidly during the 12-week degradation process. The lower mass loss was observed in samples with HA because HA contains hydroxyl groups, which can neutralize the medium by reacting with degraded acid by-products, thereby reducing the effect of acid catalysis on PCL hydrolysis [44]. Conversely, PCL/PLA blends showed a maximum mass loss rate of the samples when using the two fillers, 41% after 12 weeks. Blending PCL with PLA provides hydrophilic units and reduces the overall crystallinity of PCL, thereby improving the accessibility of water molecules and ester bonds, and increasing the hydrolysis rate [45]. The addition of fillers provides hydrophobicity, and the nano-roughness of the surface promotes interaction with water molecules and thus causes hydrolytic cracking. Therefore, the evolution of mass loss is related to the mechanical properties, which were followed through the analysis of their Young's moduli, gathered in Figure 7.

**Figure 6.** Results of degradation: mass loss (%) after 4, 8 and 12 weeks for PCL-based materials.

**Figure 7.** Results of Young modulus following 4, 8 and 12 weeks of degradation for PCL/PLA-based materials.

The Young's modulus of the PCL samples was about 275 MPa. We observed that as the incubation time and the percentage of mass loss increased, the mechanical properties gradually decreased. The neat PCL samples showed a sharp decrease in Young's modulus, but those with HA retained their Young's modulus due to the filler reinforcement. However, the samples with HA and HNT showed a gradual decrease in mechanical properties over time, since the additional threshold of the nanoloads could be exceeded as the polymer degraded, and the agglomerates could act as weak points and failure initiation points. Blending PCL with PLA resulted in the Young's modulus increasing to 800 MPa. In PCL/PLA blends with the presence of inorganic fillers, the faster degradation rate, associated with more hydrophilic properties and more amorphous phases, led to a faster decline in mechanical properties. More precisely, the PCL/PLA 20HA 7.5 HNTs blend provided a Young's modulus 60% lower than that of the PCL/PLA blend.

#### **4. Conclusions**

Studying the influence of the HA and HNTs fillers on the wettability of the selected polyesters and acrylates, we measured water contact angles, which corroborated that PCL is the most hydrophobic sample, whereas PHEMA is hydrophilic. The addition of HA and HNTs clearly improved the surface wettability of neat samples, but not of the binary samples (PCL/PLA and P(HEMA-co-EMA)).

Secondly, distribution of the fillers into the polymer matrices was studied through EDS mapping. A homogeneous distribution of HA was observed on all polymers, with the exception of PHEMA. Nonetheless, the presence of HNTs yields large aggregates on more hydrophilic polymers, derived from the hydrophobic character of the nanotubes, resulting in a good dispersion in non-polar polymers and creating agglomerations in hydrophilic polymers due to the lower interfacial adhesion. Considering HA nucleation on polymer surfaces, hydrophobic polymers did not show an efficient induction of apatite growth. Conversely, large hydroxyapatite cauliflower-shaped crystals formed on moderately hydrophilic surfaces. We discovered that PHEMA is excessively hydrophilic, so promotes HA nucleation on its surface.

Finally, the evaluation of the degradation rate of the biodegradable PCL-based samples demonstrated that both the blending PCL with PLA and the addition of HA and HNTs provide hydrophilic units, as well as decrease the overall crystallinity of PCL, thereby improving the accessibility of water molecules to ester linkages and thus the hydrolytic cleavage. Consequently, the faster degradation rate and reduced mass lead to a faster drop of mechanical properties.

**Author Contributions:** Conceptualization, I.D.-C. and S.C.-P.; Data curation, E.T. and I.D.-C.; Formal analysis, E.T. and S.C.-P.; Investigation, E.T. and I.D.-C.; Methodology, S.C.-P. and A.V.-L.; Project administration, A.V.-L. and V.F.; Resources, A.V.-L.; Supervision, A.V.-L.; Visualization, V.F.; Writing—original draft, E.T. and V.F.; Writing—review & editing, V.F. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** I. Dominguez-Candela thanks the Universitat Politècnica de València for the financial support through an FPI-UPV grant (PAID-01-19).

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

#### **References**


© 2020 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 (http://creativecommons.org/licenses/by/4.0/).

#### *Article*

## **E**ff**ects of Magnesium Oxide (MgO) Shapes on In Vitro and In Vivo Degradation Behaviors of PLA**/**MgO Composites in Long Term**

#### **Yun Zhao 1,2,3, Hui Liang 1, Shiqiang Zhang 1, Shengwei Qu 1, Yue Jiang <sup>1</sup> and Minfang Chen 1,2,3,\***


Received: 31 March 2020; Accepted: 28 April 2020; Published: 8 May 2020

**Abstract:** Biodegradable devices for medical applications should be with an appropriate degradation rate for satisfying the various requirements of bone healing. In this study, composite materials of polylactic acid (PLA)/stearic acid-modified magnesium oxide (MgO) with a 1 wt% were prepared through blending extrusion, and the effects of the MgO shapes on the composites' properties in in vitro and in vivo degradation were investigated. The results showed that the long-term degradation behaviors of the composite samples depended significantly on the filler shape. The degradation of the composites is accelerated by the increase in the water uptake rate of the PLA matrix and the composite containing the MgO nanoparticles was influenced more severely by the enhanced hydrophilicity. Furthermore, the pH value of the phosphate buffer solution (PBS) was obviously regulated by the dissolution of MgO through the neutralization of the acidic product of the PLA degradation. In addition, the improvement of the in vivo degrading process of the composite illustrated that the PLA/MgO materials can effectively regulate the degradation of the PLA matrix as well as raise its bioactivity, indicating the composites for utilization as a biomedical material matching the different requirements for bone-related repair.

**Keywords:** biopolymers composites; MgO nanoparticles; MgO whiskers; PLA; in vitro degradation; in vivo degradation

#### **1. Introduction**

The requirement for orthopedic implants in clinical medicine has been paid more attention by researchers in order to develop new materials to help patients avoiding the pain of secondary surgery, improve the recovery rate and reduce costs. Recently, polylactic acid (PLA) has been given more scientific research in the biomedical field, especially for bone repair [1–4], since it has the advantages of good biocompatibility, processability, biodegradability and bioabsorbability. However, some disadvantages have limited its widening applications [5–8]. For example, the mechanical properties of PLA cannot satisfy the requirements of load-bearing devices for bone fixation [5], and its poor bioactivity would significantly hinder the penetration and absorption of cells on the material surface [8]. In particular, predominant problems in its degradation have occurred. The pH variation in the degradation can accelerate the initial degradation rate of the implant and also increase the risk of adverse tissue reactions, such as the inflammatory response [9], which are not beneficial to the viability of bone cells, biodegradation properties and the requirements of bone formation.

In order to solve these problems, researchers have improved the performance of PLA by using inorganic fillers with a good biological activity and bone-similar ingredients [10–12]. Magnesium oxide (MgO), another kind of inorganic filler, is with good biocompatibility and non-toxic biological activity, and become a research hotspot of polymer modification [13–15]. It can release the Mg2<sup>+</sup> ions when dissolving, which is favorable for a variety of enzymes conducive to the activation and synthesis of proteins. Moreover, its unique biological activity plays an important role in cell viability and as the antibacterial agent, to improve the survival rate of cells as well as the biological activity of composites [7,16]. Additionally, MgO also perform as an alkaline degradable material with excellent biological characteristics for the tissue engineering of regenerated bone tissue [16]. It is worth mentioning that besides the release of magnesium ions in the dissolving process, MgO, including the nanoparticles and whiskers, can impact on increasing the mechanical properties of the polymer matrix [13,17] and this is very different from the pure Mg particles as a filler in the PLA matrix [18]. It shows that since the Mg materials can produce the hydrogen during the composite degradation, it is easily apt to form the interface defects between the matrix and fillers, and meanwhile, there is no surface modification for the Mg particles prior to the preparation of the composite, which also weakens the interface bonding. In previous studies [19], it is found that the PLA/MgO composite film fabricated by the solution casting method can significantly improve the mechanical properties and biological activities of the PLA matrix, and its alkalinity has a neutralizing effect on regulating the pH of short-term degradation solutions.

It has been reported that the shape of fillers plays a crucial role in the mechanical properties and crystalline behaviors of composites. As to the materials of the PLA matrix, the whiskers usually perform the more enhancing capacity for the mechanical properties of PLA, compared with the nanoparticles [7,19]. Moreover, the difference in filler shape also affects the degrading behavior of the PLA composite during a short-time period because the dissolving rates of fillers are diverse, and this can cause the different variation in the interface bonding [18]. Luo et al. studied the degradation properties of MgO whisker/PLA composites in vitro, focusing on the long-term degradation, and since, their work has proved its better mechanical property. Nonetheless, it is necessary not only to focus on the initial property, but also to understand and investigate the long-term degradability and biocompatibility of such composites, as implants for bone repair, especially the filler shape effects on bone cells and tissues. It is worth noting that the change in the hydrophilicity and crystallization of the polymeric matrix has also been improved and influenced by the addition of fillers as well as its shape. The increasing hydrophilicity can cause the variation in the water intake in materials, and the change in the crystallization is beneficial to the growth of ordered crystal structures by affecting the degradation process of the matrix, such as the hydrolysis process of PLA, as the chain and segment are more easily apt to move in the degradation [20,21]. However, few studies have focused on the effects of the MgO shape on the degrading property of the PLA/MgO composite, and there is also no report on the in vivo degradation process of the PLA/MgO composite which provides the direct performance of its degrading property and biocompatibility to bone tissue.

In this work, MgO nanoparticles and whiskers are modified with stearic acid to obtain chemical binding through their interactions [22], as demonstrated in our previous work [17]. The PLA/MgO composites soaked in a phosphate buffer solution (PBS) under long-term degradation were studied, and particularly, the influence of the filler shape on the degradation behavior of the composites was analyzed. The in vivo degradation behavior of the composites was also evaluated.

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

#### *2.1. Materials*

PLA 2002D (NatureWorks, USA) was dried at 40 ◦C for 72 h before use and all other agents used were of analytical grade and did not require further treatment.

#### *2.2. Preparations and Modifications of pMgO and wMgO*

The fillers of the MgO nanoparticles (pMgO) and MgO whiskers (wMgO) were prepared in the laboratory [17,23]. Briefly, MgCl2·6H2O (76.24 g, CAS: 7791-18-6) was dissolved in 250 mL deionized water, and 1 mL cetyl trimethyl ammonium bromide (CTAB) (1.0 wt%, CAS: 57-09-0) was added dropwise. C2H2O4·2H2O (23.64 g, CAS: 6153-56-6) was added to this solution and the mixture was allowed to react for 20 min. The suspension was collected and centrifuged at 7000 rpm for 10 min. The pMgO precipitate was dried for 1 h in a vacuum oven (85 ◦C) and then sintered in a muffle furnace for 5 h at 600 ◦C.

The 100 mL solution of Na2CO3 (0.6 mol/L, CAS: 497-19-8) was added dropwise into an equal volume solution of MgCl2 (0.6 mol/L, CAS: 7791-18-6) and stirred for 20 min. The mixture was aged at room temperature for 10 h and then filtered, washed, and dried at 80 ◦C for 3–4 h. The precursor was calcined at 750 ◦C for 4 h with a heating rate of 5 ◦C/min. Then, the resultant wMgO was obtained.

Mixtures of 0.5 g of MgO (nanoparticles or whiskers) and 50 mL of ethanol were placed in three-necked flasks, which were subjected to ultrasonic treatment in a bath to achieve a full dispersion of the mixtures. The mixtures were then heated to 45 ◦C under reflux condensation. Stearic acid (0.005 g, CAS: 57-11-4) was dissolved into 20 mL of ethanol, added dropwise into each MgO suspension, and allowed to react for 1 h in reflux condensation conditions. The mixtures were then centrifuged, washed, and dried. The resultant products were denoted as SpMgO or SwMgO. The morphologies of pMgO, wMgO, SpMgO and SwMgO are shown in the Supplementary Materials.

#### *2.3. Preparation of PLA*/*pMgO and PLA*/*wMgO Composites*

The PLA/pMgO and PLA/wMgO composites were manufactured by a micro twin screw extruder (Wuhan Ruiming Test Equipment, Ltd., China) in advance, with a screw speed of 40 rpm and the temperature profile varied from 165 ◦C at the feeding zone to 190 ◦C at the die. The SpMgO/SwMgO was dry-mixed with PLA and the mixtures were dried under a vacuum at 50 ◦C for 48 h prior to the preparing process for the PLA and composites, then followed by the pelletizing step. Detailed information on the prepared materials is presented in Table 1.


**Table 1.** Detailed information of the samples.

The extruded pellets were dried at 50 ◦C for 24 h before extrusion again by the same machine conditions to obtain the composites' wire (ϕ2 mm, controlled by the extrusion machine head) and then prepared the specimens' rods (ϕ2 × 10 mm) directly. These extrusion-molded specimens were used for the in vitro degrading experiments.

#### *2.4. In Vitro Degradation in PBS*

In vitro degradation of the pure PLA and composite rods were carried out by immersing them in 4 mL of a phosphate buffer solution (PBS) at 37 ◦C for 3, 4, 5, 6, 8 and 12 months. The specimens of each material were removed from the PBS at the end of these periods. After rinsing with deionized water and removing the surface water using filter paper, the samples were weighed and then dried in a vacuum at 40 ◦C to a constant weight.

#### *2.5. Characterization*

The morphology of the experimental samples was characterized by field-emission scanning electron microscopy (FESEM, JOEL 6700F, Japan, operating at 10 kV).

At the pre-set time points, the soaking specimens were weighed after being cleaned with deionized water and surface-dried, then they were dried to a constant weight in a vacuum at 40 ◦C and weighed again. The pH value of the PBS was determined by averaging the results of the 3 independent measurements, obtained from the three identical samples for each type of composite. The weight losses were calculated using the following equation:

$$\mathcal{W}\_{\text{Loss}}(\%) = \frac{\mathbf{m}\_{\text{d}}}{\mathbf{m}\_{\text{0}}} \times 100\% \tag{1}$$

where WLoss is the average degrading rate, and m0 and md are the initial and final weights (after removing the surface water), respectively.

The water uptake measurements were calculated using the following equation [18]:

$$\mathcal{W}\_{\text{water}}(\text{\textquotedblleft}\_{\text{\textquotedblleft}}) = \frac{\mathbf{m}\_{\text{W}} - \mathbf{m}\_{\text{d}}}{\mathbf{m}\_{\text{d}}} \times 100\text{\textquotedblright} \tag{2}$$

where Wwater% is the wateruptake rate, and mw, mo and md are the weight of specimen after conditioning, the final weight (after removing surface water) and the initial weight, respectively.

Gel permeation chromatography (GPC) measurements were taken for the degradation samples in tetrahydrofuran (THF) (analytical purity) at a concentration of 1–2 mg/mL by a Waters 2414 system (Milford, MA) equipped with a Waters Differential Refractometer. THF was eluted at 1.0 mL/min through two Waters Styragel HT columns and a linear column. The internal and column temperatures were kept constant at 35 ◦C. Calibration curves were obtained based on the standard samples of the mono-dispersed polystyrene.

The thermal analysis of the samples was performed using a differential scanning calorimetry (DSC) instrument (Netzsch Co. Ltd., Freistaat, Germany). The samples, weighing approximately 5–8 mg each, were sealed in an aluminum pan, heated under a nitrogen flow from room temperature to 220 ◦C at a heating rate of 20 ◦C/min, isothermally conditioned at 220 ◦C for 2 min, cooled to 0 ◦C and then reheated to 220 ◦C at a heating rate of 10 ◦C/min. The crystallinity degree (Xc) of the samples was estimated using the following Equation:

$$\chi\_{\rm c}(\%) = \frac{\Delta \text{H}\_{\rm m} - \Delta \text{H}\_{\rm cc}}{\Delta \text{H}\_{\rm m}^{0}} \times 100\% \tag{3}$$

where ΔHm (J/g) is the value of fusion, ΔHcc is the cold crystallization enthalpy obtained during the DSC heating process, ΔHm<sup>o</sup> is the fusion enthalpy of the completely crystalline PLA, and φ is the weight fraction of PLA in the sample. The value of PLA is selected as ΔHm<sup>0</sup> = 93.6 J/g [24].

#### *2.6. In Vivo Experiment*

#### 2.6.1. Animal Models

The in vivo experiments were carried out as described by Reference [24]. Briefly, a total of 10 healthy adult (~1 year) Japanese white rabbits weighing 3 ± 0.2 kg were selected for animal testing, and they were divided in 2 groups, with 5 rabbits in one group, of which 2 rabbits were used as standby samples. One rabbit was implanted with 2 samples, where the left and right legs were implanted with one sample each (φ2 × 6 mm). Further, sub-cage feeding was performed for a week and no adverse reactions were found. Rabbits were anesthetized with an intramuscular injection of ketamine (0.2 mL/kg). After that, the hair on the side of the knee in a roughly 5-cm range was shaved. The iodine disinfectant was used to disinfect the knee parts of the rabbit. The anterior lateral patella of the knee was incised to about 4 cm, followed by cutting the skin, lateral support and a joint capsule. A hole with a depth of 1 cm was drilled in the femur and tibial cancellous bone of the knee. The PLA and WPLA rods were implanted into the hole separately; then, postoperative suture, iodophor disinfection and sub-cage feeding were performed. Rabbits were sacrificed with ear veins

injected with air. The bone with the implanted rod was removed from the euthanized rabbits after 3, 6, 12 and 18 months, and preserved by a 10% formalin solution. For all the animal experiments, the materials and surgical instruments were radiation-disinfected in the Tianjin Jinpeng far radiation Co., Ltd. after Co60 for 24 h, with a radiation dose of 25 KGy, and the experiments were implemented by Tianjin Hospital (approval No. 2015-11155).

#### 2.6.2. Routine Pathological Examinations

Hard histological biopsies were performed to evaluate the structure variation of the implants under long-term degradation behaviors and the tibial cancellous bone response after surgery. The surgical sites were fixed in a 10% formaldehyde solution, and then the samples were dehydrated in the order of the graded series of alcohols. Following the dehydration and decalcification, the specimens were embedded in paraffin, and the tissue sections were stained with hematoxylin and red staining.

The whole process of experiment is present in Figure 1.

**Figure 1.** Schematic illustration of the experiment.

#### **3. Results**

#### *3.1. Microstructures of In Vitro Degradation of PLA and Composites*

Figure 2 shows the surface morphologies and fracture morphologies (brittle fractures treated with liquid nitrogen) of the PBS-soaked samples at the different degradation periods. It is obvious that the surface degradation of all specimens is gradually aggravated with the extension of the immersion time. Specifically, some microcracks on the surface of all the samples in the 6 months appear, and the degrading holes are also shown by the 8 and 12 months in Figure 2a, which are produced by the PLA decomposition [25]. Meanwhile, the higher number of decomposing holes of the PPLA and WPLA composites are displayed and indicate that the decomposition of composite materials is more intensive than that of PLA under the long immersion period, probably due to the presence of MgO affecting the matrix's hydrophilicity. In Figure 2b, the fracture morphologies of the composites also form the greater number of bigger holes of WPLA and PPLA observed in comparison with the contemporary PLA sample and the intensive hydrolysis behavior of the matrix over the 12 months of composition is also displayed, suggesting their accelerating degradation under the long-term immersion. However, it can be observed that the morphologies of the samples in some areas (marked by red arrows) are greatly different between WPLA and the others, particularly in the graphs of 12 months. The former performs

brittle-similar fracture behavior, while the latter exhibits noticeable plastification areas, which are probably a result of the degradation product and water [20,21].

**Figure 2.** SEM microstructures of polylactic acid (PLA) and composites at different degradation times: (**a**) surface morphologies; (**b**) fracture morphologies.

#### *3.2. Weight Loss, Water Intake and pH Value*

The weight changes caused by the sample's degradation in the PBS are shown in Figure 3a. It can be seen that the variation in all samples is similar, namely by decreasing gradually. Obviously, the residual weights of the composites are comparatively less than that of the neat PLA over the 12 months, which is also mainly due to the MgO presence increasing the water uptake and accelerating the matrix decomposition. Initially, from the 4 months, there is an apparent increment in the degrading rate of PLA possibly because of its self-catalytic degradation [26], which accelerates its weight loss. Compared with the control PLA sample, the variation in the composites' weight loss is not greatly remarkable during the whole testing process. However, it is notable that WPLA provides the slightly more residual weight from the 5 months probably due to the hard degradation of its crystalline part leading to much more water, and there is also the increment in the water intake of WPLA over 5 months. Consequently, PPLA has the higher residual weight in comparison with WPLA in the 12 months. This is probably related to the MgO shape affecting the crystal behaviors and water uptake of the PLA matrix [18]. In Figure 3b, the composites exhibited the higher value of water absorption, which is consistent with their weight loss results, and attributed to MgO enhancing the hydrophilicity of the PLA matrix. Moreover, compared with the contemporary WPLA specimens,

the samples of PPLA perform much a stronger hydrophilic capability during the experimental period and this is probably caused by the higher quantity of nanoparticles compared with the same weight whiskers. Undoubtedly, the density of the crystal areas of PPLA can also be increased more notably by nanoparticle nucleation [23] but these forming areas are probably smaller due to the inhibiting growth effect between each other. Moreover, there may be much more parts of the imperfect crystal and amorphous region present, which can easily occur at the beginning of the degradation and lead to the acceleration of the degrading process, since metal oxides can effectively accelerate the decomposition of the PLA matrix, especially in a comparatively short-term degradation [27]. As to the variation in the pH value shown in Figure 3c, the pure PLA has a significantly lower value than that of the composites. With the addition of MgO, the pH values of the degrading solution are higher, in spite of decreasing gradually, especially at the final stage. The delay in the pH reduction in the composite solution may be mainly due to the neutralization of the alkaline magnesium ions released by MgO in the degradation medium [28]. Additionally, it is noted that PPLA also performs slightly better in controlling the solution's pH value, although the pH value of the later period was around at 6.4.

**Figure 3.** The (**a**) weight loss, (**b**) water intake and (**c**) pH value as a function of the degradation time of the neat PLA and composite immersed in the phosphate buffer solution (PBS).

#### *3.3. GPC and DSC Results*

In Table 2, the molecular weight data obtained from the GPC analysis are included. It is observed that the *Mw* of PLA and the composites decrease as the time is prolonged, as a consequence of the hydrolytic process. Initially, there is no intensively significant difference between the composites and PLA. The *Mw* of WPLA over 3 months is a little higher and this tendency is maintained until the 12-months of the degradation. As to the variation in PI (dispersity indexes), initially, the neat PLA performs the lower value compared with the composites for the samples of 3 and 4 months, while with

the time prolonging, there is an increasing tendency of PLA PI. This change is probably due to the hydrolysis of the matrix and the formation of the degradation product with a lower *Mw* from the amorphous areas in the PLA sample. It is noticeable that the variation in the *Mw* of PPLA is with the lowest value from 3 months to 6 months, and this is consistent with the previous study's report that metal oxides can enhance the degradation of the PLA matrix before 3 months [29]. However, with the extension of the immersing time, PPLA has the higher *Mw* value than that of PLA in 8 and 12 months, indicating the delaying degradation of PPLA. This is probably due to the more crystalline structure formed by the nanoparticles and the slowing degradation of the crystalline part occurring in the polymeric matrix. The MgO, especially the whiskers, is favorable for forming the more crystal regions, resulting in it effectively inhibiting and alleviating the degradation of the specimen [7], in spite of increasing the water uptake and hydrophilicity of the PLA matrix. In addition, there is a prominent difference between the *Mw* of PPLA and WPLA in the 12 months, and the higher *Mw* value of WPLA observed is possibly ascribed to the presence of bigger and more complete crystal regions, which lead to a higher crystallinity and is formed by the tighter arrangement of the PLA molecule induced by the whiskers [30]. This also indicates that the large amount of water entering the matrix is apt to accelerate the hydrolyzing of the amorphous material, but it slightly impacts on the crystal area even for the long-term degradation.


**Table 2.** Gel permeation chromatography (GPC) data corresponding to the samples.

Weight averaged molecular weight (*Mw*), number averaged molecular weight (*Mn*) and dispersity indexes (PI = *Mw*/*Mn*) of the specimens as determined by GPC.

The DSC data for the pure PLA and the composites in the experiment are displayed in Figure 4. During the degradation process, the Tg peaks observed for the samples is gradually shifted to the lower value, and the same variation is also observed for Tm. There are also two endothermic peaks for some curves and this is a common phenomenon for PLA degradation due to the recrystallization process of the defective crystals as decomposed products [7]. The whole degrading process starts with an amorphous degradation and then gradually combines with the destruction of the crystalline structure in a long-term degradation, which is also with the similar decline process of the molecular weight proved by the GPC results. It is found that there is only one endothermic peak in the PLA sample in 12 months with the lower value of Tm, and according to the GPC results in the corresponding period, it probably implies the formation of a decomposed product with a low molecular weight during the long-term degradation, which is distinguished from the crystalline region in PPLA and WPLA, as observed in the contemporary comparison of the curves. For the samples in 3 months, the curve of the PPLA specimen has double the endothermic peaks, possibly due to its accelerating degradation caused by the MgO nanoparticles [29] at the former 3-months stage. Moreover, in comparison with the curves of PPLA, the variation in the WPLA DSC line implies the lower degrading rate of WPLA, and this can be ascribed to its high crystallinity by the whiskers, demonstrated by the results of Xc in Table 3. Additionally, there is also no obvious cold crystallization phenomenon of WPLA in 12 months, differing from the PPLA performance, which also verifies the high crystallization of the WPLA composite. It is also greatly remarkable that the whole variation of PPLA in the DSC from 3 months to 12 months is not similar to that of WPLA, and the double endothermic peaks are maintained in all the curves of PPLA. It is probably due to the relatively denser or more compact crystalline network structure in the

PPLA matrix after the short-term degradation, which may be helpful to delay its degradation, in spite of its lower crystallinity.

**Figure 4.** The differential scanning calorimetry (DSC) curves of the secondary heating curves obtained for the degraded samples.


**Table 3.** DSC data corresponding to the samples.

#### *3.4. Dying of In Vivo Degradation*

According to the degradation results in vitro, the in vivo degradation of PLA and WPLA were carried out for the comparatively long-term implanting. More attention was given to the changes in the materials in vivo with the time of 3 months, 6 months, 12 months and 18 months, and the graphs of the histological examinations stained by hematoxylin and red staining are displayed in Figure 5. After 3 months, the edge of the implanted WPLA composite was not flat, while the control PLA was still relatively smooth, as shown. Combined with the results of the water intake and molecular weight of the materials, it indicates that the swelling stage of the decomposing process is more intensive for WPLA, due to its enhanced hydrophilicity. Meanwhile, after the 6-month implantation, it is noteworthy that the implants of PLA and WPLA exhibited a large number of decomposition cracks, and the apparent destruction of the implants was also observed (12 months of Figure 5), but the cracks for both samples are not similar. Specifically, there are much more and parallel cracks displayed in the control PLA in spite of the insignificant swelling, nevertheless the specimen of WPLA seems to swell significantly and break into some block regions by less random cracks. The results illustrate that the water intake rate

of WPLA with the MgO whiskers is faster than that of the pure PLA, but the degradation rate of the pristine PLA may be more prominent, suggesting that the degradation process of the matrix in the control and composites were different from each other. Moreover, after the 18 months, the implant of WPLA is basically decomposed, the partial degrading matrix enter into the newly-formed bone trabecula with a tight touch (marked by red arrows in 18 months of Figure 5) and PLA is still in the comparatively complete rod shape with a slight expansion, implying the accelerating decomposition of the swelled WPLA in the final degrading stage. This illustrates that MgO has a significant effect on the degradation process of the PLA matrix. The swelling rate and process are more accelerated and more intensively impacted on the degradation, along with the matrix decomposition. Additionally, from the parts noted by the red lines in the graph, it can be seen that the presence of MgO is favorable for inducing the tight connect between the matrix and bone trabeculae, due to both the better hydrophilicity and biological activity of the significant effect of Mg2<sup>+</sup> on bone formation and healing [28].

**Figure 5.** Histological morphologies of the implanted PLA and WPLA after 3-, 6-, 12- and 18-months duration of implantation.

#### **4. Discussions**

The nucleation abilities of the nanoparticles and whiskers for the polymeric composites are different, leading to the variation in their crystallinization. Undoubtedly, with the equal quality, the quantity of the nanoparticles is much more than that of the whiskers, and the hydrophilicity improvements of PPLA are more prominent, as shown in Figure 3b. The hydrolysis process of the composites is mainly impacted by water absorption. The SEM graphs reveal that the degradation behaviors of the composites PPLA and WPLA in the long-term degradation process are similar to that of the pure PLA, and there is the beginning of the amorphous region's decomposition for the polymeric matrix. However, the decomposition of the PLA amorphous area is obviously faster, which shows the "tough fracture", although no obvious hole is shown in the 12-months sample of Figure 1. It is probably due to the amorphous status from the degradation product of a low molecular weight [31]. Meanwhile, it also seems that compared with PPLA, the crystalline region of WPLA is more difficult to destroy, possibly due to two aspects. One is that the whiskers in WPLA induce the PLA molecular chain to form a crystal structure along the whiskers, which enhance their interface bonding and increase the crystallinity [30]. The other is that the whiskers have the ability to promote the ordering of the

hydrolyzed PLA molecular chains, and its looser percolation network structure is more conducive to the growth of the PLA chip [31], which is also confirmed by the results of a higher value of crystallinity in the DSC and molecular weight in the GPC of the 12-months degradation. Additionally, it should be noted that although PPLA is easier to lose the amorphous area and perform a fast degradation during a short-time immersion, the nanoparticles with a smaller size are apt to fabricate the relatively denser or more compact crystalline structure and this is helpful to delay the PPLA degradation in the long-term degradation, which can also maintain the pH stability, as shown in Figure 2.

As to the in vivo degradation, the variation in the decomposing process of the WPLA composite affected by the whiskers is also distinct from that of the control PLA. The long-term degradation results show that the composites swell more prominently and faster due to the enhanced hydrophilicity which increases the water intake, and the swelling rate of the composite is also much faster than the degradation rate of the polymeric matrix. Meanwhile, the degraded status of the WPLA implant observed directly from the histological morphologies differs from the sample without whiskers. Due to the swelling effect, the composite is decomposed into the "block" shape, although it still degraded through bulk erosion [32], which is identical to the degrading mechanism of the pristine PLA. However, with the higher water uptake, it indicates that the degradation of WPLA is greatly controlled by the diffusion rate of the water in the PLA matrix due to the significant enhancement of hydrophilicity. Moreover, the biological activity of the PLA matrix is obviously improved, the specimen is tightly surrounded by bone tissue and the materials and newly formed bone trabecula are integrated into each other.

#### **5. Conclusions**

The in vitro and in vivo degradation of the biodegradable composites prepared by the PLA matrix blending with the 1% wt. MgO were investigated to determine the effect of the MgO shape (nanoparticles and whiskers) on the degradation process of PLA/MgO composites. We have demonstrated that the addition of MgO can accelerate the water uptake rate of the PLA matrix, and the degrading procedure of composites begins with the loss of the non-crystalline region prior to the destruction of the crystalline parts, demonstrated by the DSC results, but the water uptake rate of the composites, especially with MgO whiskers, is obviously faster than its bulk degradation in the degrading procedure, caused by the presence of MgO. The in vivo results of the histological morphologies suggest that the PLA matrix of the composite has the prominently enhanced bioactivity and that this PLA/MgO biocomposite can be potentially utilized for bone repair with better performance.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4360/12/5/1074/s1. Figure S1. The SEM graphs of the MgO nanoparticles and whiskers.

**Author Contributions:** Conceptualization, M.C.; Data curation, Y.Z., H.L. and S.Q.; Formal analysis, Y.Z.; Funding acquisition, M.C.; Investigation, H.L.; Methodology, S.Z.; Project administration, M.C.; Resources, Y.Z., H.L., S.Z. and Y.J.; Supervision, M.C.; Validation, Y.Z.; Writing—original draft, Y.Z.; Writing—review & editing, Y.Z. All authors have read and agreed to the published version of the manuscript.

**Funding:** The work was funded by National Nature Science Foundation of China (No. 51801137, No. 51871166); the Tianjin Natural Science Foundation (No. 17JCQNJC03100); the Joint Foundation of National Natural Science Foundation of China (Grant No. U1764254).

**Conflicts of Interest:** The authors do not have any financial/personal conflict of interest that could affect their objectivity, or inappropriately influence their actions.

#### **References**


© 2020 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 (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Manufacturing and Characterization of Functionalized Aliphatic Polyester from Poly(lactic acid) with Halloysite Nanotubes**

**Sergi Montava-Jorda 1, Victor Chacon 2, Diego Lascano 2,3,\*, Lourdes Sanchez-Nacher <sup>2</sup> and Nestor Montanes <sup>2</sup>**


Received: 12 July 2019; Accepted: 3 August 2019; Published: 6 August 2019

**Abstract:** This work reports the potential of poly(lactic acid)—PLA composites with different halloysite nanotube (HNTs) loading (3, 6 and 9 wt%) for further uses in advanced applications as HNTs could be used as carriers for active compounds for medicine, packaging and other sectors. This work focuses on the effect of HNTs on mechanical, thermal, thermomechanical and degradation of PLA composites with HNTs. These composites can be manufactured by conventional extrusion-compounding followed by injection molding. The obtained results indicate a slight decrease in tensile and flexural strength as well as in elongation at break, both properties related to material cohesion. On the contrary, the stiffness increases with the HNTs content. The tensile strength and modulus change from 64.6 MPa/2.1 GPa (neat PLA) to 57.7/2.3 GPa MPa for the composite with 9 wt% HNTs. The elongation at break decreases from 6.1% (neat PLA) down to a half for composites with 9 wt% HNTs. Regarding flexural properties, the flexural strength and modulus change from 116.1 MPa and 3.6 GPa respectively for neat PLA to values of 107.6 MPa and 3.9 GPa for the composite with 9 wt% HNTs. HNTs do not affect the glass transition temperature with invariable values of about 64 ◦C, or the melt peak temperature, while they move the cold crystallization process towards lower values, from 112.4 ◦C for neat PLA down to 105.4 ◦C for the composite containing 9 wt% HNTs. The water uptake has been assessed to study the influence of HNTs on the water saturation. HNTs contribute to increased hydrophilicity with a change in the asymptotic water uptake from 0.95% (neat PLA) up to 1.67% (PLA with 9 wt % HNTs) and the effect of HNTs on disintegration in controlled compost soil has been carried out to see the influence of HNTs on this process, which is a slight delay on it. These PLA-HNT composites show good balanced properties and could represent an interesting solution to develop active materials.

**Keywords:** poly(lactic acid), halloysite nanotubes; mechanical characterization; morphology; thermal characterization

#### **1. Introduction**

In the last decade, the polymer industry has faced important challenges related to new regulations, increasing concern about environment, petroleum depletion and others. Sustainability has consolidated as a leading force in the development of new high environmentally friendly materials [1–3]. Petroleum-derived polymers have, in general, a remarkable effect on the overall carbon footprint, so that many researches have focused on the development of new polymers (thermoplastics, thermosetting

and elastomers) with a positive effect on their carbon footprint [4–7]. These environmentally friendly polymers could be classified into three different groups with different environmental connotations [8]. One group is composed of biodegradable petroleum-derived polymers such as most aliphatic polyesters, i.e., poly(ε-caprolactone)—PCL, poly(butylene succinate)—PBS, poly(glycolic acid)—PGA and so on [9–11]. Although they are petroleum-derived polymers, they show interest from an environmental standpoint as they can be disintegrated in controlled compost soil conditions. Another interesting and growing group is that of bio-based and non-biodegradable polymers which include polymers such as poly(ethylene)—PE, poly(ethylene terephthalate)—PET, poly(amides)—PAs and so on, which are obtained from renewable resources but are not biodegradable [12–16]. In fact, they show almost identical properties to their corresponding petroleum-derived counterparts. Their interests in that they can contribute to reduce the carbon footprint as they are obtained from plants that are able to fix CO2. Finally, there is an increasing interest on a group of polymers that are both bio-based and biodegradable. This group includes polysaccharides and derivatives, e.g., cellulose, chitin, chitosan, starch and derivatives, among others [17–19]. Protein-based polymers, e.g., gluten, soybean and collagen, among others are also included in this group [20–23]. Finally, bacterial polyesters or poly(hydroxyalkanoates)—PHAs, which derive from bacterial fermentation are gaining interest but up today, their synthesis process and purification is still an expensive process but these materials are thought to be a real alternative to a wide variety of polymers due the high number of poly(esters) that can be synthesized by different bacteria. Among all poly(hydroxyalkanoates), it is worthy to note the key role that poly(hydroxybutyrate)—PHB and its copolymers could acquire in the future [24–28].

Above all these polymers, it is worthy to note the current interest of poly(lactic acid)—PLA, which can be obtained by starch fermentation and, subsequently, polymerization. PLA and its blends have gained a privileged position in the polymer industry as it shows balanced properties (mechanical, thermal, barrier, physical and so on) with an everyday most competitive price [29–31]. Today it is possible to find PLA in a wide variety of industrial applications that include automotive, construction and building, packaging, houseware and so on [32–36], Behera et al. [37] proposed that by controlling the manufacturing process, the properties of PLA could be tailored e.g., by lowering Mw and the glass transition temperature (Tg). On the other hand, PLA is one of the most widely used material employed for fused deposition modeling (FDM) for 3D printed parts for whatever industry [38–41]. In addition to its biodegradability, PLA is resorbable and therefore it finds increasing use in the medicine industry (stents, screws, fixation plates, scaffolds for tissue engineering, surgical suture and so on) [42–46]. In these applications, PLA could also act as a drug carrier to provide controlled drug release [47,48]. Micro and nanoencapsulation have given interesting results for different purposes. This can be accomplished by loading the PLA matrix with encapsulated microor nanoparticles. In the recent years, nanotubes (NTs) have gained interest in both pharmacy and medicine applications as they can be used as carriers to deliver different active compounds such as antioxidants, antibiotics, antimicrobials and wound healing, among others [49–51]. Different nanotubes have been proposed with this aim and different possibilities. Despite carbon nanotubes (CNTs) have attracted most interest, other nanotubes are being studied as potential drug carriers in medicine and pharmacy [52–55], e.g., titanate and titania nanotubes (TiNTs) [56,57], hydroxyapatite (HApNTs) [58] and zinc oxide (ZnONTs) [59]. One important drawback of nanotubes is their synthesis process, which is, usually, complex and expensive. Nevertheless, aluminosilicates offer interesting properties as potential carriers. Halloysite nanotubes (HNTs) are naturally occurring nanotubes derived from a particular aluminosilicate structure, composed of an external silicate layer and an internal alumina layer with a different specific volume. This particular structure allows growing nanotubes in the form of rolled silica-alumina layers. They are worldwide available at a competitive price and they can be selectively etched to increase their carrying capacity and above all, they are completely biocompatible [60,61]. HNTs have been loaded into different polymer matrices such as poly(lactic acid), *N*,*N* -ethylenebis(stearamide) (EBS), poly(propylene), poly(amide) 11, poly(ethylene-co-vinyl acetate) (EVA), high impact poly(styrene) (HIPS) [62–64] and different loaded drugs such as insulin,

thymol and PEG-hirudin [65,66]. In addition to the bioactivity of these composites, they must fulfill some mechanical, thermal, chemical and physical properties to be used in a particular application [67].

This work explores the effect of halloysite nanotubes (HNTs) on mechanical, thermal and thermomechanical properties of a poly(lactic acid) matrix for potential uses in engineering applications. The effect of the loading amount comprised between 3 and 9 wt% is studied.

#### **2. Experimental**

#### *2.1. Materials*

A poly(lactic acid) commercial grade Ingeo 6201D in pellet form was supplied by NatureWorks LLC (Minnesota, USA). This grade possesses a density of 1.24 g cm−<sup>3</sup> and a melt flow index comprised between 15–30 g/10 min measured at a temperature of 210 ◦C. Despite this commercial grade is intended for melt spinning of fibers, this melt flow index is also suitable for injection molding and this grade has been previously used as a base material for plasticization, blending and manufacturing of wood plastic composites. Halloysite nanotubes (HNTs) with a chemical composition of Al2Si2O5(OH)4·2H2O and CAS 1332-58-7, were purchased from Sigma Aldrich (Madrid, Spain). HNTs had an average molecular weight of 294.19 g mol<sup>−</sup>1. The average length was comprised between 1 and 3 μm while the external diameter varied from 30 to 70 nm. The typical morphology of these HNTs is shown in Figure 1 where the typical tubular shape can be detected and even the lumen diameter can be seen by transmission electron microscopy technique, it was performed in a Philips microscope model CM10 (Eindhoven, the Netherlands), the acceleration voltage was set at 100 kV. The HNTs samples were spread in acetone then immersed in an ultrasound bath; afterward, a small drop of this solution was poured onto a carbon grid, it was subjected to solvent evaporation at 25 °C. From the TEM images, the lumen size of HNTs was obtained. At least 50 measurements were performed to obtain the size distribution. As reported in a previous work by Garcia-Garcia et al. [60], the composition of HNTs is mainly SiO2 (53.75%) and Al2O3 (44.57%) and, in a less extent, some additional oxides at less than 1% (P2O5, Fe2O3, SO3 and CaO, among others). As it can be seen in Figure 1, HNTs were also characterized by field emission electron microscopy (FESEM) operated at an acceleration voltage of 2 kV. Due to the high hydrophilicity of HNTs, they tended to form aggregates (Figure 1c), which shows an aggregate of dimensions 5.1 × 71 μm2. The tubular shape could also be observed by FESEM as it is shown in Figure 1d, in which a single HNT can be observed with an external diameter of 105 nm and a length of about 0.5 μm. Nevertheless, as indicated by Garcia-Garcia et al. [60] the size distribution of HNTs is rather heterogeneous.

Both PLA pellets and HNTs were dried at 60 ◦C for 24 h in an air-circulating oven to remove residual moisture. Different formulations were prepared containing different HNTs loading as summarized in Table 1. Poly(vinyl acetate)—PVAc was obtained from Sigma-Aldrich (Madrid, Spain), it was used as a compatibilizer due to its high hydrophilicity.


**Table 1.** Labeling and composition of poly(lactic acid) composites with different HNTs loadings with poly(vinyl acetate) compatibilizer.

\* phr represents the weight parts of additive per one hundred weight parts of the PLA/HNT composite material.

**Figure 1.** Different images showing the structure of halloysite nanotubes (HNTs). Transmission electron microscopy (TEM) images of halloysite nanotubes at different magnifications (**a**) 30,000×, (**b**) 80,000×. Field emission scanning electron microscopy (FESEM) images of (**c**) HNT aggregate at 10,000× and (**b**) isolated HNTs showing dimensions and the tubular structure at 100,000×.

#### *2.2. Manufacturing of PLA*/*HNTs Composites*

These formulations were placed in a zipper bag and subjected to an initial homogenization process. The mixtures were fed into a twin-screw co-rotating extruder from DUPRA S.L. (Alicante, Spain) with the following temperature profile: 165 ◦C (hopper), 170 ◦C, 175 ◦C and 180 ◦C (dye). The rotating speed of the screw was set to 40 rpm and the maximum extruder capacity was set at 60%. The obtained compounds, were pelletized for further processing by injection moulding in a Meteor 270/75 from Mateu & Solé (Barcelona, Spain) with the following temperature profile (from the hopper to the injection nozzle): 170 ◦C, 180 ◦C, 190 ◦C and 200 ◦C. After this, standard samples for characterization were obtained as can be seeing in Figure 2. As it can be seen, the addition of HNTs contribute to a remarkable change in color from white (neat PLA) to dark brown for the composite with 9 wt% HNTs. Therias et al. [68] have reported identical change in color with increasing HNTs loading.

**Figure 2.** Digital images of PLA-HNT composites with different HNTs loading.

#### *2.3. Thermal Characterization*

Thermal characterization was carried out by differential scanning calorimetry (DSC) in a DSC 821 from Mettler-Toledo Inc. (Schwerzenbach, Switzerland). The selected thermal cycled was divided into three different stages: A first heating from 30 ◦C to 200 ◦C was followed by a cooling down to 0 ◦C and, finally, a second heating stage from 0 ◦C up to 350 ◦C was scheduled. All the stages were run at a rate of 10 ◦C min−<sup>1</sup> in nitrogen atmosphere (66 mL min<sup>−</sup>1). Different parameters were obtained from the second heating cycle, namely, the glass transition temperature (*T*g), the cold crystallization peak temperature (*T*cc) and enthalpy (Δ*H*cc), the melt peak temperature (*T*m) and enthalpy (Δ*H*m) and the degradation temperature (*T*d). The degree of crystallinity (χc%) was calculated following Equation (1).

$$\chi\_{\rm cPLA} \left( \text{\textsuperscript{\text{\textsuperscript{\text{\textsuperscript{\text{\textsuperscript{\text{\textsuperscript{\text{\textsuperscript{\text{\textsuperscript{\text{\comfrown}}}}}}}}}} \text{\color{\text{\textsuperscript{\text{\textquotedblleft}}}}}{\text{\color{\text{\textquotedblleft}}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\color{\text{\textquotedblleft}}} \text{\textquotedblright}} \text{\color{\text{\textquotedblleft}}} \text{\textquotedblright}} \text{\textquotedblleft}} \text{\text{\textqu$$

where Δ*H*100% is a theoretical value that represents the estimated melt enthalpy of a fully crystalline PLA polymer, i.e., 93.7 J g−<sup>1</sup> as reported in literature [69]. Finally, the term wPLA represents the weight fraction of PLA on composites [70–72].

In addition to this thermal characterization, samples were analyzed by thermomechanical analysis (TMA) to obtain the dimensional stability of the obtained composites. In particular, the coefficient of linear thermal expansion (CLTE) was calculated from the slope of the characteristic TMA curves. To this, a thermomechanical analyzer TA Q400 (DE, USA) was used with a constant force of 20 mN. The heating program was set from –80 ◦C up to 100 ◦C at a heating rate of 2 ◦C min<sup>−</sup>1.

To better understand the thermomechanical behavior of the samples, they underwent dynamic mechanical thermal analysis (DMTA) in an oscillatory rheometer AR-G2 supplied by TA Instruments (New Castle, USA). A special clamp system designed for solid samples (40 <sup>×</sup> <sup>10</sup> <sup>×</sup> 4 mm3) were used, it works in torsion-shear conditions. The heating program was a temperature sweep from 30 ◦C up to 140 ◦C at a constant heating rate of 2 ◦C min<sup>−</sup>1. The maximum shear/torsion deformation (γ), and the oscillations frequency were defined as a percentage of 0.1% and 1 Hz, respectively.

#### *2.4. Mechanical Properties*

The most relevant information about mechanical properties was obtained by tensile, flexural, impact and hardness tests. Tensile and flexural tests were carried out as indicated in ISO 527 and ISO 178 respectively in a universal test machine ELIB 30 from S.A.E. Ibertest (Madrid, Spain). A load cell of 5 kN was used for both tests and the crosshead rate was set to 5 mm min<sup>−</sup>1. With regard to the impact properties, a Charpy pendulum with a total energy of 6 J from Metrotec S.A. (San Sebastián, Spain) was used on unnotched samples. The hardness was estimated using a Shore durometer with the "D" scale, model 673-D from J. Bot S.A. (Barcelona, Spain). All mechanical tests were run on five different samples and the average values of the corresponding properties were obtained.

#### *2.5. Morphology Characterization*

The morphology of the fractured samples (after failure in the impact test) was analyzed by field emission scanning electron microscopy (FESEM). A Zeiss Ultra55 FESEM microscope from Oxford Instruments (Oxfordshire, UK) working at an acceleration voltage of 2 kV was used. As PLA/HNTs composites are not electrically conducting, samples were subjected to a sputtering process with gold-palladium in a high vacuum sputter coater EM MED020 from Leica Microsystems (Milton Keynes, UK).

#### *2.6. Water Uptake of PLA*/*HNTs*

Water uptake of PLA/HNTs composites was carried out as indicated in ISO 62:2008 with distilled water at 30 <sup>±</sup> 1 ◦C for a period of 98 days. Rectangular samples with dimensions 80 <sup>×</sup> 10 <sup>×</sup> 4 mm3 were initially dried at 60 ◦C for 24 h to remove residual moisture in an air circulating oven 2001245 DIGIHEAT-TFT from J.P. Selecta, S.A. (Barcelona, Spain). After this drying process, samples were submerged in distilled water and removed after planned times. These samples are dried with a cotton cloth and subsequently, are weighed in an analytic balance AG245 from Mettler-Toledo Inc. (Schwerzenbach, Switzerland); after weighting, samples were submerged aging in distilled water. The amount of absorbed water during the water uptake was calculated from Equation (2).

$$
\Delta m\_t \left( \% \right) = \left( \frac{\mathcal{W}\_t - \mathcal{W}\_0}{\mathcal{W}\_0} \right) \times 100 \,, \tag{2}
$$

where *Wt* is the mass after a time *t* while *W*<sup>0</sup> represents the dry weight of PLA/HNTs composites before the water uptake process. After a particular immersion time (saturation time) changes in water absorption can be neglected as they are extremely low. Then, it is possible to obtain the saturation mass that is denoted as Δmass.

#### *2.7. Disintegration in Controlled Compost Soil*

Biodegradation test, or more correctly, disintegration in the controlled compost soil test, was carried out following ISO 20200. The disintegration process was carried out at a temperature of 58 ◦C and a relative humidity on soil of 55%. Squared samples (20 <sup>×</sup> 20 mm2) and a thickness of 1 mm were buried into a biodegradation reactor, prepared as indicated in the corresponding standard. After some periods, samples were unburied, washed with distilled water and dried at 50 ◦C and, finally, weighed in an analytic thermobalance. The weight loss during the biodegradation process was calculated from Equation (3).

$$\text{Weight loss } \left( \% \right) = \left( \frac{\mathcal{W}\_0 - \mathcal{W}\_t}{\mathcal{W}\_0} \right) \times 100. \tag{3}$$

In this equation, *Wt* represents the mass after a degradation time *t* and *W*<sup>0</sup> stands for the dry weight of PLA/HNTs composites before disintegration.

#### **3. Results and Discussion**

#### *3.1. Influence of HNTs Content on Mechanical Performance of PLA*/*HNTs Biocomposites*

Table 2 shows a summary of the main mechanical parameters obtained from different tests, i.e., tensile, flexural, impact and hardness, as a function of increasing HNTs content. The first thing one can realize is that the tensile strength (σt) was not highly affected by the presence of HNTs. It is true that we could observe a slight decrease in σ<sup>t</sup> with increasing HNTs content. In particular, neat PLA showed a σ<sup>t</sup> of 64.6 MPa while all composites with HNTs showed a σ<sup>t</sup> comprised between 58–59 MPa, which represents a maximum percentage decrease of about 10%. On the other hand, the effect of HNTs on ductile properties such as elongation at break (%εb) was much more pronounced. PLA is an intrinsically fragile polymer with a restricted elongation at break of only 6.1%. As we can see in Table 2, HNTs produced an embrittlement effect on PLA, leading to %ε<sup>b</sup> lower than 4%. The maximum decrease corresponded to the PLA composite with 9 wt % HNTs with an elongation at break of 3.3% (which represents a percentage decrease of about 45%). HNTs were dispersed into the PLA polymer matrix and promoted a loss on material cohesion. Therefore, the stress transfer from the embedded particles to the surrounding matrix was restricted, and this had a negative effect on elongation at break. Similar findings have been reported in literature with PLA matrices [73,74]. As expected, the tensile modulus increased with increasing HNTs content. It is important to remember that the tensile modulus (*E*t) represents the ratio between the applied stress (σ) and the obtained elongation (ε) in the linear region of a stress-strain curve. As above-mentioned, both tensile strength and elongation at break decrease with increasing HNTs content, but the decrease in %ε<sup>b</sup> was much more pronounced than that observed for σt. As the strain was in the denominator, as the denominator (ε) decreased more than the numerator (σ), the overall effect was an increase in *E*t. While neat PLA was characterized by an

Et value of 2086 MPa and the composite with 9 wt% HNTs was more brittle, with a tensile modulus of 2311 MPa (>10% increase). These results are in accordance to those reported by Chen et al. [45]. This embrittlement could be related to an increase in the structural stiffness of PLA chains due to HNTs-PLA interactions [75,76]. These low decrease in tensile properties is related to the use of the PVAc compatibilizer, which is highly hydrophilic that can establish interactions with both hydroxyl groups (–OH) contained in PLA (end chains) and HNTs surface, thus leading to good embedding of the HNTs into the PLA matrix with a positive effect on mechanical performance. Pracella et al. [77] reported the interesting compatibilizing effect of poly(vinyl acetate), PVAc to enhance compatibilization in PLA/cellulose nanocrystals, as it provides increased adhesion between these two components.

**Table 2.** Mechanical properties of PLA/HNTs composites obtained from tensile tests (tensile modulus—Et, tensile strength—σ<sup>t</sup> and elongation at break—%εb), flexural tests (flexural modulus—Ef and flexural strength—σf), hardness (Shore D) and impact (impact strength from the Charpy test).


With regard to flexural properties, the behavior of PLA-HNT composites followed the same tendency observed for tensile tests. All the herein developed composites with PLA and HNTs show increased flexural modulus (*E*f) compared to neat PLA. It is worthy to remark that PLA composites with 9 wt% HNTs offered a *E*<sup>f</sup> value of 3927 MPa, which was noticeably higher than neat PLA (3570 MPa). As the above-mentioned, addition of HNTs provided increased structural stiffness on PLA polymer chains [75,76,78]. On the contrary, the flexural strength (σf) decreased with increasing HNT content. At this point it is worthy to remark that deformation and strength were highly sensitive to material cohesion. Although some HNTs-PLA interactions are expected, presence of HNTs aggregates breaks cohesion and this has a negative effect on both properties in tensile or flexural conditions. Therefore, the flexural strength (sf) of neat was 116 MPa while this was reduced down to values of 107 MPa for PLA composites containing 9 wt % HNTs (around 7.7% decrease). Although it has been reported the reinforcing properties of HNTs on a PLA matrix, Therias et al. [68] have reported a slight decrease (almost negligible) of tensile strength with increasing HNTs content (up to 12 wt %), the most noticeable effect is an increase in stiffness. It is important to remark that the modulus represents the stress to strain ratio in the linear region of a tensile or flexural diagram. As the tensile strength remains with high values and elongation at break is remarkably reduced, then this ratio increases. Therefore, the tensile and flexural moduli increase with increasing HNTs content. In contrast, Chen et al. [79] have reported alternating values of tensile strength (increase and decrease) with different HNTs loading, but, in general, the stiffness is always increased with HNTs loading. It is worthy to remark the work by Guo et al. [80]. In their work they report the effect of pristine HNTs and alkali-modified HNTs. The show a decrease in tensile strength and elongation at break by increasing the loading of pristine HNTs but, in contrast, alkali-modified HNTs contribute to an increase in tensile strength, which is attributed to improved PLA/HNTs interactions due to the alkali treatment, which enhances more available hydroxyl groups in the outer surface.

Table 2 also summarizes Shore D hardness values for all composites, compared with neat PLA. All values were close to 82 and there was a very slight increase with the presence of HNTs. Nevertheless, this slight change was included in the deviation itself and, therefore, it was not possible to conclude a clear tendency.

Table 2 also includes the impact strength values obtained from the Charpy test. As expected, addition of HNTs provided an embrittlement effect as indicated by the modulus (in both tensile and flexural conditions). This embrittlement was also evident from the impact strength values. Neat PLA offered an impact strength of 1.46 kJ m−<sup>2</sup> and this was reduced down to values close to half (0.71 kJ m−2) in composites with 9 wt% HNTs. Impact strength was also related to material cohesion and, as indicated previously, presence of HNTs aggregates led to poor material cohesion and this had a negative effect on both resistance (maximum stress that the material can withstand) and deformation, both parameters playing a key role in impact strength. The embedded HNTs particles acted as starting points for crack initiation and growth thus leading to a clear embrittlement, which can be observed by morphology analysis.

The morphologies of the fractured samples after impact tests are gathered in Figure 3 at different magnifications. Neat PLA (Figure 3a,b) showed a typical brittle fracture surface, with low roughness and a smooth fracture surface. As the HNTs loading increased, it was possible to detect an increase roughness, which was attributed to the presence of HNTs. HNTs promoted an embrittlement against impact, which favored crack formation and growth. This is because HNTs are highly hydrophilic and do not establish strong interactions with PLA matrix. At higher magnification (5000×) it was possible to see a fine particle dispersion of HNTs embedded into the PLA polymer matrix (Figure 3d,f,h). It is possible to observe the presence of HNTs aggregates as well as some individual HNTs embedded in the PLA matrix. As the HNTs content increased the density of these aggregates/individual HNTs increased. Although it seems these particles were fully embedded into the PLA matrix, there was poor interaction between them and the surrounding matrix even with the presence of PVA compatibilizer. This phenomenon provided poor material cohesion and, therefore, stress concentration phenomena could take place. Some research works indicate that HNTs usually give good particle dispersion due to the tubular structure if HNTs, which is responsible for a weakening of nanotube–nanotube interactions. In addition, weak interactions can be obtained between the PLA matrix and the embedded HNTs particles by hydrogen bonds between the carbonyl groups in PLA and the hydroxyl groups in halloysite [81,82].

Magnified FESEM images taken at 10,000× (Figure 4) showed in a clearer way, the presence of these aggregates and individual HNTs.

As it can be seen in Figure 4a, good HNT dispersion is achieved for this relatively low HNT loading of 3 wt%. Individual HNTs can be detected as well as some aggregates (in de upper-right side). This aggregate formation is also detectable in PLA/HNTs composites containing 6 wt% HNTs (Figure 4b) located at the left side; nevertheless, the presence of individual well-dispersed HNTs can also be observed. Finally, Figure 4c shows the fracture surface corresponding to the PLA/HNT composite with 9 wt% HNTs. It is clearly detectable the presence of larger aggregates and some individual HNTs embedded in the PLA matrix. To check this dispersion Figure 5 gathers the FESEM image of a PLA/HNT composite (9 wt% HNT) and their corresponding energy-dispersive X-ray spectroscopy mapping images for oxygen carbon (C Kα1\_2), aluminum (Al Kα1; O Kα1) and silicon (Si Kα1). As the above-mentioned, HNTs are aluminosilicate structures and, therefore, the presence of Al2O3 (Al Kα1) and SiO2 (Si Kα1 and O Kα1) was evidenced by the EDX analysis. In addition, these elements mapping suggest the presence of HNTs aggregates as the size was of 2.5 <sup>×</sup> <sup>5</sup> <sup>μ</sup>m<sup>2</sup> (see upper-left side of O Kα1, Al Kα1 and Si Kα1 EDX mapping images), as well as some individual points related to individual HNTs as observed by the FESEM analysis.

**Figure 3.** Field emission scanning electron microscopy (FESEM) images at different magnifications (2500×: left column; 5000×: right column) of fractured samples from impact tests corresponding to PLA-HNT composites with different HNTs loading. (**a**,**b**) neat PLA, (**c**) and (**d**) 3 wt% HNTs, (**e**) and (**f**) 6 wt% HNTs and (**g**) and (**h**) 9 wt% HNTs.

**Figure 4.** Field emission scanning electron microscopy (FESEM) images at 10,000× of fractured samples from impact tests corresponding to PLA-HNT composites with different HNTs loading. (**a**) 3 wt% HNTs, (**b**) 6 wt% HNTs and (**c**) 9 wt% HNTs.

**Figure 5.** Field emission scanning electron microscopy (FESEM) image of a PLA/HNT composite with 9 wt% HNTs at 5000× and EDX mapping corresponding to carbon (C Kα1\_2), aluminum (Al Kα1), oxygen (O Kα1) and silicon (Si Kα1).

#### *3.2. Influence of HNTs Content on Thermal Behaviour of PLA*/*HNTs Biocomposites*

The effect of HNTs on thermal properties was assessed by differential scanning calorimetry (DSC). Figure 6 and Table 3 gathers the main parameters from DSC analysis for neat PLA and PLA-HNTs composites with increasing HNTs loading. With regard to the glass transition temperature (*T*g), the addition of HNTs did not provide any remarkable change in this parameter, which indicates poor HNTs-PLA interactions. In fact, the *T*<sup>g</sup> values remained almost constant with values around 64 ◦C. Regarding the cold crystallization process, it is possible to observe a slight decrease in the cold crystallization peak temperatures (*T*cc) as observed by Prashanta et al. [74]. This is directly related to the nucleant effect that HNTs can exert in PLA chains, thus allowing PLA chains to fold in a packed way to form crystallites. The melt peak temperature (*T*m) was not affected by the presence of whatever loading of HNTs, remaining at values of about 172–173 ◦C. Regarding the degradation onset temperatures (*T*d), no clear effect of HNTs could be detected as all values were comprised between 322 and 327 ◦C [83]. With the obtained values of the melt and cold crystallization enthalpies (Δ*H*<sup>m</sup>

and Δ*H*cc) respectively, the percentage degree of crystallinity (%χc) was calculated and it is shown in Table 3. As observed by Murariu et al. [84], addition of halloysite provides slightly lower crystallinity values. Table 3 shows two different crystallinity values, one is %χc\_s, which represents the degree of crystallinity of the injection molded sample without any removal of the thermal history (this takes into account the difference between the melt and cold crystallization enthalpies) and a second %χc\_max, which stands for the maximum crystallinity PLA can reach in these composites (this only considers the melt enthalpy). As it can be seen, this maximum crystallinity followed the same tendency, it means, a decrease with increasing HNT loading. This could be related to formation of less perfect crystals due to the presence of HNTs as reported by Chen et al. [79] with similar crystallinities for neat PLA of about 37% and lower values of 33–34% for a HNT loading of 10–15%. They also reported a clear decrease in the cold crystallization process as observed in this work, while very slight changes were observed for the glass transition temperature, Tg, in accordance to the results in this research.

**Figure 6.** Comparative plot of the differential scanning calorimetry (DSC) thermograms corresponding to PLA-HNT composites with different HNTs loading.

**Table 3.** Main thermal parameters of PLA-HNT composites with different HNTs loading obtained by differential scanning calorimetry (DSC) analysis.


\* χc\_s represents the crystallinity of the injection-molded materials without removing the thermal history. \*\* χc\_max stands for the maximum crystallinity that can be obtained in PLA/HNT composites.

In addition to DSC characterization, the dimensional stability was assessed by thermomechanical analysis (TMA) through determining the coefficient of linear thermal expansion (CLTE) below and above the glass transition temperature (*T*g). Obviously, the CLTE was much lower at temperatures below the *T*<sup>g</sup> compared to temperatures above *T*g. This is due to the glassy behavior of the material below the *T*g, as expected. On the contrary, above *T*g the material becomes more plastic and this allows higher expansion. Nevertheless, the effect of HNTs on overall CLTE can be neglected as all the CLTE values for neat PLA and all its composites with HNTs changed in a very narrow range comprised between 76.1 and 78.7 μm m−<sup>1</sup> ◦C<sup>−</sup>1. This identical tendency can be observed for the CLTE above the Tg with values comprised in the 130.5–133.7 μm m−<sup>1</sup> ◦C−<sup>1</sup> range. It has been reported the positive effect of HNTs on dimensional stability of [85] polyimide (PI) composites up 40–50 wt% loading of HNTs and polyamic acid (PAA). This high loading contributes to lowering the CLTE of polyimide (PI) films to be compatible with most metals, thus leading to polymer matrix composites with application as film capacitors. The obtained results in the present work suggest a slight improvement on the dimensional stability but it is not highly pronounced due to the relatively low HNTs loading compared to other works.

Dynamic-mechanical thermal analysis (DMTA) is very helpful to assess the effect of HNTs on both thermal and mechanical properties. Figure 7 shows the plot evolution of the storage modulus *(G* ; Figure 5a) and the dynamic damping factor (*tan* δ; Figure 7b). Regarding to neat PLA, its characteristic DMTA curve gave interesting information. Below 50 ◦C, *G* remained almost constant and in the temperature, range comprised between 55–70 ◦C, a threefold decrease could be detected. This dramatic decrease in the storage modulus, *G* was directly related to the temperature range in which the glass transition occurs. Another important process that could be observed by DMTA in neat PLA was the cold crystallization process. Crystallization is related to the formation of a packed-ordered structure, which involves an increase in the storage modulus, *G* . Therefore, as we can see, *G* increased in the temperature range of 80–100 ◦C. Regarding the effect of HNTs on DMTA behaviour, it is possible to see that all PLA-HNTs DMTA curves showed the same shape of that of neat PLA but with slight changes. On one hand, all *G* curves were shifted to higher values thus indicating an increase in the storage modulus, *G* which is directly related to the tensile and flexural modulus as described previously. Similar findings have been reported by Prashanta et al. [74]. Regarding the glass transition process, it seems that all curves were overlapped, which indicates slight or no changes in Tg as observed by DSC. On the other hand, regarding the cold crystallization process, DSC revealed a decrease in the peak temperature, and this is in total agreement with the DMTA curves of neat PLA and PLA-HNTs composites. It is possible to see in Figure 7a that the characteristic curves were moved to lower temperatures as the HNTs content increased. On the other hand, Figure 7b shows the evolution of the dynamic damping factor (*tan* δ) with the temperature. Despite that there are several methods to obtain the *T*g from the DMTA analysis, one of the most accepted methods considers *T*g as the peak maximum of the dynamic damping factor. The peak was located at 65 ◦C, which is in total agreement with the values obtained by DSC and there were no detectable changes in *T*g. It is possible to find a correlation with the maximum values of the dynamic damping factor with increasing HNTs loading since *tan* δ represents the ratio between the loss modulus (*G"*) and the storage modulus (*G* ). As indicated, *G* increased with HNTs loading and as it is in the denominator of the dynamic damping factor, it led to a decrease in *tan* δ as it can be observed in Figure 4b. Table 4 shows a summary of some DMTA parameters taken at different temperatures, so that one can see in a clear way how the storage modulus of the samples behaved at different temperatures.

**Table 4.** Summary of some dynamic mechanical thermal properties of PLA-HNTs composites with different HNTs loading obtained by DMTA.


**Figure 7.** Plot evolution of the dynamic mechanical thermal properties (DMTA) for PLA-HNTs composites with different HNTs loading (**a**) storage modulus, *G* and (**b**) dynamic damping factor, *tan* δ.

#### *3.3. Study of the Water Uptake of PLA*/*HNTs Biocomposites*

δ The evolution of the water uptake of PLA-HNTs composites is shown in Figure 8. As it can be seen, there was an increase in the absorbed water as a function of the immersion time and followed the typical Fick s Law. An initial stage with high slope related to water absorption (Δmass) could be seen. In a second stage this increase was less pronounced and finally, a stationary mass was obtained (equilibrium water uptake), denoted as (Δmass∞).

**Figure 8.** Evolution of the water uptake process in PLA-HNTs composites with different HNTs loading.

The lowest water uptake values were observed for neat PLA with a saturation water content of 0.95% reached after 91 days. Halloysite nanotubes are highly hydrophilic materials due to the aluminosilicate structure and, consequently, as the HNTs loading increases, the saturation water increases as observed by Russo et al. [73]. In particular, the saturation for the PLA composite with 3 wt% HNTs was 1.31% while this was still higher for the composite containing 9 wt% HNTs (1.67%). The hydrophilic nature of HNTs was responsible for this increase. HNTs offer a high surface area with a high number of hydroxyl groups that contribute to the water uptake rate and equilibrium water. Therefore, as the HNTs loading increases, the Δmass<sup>∞</sup> also increases.

#### *3.4. Disintegration in Controlled Compost Soil of PLA*/*HNTs Biocomposites*

The weight loss of PLA-HNTs composites during the biodegradation (disintegration in controlled compost soil) is shown in Figure 9. All materials showed an incubation time of about nine days in which very slight weight changes took place. Above this induction time, all materials started a slow weight loss of up to 30 days and above this, the disintegration rate increased. Disintegration of PLA and other aliphatic polyesters was directly related to hydrolytic degradation of the ester groups; for this reason, it is necessary to maintain a relative humidity of 55% at a moderate temperature of 58 ± 2 ◦C to speed up the process. Neat PLA is the one with the highest disintegration rate as reported in literature [86,87]. After 35 days the weight loss was close to 30% and a 90% weight loss was achieved after 49 days. Above 49 days, PLA was fully disintegrated, and it was not possible to recover any piece to weight it. The presence of HNTs delayed the disintegration process. At 48 days, neat PLA was fully disintegrated while all PLA-HNTs composites showed a weight loss of about 80%. The overall effect of HNTs on the disintegration was a delay in the rate and lower weight loss after the same period compared to neat PLA as halloysite did not degrade due to its inorganic nature.

**Figure 9.** Follow up of the disintegration process in controlled compost soil of PLA-HNTs composites with different HNTs loading.

Figure 10 gathers some optical images of the disintegration process of neat PLA and PLA-HNTs composites with varying HNTs loading. Above two weeks, it was possible to appreciate a clear change in surface in almost all materials, which is representative of embrittlement (crack formation and growth), as Aguero et al. [88] reported, during the first two and three weeks, neat PLA is in the induction period. After four weeks all materials show some disintegration level. Neat PLA shows a less fragmented structure at this time. As it has been detected previously with the weight loss, presence of HNTs delay the disintegration process. After six weeks, the consistency of all materials has been lost and there is not a clear effect of the HNTs content, Paul et al. [89] have investigated the effect of organo-nanomodifiers in PLA blends in which it can be seen that they tend to degrade faster than neat PLA, but it is expectable that the final weight loss is directly related to the HNT content as this inorganic component does not degrade in these conditions, unlike this, in other researches [68], halloysite nanotubes have an effect that promotes degradation in photooxidation media.

**Figure 10.** Optical images of the disintegration in controlled compost soil of PLA-HNT composites with different HNTs loading.

#### **4. Conclusions**

This work assessed the technical viability of PLA composites with different halloysite nanotubes content (HNTs) for further uses. In particular, this work focused on the effect of HNTs loading (in the 3–9 wt% range) on mechanical, thermal, thermomechanical and disintegration properties of PLA-HNT composites. In general, the mechanical properties were slightly lower than those of neat PLA but the decrease in tensile or flexural strength was less than 7%, which was a positive effect. On the contrary, the elongation at break was reduced to a half with the presence of HNTs. Nevertheless, it is worthy to note that PLA itself was extremely brittle. With regard to thermal properties, presence of HNTs led to lowering the cold crystallization process but the glass transition temperature (*T*g) did not change. Identical behavior was obtained with differential scanning calorimetry (DSC) and dynamic-mechanical thermal analysis (DMTA) thus showing the consistency of the obtained results. Regarding the water uptake, the presence of highly hydrophilic HNTs contributes to increased water uptake (which is a negative effect on most industrial applications, but could be a positive effect on medical applications). Finally, the disintegration at PLA with HNTs was slightly delayed but, in general, all composites were almost disintegrated in a reasonable time. This work opens new possibilities to these composites for further applications in medicine as the overall properties are maintained and HNTs could be used as carriers for controlled drug delivery.

**Author Contributions:** Conceptualization, N.M. and S.M.-J.; methodology, L.S.-N. and D.L; validation N.M., L.S.-N. and S.M.-J.; data analysis, V.C. and D.L.; original draft preparation, V.C. and D.L.; review and editing, S.M.-J. and L.S.-N.; supervision, L.S.-N.; project administration, N.M.

**Acknowledgments:** This research was supported by the Ministry of Science, Innovation, and Universities (MICIU) through the MAT2017-84909-C2-2-R program number. D. Lascano wants to thank UPV for the grant received though the PAID-01-18 program. Microscopy services at UPV are acknowledged for their help in collecting and analyzing FESEM images.

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

#### **References**


© 2019 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 (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Characterization on Polyester Fibrous Panels and Their Homogeneity Assessment**

**Tao Yang 1,\*, Ferina Saati 2, Jean-Philippe Groby 3, Xiaoman Xiong 4, Michal Petr ˚u 1, Rajesh Mishra 4, Jiˇrí Militký <sup>4</sup> and Ste**ff**en Marburg <sup>2</sup>**


Received: 1 September 2020; Accepted: 14 September 2020; Published: 15 September 2020

**Abstract:** Nowadays, fibrous polyester materials are becoming one of the most important alternatives for controlling reverberation time by absorbing unwanted sound energy in the automobile and construction fields. Thus, it is worthy and meaningful to characterize their acoustic behavior. To do so, non-acoustic parameters, such as tortuosity, viscous and thermal characteristic lengths and thermal permeability, must be determined. Representative panels of polyester fibrous material manufactured by perpendicular laying technology are thus tested via the Bayesian reconstruction procedure. The estimated porosity and airflow resistivity are found in good agreement with those tested via direct measurements. In addition, the homogeneity of polyester fibrous panels was characterized by investigating the mean relative differences of inferred non-acoustic parameters from the direct and reverse orientation measurements. Some parameters, such as tortuosity, porosity and airflow resistivity, exhibit very low relative differences. It is found that most of the panels can be assumed homogeneous along with the panel thickness, the slight inhomogeneity mostly affecting the thermal characteristic length.

**Keywords:** Bayesian reconstruction; homogeneity; porous materials; fibrous polyester materials

#### **1. Introduction**

It has been demonstrated that polyester fibrous material is a good alternative to conventional sound absorbing material [1]. In textile industry, polyester fiber assemblies are in the form of woven, knitted and nonwoven structures. The woven structure fibrous material is made by using two or more sets of yarn interlaced to each other. Woven structure is generally more durable in comparison with knitted and nonwoven structures. However, woven structure is not widely used in noise treatment at low frequency range due to its relatively small thickness. The sound absorption of polyester woven structure determined by impedance tube and reverberant field method has been reported [2]. The single layer woven structure with a small thickness (i.e., 2.16–2.41 mm) exhibits poor sound absorption at low frequency range. Knitted structures are made of interlocking loops by using one or more yarns. Spacer fabrics, a special type of knitted structure, attracted great attention for sound absorption because of their thick structure possibility and designable appearances [3,4].

Unlike woven and knitted structures, nonwoven structure fibrous material has more advantages as a sound absorbing material, such as high porosity, economical price, light weight, recyclability, good elasticity and a large thickness range. Thus, nonwoven structure is more widely used for different noise reduction requirements. Nonwoven structure is produced in three stages: web formation, web bonding and finishing. Drylaid (i.e., carded, airlaid), wet-laid, spunmelt are the main technologies for web formation. Web bonding can be achieved through mechanical, thermal and chemical methods. The finishing aims to improve the outward appearance and the quality of the fibrous structure. Finishing methods are various according to different required specific properties. Nonwovens have an important role in sound absorption within the automotive, construction and a variety of industrial uses [5]. Nonwoven structure and its combination with other materials (such as woven structure, polyurethane foam, polypropylene foam, etc.) are used in seating area, headliners, side panels, carpets, trunks, bonnet liners for interior vehicle noise control [6]. Nonwoven structure material coupled with a hard wall can significantly reduce noise transmission and reduce the reverberation by improving the sound absorption [7,8].

The polyester nonwoven structure used in noise reduction is normally in the form of panel. The acoustic properties of polyester nonwoven panels have been well studied [9–11]. Kino et al. investigated the effect of various cross-sectional shapes polyester nonwoven structure on the acoustic and non-acoustic properties [10]. They concluded that cross-sectional shape has a slight effect on sound absorption, while there is a significant effect on the airflow resistivity, the thermal and viscous characteristic lengths. The accuracy of several prediction models for polyester materials was investigated by Garai and Pompoli, they recommended a new model to predict the acoustic characteristics of polyester fibrous materials [12].

Investigation of the homogeneity of an acoustic absorber can facilitate optimization of acoustic properties in practical applications. For instance, an absorber which is inhomogeneous in thickness direction will usually exhibit different acoustic performance at direct and reverse orientations. However, very few research studied characterization of the homogeneity of porous materials. In one paper, researchers applied X-ray computerized tomography (CT) to characterize the homogeneity of prebaked anode by analyzing the distribution of coke, pitch and porosity throughout the anodes as well as the variations in binder matrix thickness [13].

High-loft nonwoven panel is known to have useful acoustic properties [14]. Although some studies related to the acoustic properties of this material can be found in the literature [14,15], there is a lack of research on recovered non-acoustic parameters via inverse method for this type of material. Moreover, there are only a few publications focusing on characterization of the homogeneity of nonwoven panels. This paper presents an investigation of estimating non-acoustic parameters of polyester high-loft nonwoven panel via Bayesian reconstruction procedure. In addition, the homogeneity of polyester panels has been numerically assessed by comparing the results from panels' direct and reverse orientations.

#### **2. Experiment**

#### *2.1. Materials*

The polyester fibrous panels (Technical University of Liberec, Liberec, Czech Republic) are composed of 45 wt.% staple polyester, 30 wt.% hollow polyester and 25 wt.% bicomponent polyester. Low-melting polyester fiber consists of the sheath part of bicomponent fiber which is used to thermally bond the fiber structure. The cross-sectional images of three types of polyester fibers are shown in Figure 1. A microscope was used to measure fiber diameter, and fifty fibers were measured for each type of fiber to obtain an accurate average value. The mean diameters of the staple, hollow and bicomponent fibers are 13.19, 24.45 and 17.94 μm, respectively.

**Figure 1.** Cross-sectional images of fibers. (**a**) Staple fiber; (**b**) hollow fiber; (**c**) bicomponent fiber [9].

Polyester fibrous panel samples were produced by perpendicular laying technology which consists of carded and thermal bonding procedures (see Figure 2) [16,17]. The carded web, consisting of a proportion of bicomponent polyester fibers in the blend, is fed onto the conveyor belt. The reciprocating forming comb and pressure bar are two main working elements used to create vertical folds. The forming comb strokes the lower part of the carded web and pushes the carded web to form a vertical fold. The reciprocating pressure bar moves the folded web along with the wire guide and the conveyor belt to the batt layer. With the movement of the pressure bar, the needles placed on the pressure bar penetrate the folded web and strengthens the folds, which improves the vertical orientation of the fibers in the nonwoven structure. The web is subsequently stabilized by melting the bonding fibers present in the fiber blend when it passes through the through-air bonding chamber. Thereafter, the nonwoven fabric is cooled. Panel thickness is controllable by setting the distance between the grid and conveyor as well as the dimension of the pressure bar. Panel density is adjusted via the velocity of the conveyor belt. By selecting proper fiber blend and adjusting the lapping device, various end products providing high absorption and insulation performance to meet a variety of applications could be achieved. An example of so manufactured polyester fibrous panel is illustrated in Figure 3.

**Figure 2.** Sketch of perpendicular laying technology [17].

**Figure 3.** Example of manufactured polyester fibrous panel. Images are taken from different directions: (**a**,**b**) top view, and (**c**,**d**) side view.

#### *2.2. Direct Characterization*

Some non-acoustic parameters, such as thickness, porosity and airflow resistivity, can be easily determined according to ASTM D1777-96, ASTM C830-00 and ISO 9053–1, respectively [18–20]. The micro-CT (micro computed tomography), SEM (scanning electron microscope) and length weighted methods can be applied to determine fiber mean diameter [21,22]. The tomographic reconstruction method used to measure tortuosity of porous materials has been reported in 2009 [23].

Fiber diameter, thickness, porosity and airflow resistivity were directly measured in this study. The characteristics of the polyester fibrous panel specimens are listed in Table 1. An Alambeta device (SENSORA, Liberec, Czech Republic) was used to determine the panel thicknesses. Sample porosities were determined according to ASTM C830-00 [18]. In this standard, the porosity was determined as <sup>φ</sup> = 1 <sup>−</sup> <sup>ρ</sup>/ρ*<sup>f</sup>* , where <sup>ρ</sup> is the fabric bulk density and <sup>ρ</sup>*<sup>f</sup>* is the fiber density which is 1141.82 kg/m3. The density of three types of fibers was measured by liquid pycnometry technique [24]. Closed pores have less or no effect on airflow resistivity and sound absorption compared to open pores [25]. As a consequence, the voids in hollow fibers were not included in the analysis.


The 100 mm diameter circular specimens were punched by an Elektronische Stanzmaschine Type 208 machine for the standard airflow resistivity test. The airflow resistivity of samples was measured on AFD300 AcoustiFlow device (Gesellschaft für Akustikforschung Dresden mbH, Dresden, Germany) according to ISO 9053:1991 [20]. Ten samples were measured for each polyester nonwoven panel. Normally, the decrease of porosity results in the increase of airflow resistivity for the samples made by the same fiber content and manufacturing technology. While the airflow resistivity exhibits a drop in Samples 7–9, the phenomenon can be attributed to the slightly different manufacturing technology for these three samples compared to other samples.

#### *2.3. Acoustic Characterization*

The acoustic properties of the materials can be directly evaluated via reverberant chamber measurements, steady-state measurements and impedance tube measurements. Moreover, acoustic methods can be used to characterize the morphologic characterizations of reticulated foams, fibrous materials, granular materials and sandy soils [26].

#### 2.3.1. Impedance Tube Measurement

A four-microphone impedance tube was applied to carry out the measurements to recover the reflection *<sup>R</sup>* and transmission *<sup>T</sup>* coefficients. Then, the dynamic density <sup>ρ</sup>*eq* and dynamic bulk modulus *K eq* can be easily recovered [27–29].

Figure 4 illustrates the 30 mm impedance tube used for this study. It consists of two 1/4 in. G.R.A.S. microphones flush mounted on both sides of the test sample and the tube ends with an anechoic termination. The distances between microphone positions and the sample front surface are 50, 30, 150 and 170 mm, respectively. The excitation signal was a logarithmic swept sine, over the frequency range of 800–5500 Hz. To achieve accurate results, the microphones were calibrated, and phase matched with each other [27].

**Figure 4.** Four-microphone impedance tube configurations.

2.3.2. The Johnson-Champoux-Allard-Lafarge Model

In the Johnson-Champoux-Allard-Lafarge (JCAL) model, the equivalent dynamic density is associated with the viscous losses and the equivalent dynamic bulk modulus with the thermal losses [27]. The dynamic density of porous media was first proposed by Johnson et al. in 1987 [30]. Champoux, Allard and Lafarge et al. then modified the dynamic bulk modulus of porous media [31,32].

If saturating fluid is air, the JCAL model assumes the porous media are rigid and motionless at the frequency over the phase decoupling frequency. In the model, the equivalent dynamic density is described as:

$$
\widetilde{\rho}\_{\text{eq}} = \frac{\rho\_0}{\phi} \widetilde{\alpha}(\omega) \tag{1}
$$

where α(ω) is the dynamic tortuosity, given by:

$$
\widetilde{a}(\omega) = a\_{\rm os} + \frac{jv}{\omega} \frac{\phi}{k\_0} \sqrt{1 - \frac{j\omega}{v} \left(\frac{2\alpha\_{\rm os} k\_0}{\phi \Lambda}\right)^2} \tag{2}
$$

where ρ<sup>0</sup> is the density of the saturating fluid, φ is the open porosity, α<sup>∞</sup> is the dynamic tortuosity, *j* is the complex number, *v* = η/ρ<sup>0</sup> is the kinematic viscosity, where η is the dynamic viscosity, ω = 2π*f* is the angular frequency, *k*<sup>0</sup> = η/σ is the static viscous permeability, where σ is the airflow resistivity, and Λ is the viscous characteristic length [30].

The dynamic bulk modulus is described as:

$$\widetilde{\mathcal{K}}\_{\text{eq}} = \frac{\mathcal{V}^{P\_0}}{\phi} \left( \mathcal{V} - \frac{\mathcal{V} - 1}{\overline{\alpha}'(\omega)} \right)^{-1} \tag{3}$$

where α (ω) is the thermal tortuosity, given by

$$\widetilde{\alpha}^{\prime}(\omega) = 1 + \frac{j\upsilon^{\prime}}{\omega} \frac{\phi}{k\_0^{\prime}} \sqrt{1 - \frac{j\omega}{\upsilon^{\prime}} \left(\frac{2k\_0^{\prime}}{\phi \Lambda \nu}\right)^2} \tag{4}$$

where *v* = *<sup>v</sup> Pr*, where *Pr* is the Prandtl number, *k* <sup>0</sup> is the static thermal permeability, Λ is the thermal characteristic length [31,32].

#### 2.3.3. The Inversion Procedure

Inverse characterization methods are becoming more popular since they are able to simultaneously reckon several parameters [27]. The comparison between inferred porosity, airflow resistivity and characteristic lengths of porous materials using a commercial inversion software (i.e., FOAM-X) and those obtained from ultrasonic measurements, Bies-Allard and Kino-Allard models was studied in

2012 [33]. The deterministic inverse methods normally fit the measurements and predictions through models, then find the parameter that best describes the material. The Bayesian approach was applied to carry out the inversion procedure in this paper. Bayesian approach can inversely determine some physical parameters of porous materials from the impedance tube measurement [34]. The unknown parameters are assumed as random variables distributed according to a probability distribution in the Bayesian approach. This distribution is based on the existing knowledge or experience about the parameter values. However, for many of the non-acoustic parameters, it is possible to specify lower and upper bounds. Therefore, the prior parameter distribution can be defined. Posterior to collecting data, the parameter distribution is given by Bayes' theorem. The posterior parameter distribution depends both on the prior distribution and on the measurements. Thus, this distribution contains all the available information about the parameters. By using the optimization method, good estimation of parameters can be obtained. One optimization technique called Markov chain Monte Carlo (MCMC) has been well applied to solve the problem of non-acoustic parameters estimation. Not only the probability density function (pdf) of each parameter, but also the joint probability density functions of the parameters can be retrieved according to the Bayesian approach in conjunction with MCMC [35].

#### *2.4. Homogeneity Assessment*

A homogeneous material has the same properties everywhere, i.e., uniform without irregularities. Some macroscopic pore structure parameters (e.g., permeability or porosity) can be used to verify the homogeneity of porous media [36]. However, homogeneity strongly depends on the selected sample size. At the initial increase of sample size, those parameters vary with the sample size and exhibit random fluctuations. As the sample size increases, the amplitude of the fluctuations gradually diminishes. The values of the pore structure parameters remain stable once a certain sample size is reached. Dullien [36] stated that when the structure parameters of a porous material maintain close values to the increasing sample size, the media is said to be macroscopically, or statistically, homogeneous.

As stated above, researchers characterized the homogeneity of prebaked anodes by analyzing the distribution of pores and different types of particles based on X-ray computerized tomography and image analysis [13]. However, the first step, i.e., computerized tomography on porous material, is a time-consuming work. For instance, one cubic sample with 3 mm × 4 mm × 10 mm cube lengths takes 6 to 8 h to accomplish the tomography. By contrast, the acoustic method is much more efficient. The inferred non-acoustic parameters (e.g., porosity, tortuosity, airflow resistivity) can represent the geometric structure in the polyester panels. Moreover, homogeneity in through-plane orientation is more important than in-plane orientation since sound waves mainly propagate from surface to inner structure. Thus, the homogeneity at thickness direction of polyester panels will be analyzed by comparing the inferred non-acoustic parameters.

#### **3. Results**

In order to figure out the homogeneity of polyester panels, the measurements from both direct and reverse orientations of samples have been carried out. The surface texture of front and back sides is shown in Figure 5.

**Figure 5.** The surface texture of polyester fibrous panel sample: (**a**,**c**) front side, and (**b**,**d**) back side.

In order to prepare specimens with proper radius, specimens are carefully cut by scissors. A transparent plastic tube having the same radius as the impedance tube was adapted to ensure the specimens fitting exactly into the tube's cross-section. The samples are cautiously mounted in the connecting tube in the middle of four microphones before testing. The surface of each specimen is at the same level to the edge of the connecting tube. When changing the measurement direction, samples are rotated to another direction to avoid displacement. The front surface orientated to the speaker was referred to as direct orientation. The recovered parameters are listed in Table 2. Since airflow resistivity was used more often compared to static viscous permeability and these two parameters can be easily converted through the formula *k*<sup>0</sup> = η/σ, only the airflow resistivity is presented in Table 2. In addition, the values of the recovered porosity (φ), tortuosity (α∞), viscous characteristic length (Λ), thermal characteristic length (Λ ) and static thermal permeability (*k* <sup>0</sup>) are presented. The standard deviations for each value are reported in brackets. The tortuosity of common fibrous absorbent ranges from 1 to 1.06 [37]. As porosity approaches the value of 1, tortuosity reduces to the minimal value of 1 [38]. Some existing empirical formula between porosity and tortuosity, α<sup>∞</sup> = 1/ φ and α<sup>∞</sup> = 1 + (1 − φ)/2φ, can also explain this correlation [39,40]. The reconstructed tortuosity for all of the samples is 1 with low standard deviation as shown in Table 2. Although the recovered porosities seemingly display big differences compared to the directly measured values, the maximum relative error of inferred porosity is less than 4%. It was considered that the results with an error less than 10% are accurate enough for inverse analysis, as the value of porosity for a porous material can vary due to several uncertainties during measurements. Thus, it can be concluded that the inferred porosity is reasonable.


**Table 2.** Inferred non-acoustic parameters of polyester panels.

Figure 6 presents the comparison of some recovered non-acoustic parameters between two orientations. It can be easily seen that, with an increase of density the viscous characteristic length decreases, while the airflow resistivity increases. However, no clear trend can be found between thermal characteristic length, thermal permeability and density. Moreover, the difference on viscous characteristic length, airflow resistivity and thermal permeability are relatively small. It is obviously found that the inferred thermal characteristic length yields a significant difference on the two sides, especially for Samples 4 and 6 with densities of 23.54 kg/m3 and 24.54 kg/m3. In addition, the standard deviation is extremely high. This phenomenon can be attributed to the following reasons: interface difference on the two sides, complication on determination of thermal characteristic length and small frequency range or large measurement uncertainty [27]. The front side of samples is more even and continuous, while the back side is rough and uneven (see Figure 5). In addition, slight inhomogeneity can significantly affect thermal characteristic length. Estimation of the thermal characteristic length based on impedance measurement is very sensitive to boundary conditions. If the sample radius does not properly fit the impedance tube's radius, the sample can be compressed, or air leakage existed between the sample and the tube. Consequently, erroneous values will occur because of the modified sample microstructure or the influence of air leakage on the inversion procedure.

**Figure 6.** Inferred non-acoustic parameters from direct and reverse orientations: (**a**) viscous characteristic length Λ, (**b**) thermal characteristic length Λ , (**c**) airflow resistivity σ, and (**d**) thermal permeability *k* 0.

Since the values of inferred thermal characteristic length on the direct orientation are not reliable, the inferred airflow resistivity on the reverse direction was chosen to compare with the measured value in Figure 7. The correlation between the directly measured and recovered values is presented. It can be seen that the regression line has slope value close to 1 and the coefficient of determination is over 0.94. It can be further demonstrated that the recovering method can accurately estimate airflow resistivity of polyester fibrous panel. The relative difference was defined as <sup>δ</sup> = σ*meas* <sup>−</sup> <sup>σ</sup>*in f er* /σ*meas*, where σ*meas* is the measured airflow resistivity and σ*in f er* is the inferred value. The relative difference exhibits the biggest value for Sample 7 with a value of 0.489. This phenomenon can be explained by the small change that exists in manufacturing technology, as was stated above. The most accurate inversion of airflow resistivity occurs on Sample 12 with the smallest δ which is 0.001. As seen in Table 1, from Sample 1 to Sample 16 the density of the panels increases from 16.93 to 45.56 kg/m3. It can be found that δ is relatively lower when the samples are denser. For instance, Samples 9–16 have lower relative difference (e.g., <0.2). Meanwhile, among the low density panels (Samples 1–8) only two samples reach this relative difference level. It can be concluded that the inversion method of airflow resistivity is more accurate for denser polyester panel materials.

**Figure 7.** Comparison between measured and inferred airflow resistivity: (**a**) correlation between the measured and inferred values, and (**b**) their relative difference.

To compare the inferred non-acoustic parameters from direct and reverse orientations, the mean relative difference was calculated according to the following equation:

$$\delta = \frac{1}{N} \sum\_{n=1}^{N} \delta\_n = \frac{1}{N} \sum\_{n=1}^{N} \frac{\left| \mathbf{x}\_{f,n} - \mathbf{x}\_{b,n} \right|}{\mathbf{x}\_{f,n}} \tag{5}$$

where δ is the mean relative difference, *xf* and *xb* are, respectively the inferred non-acoustic parameters from direct and reverse orientations, and *N* is the total number of studied material specimens (*N* = 16).

The mean relative differences of inferred porosity (φ), tortuosity (α∞), viscous characteristic length (Λ), thermal characteristic length (Λ ), airflow resistivity (σ) and thermal permeability (*k* <sup>0</sup>) are presented in Figure 8. The values of mean relative differences are demonstrated on the top of each bar. It was assumed that the material is homogeneous when the physical parameters having variances less than 0.05 [41]. The mean relative differences of tortuosity, porosity and airflow resistivity are much smaller than the critical value. The differences of viscous characteristic length and thermal permeability are 0.109 and 0.121, respectively. However, the inferred thermal characteristic length exhibits the highest mean relative difference with the value of 0.397. Due to the sensitivity of estimating thermal characteristic length, its inferred values are not recommended for assessing materials homogeneity. Furthermore, tortuosity, porosity and airflow resistivity can well represent the pore size, fiber size and their distributions. Thus, the polyester fibrous sample panel is nearly homogeneous or slightly inhomogeneous at thickness direction. Moreover, the acoustic method can be an alternative approach to characterize the homogeneity of porous material. It can be used not only in the thickness direction, but also in arbitrary directions. Nevertheless, the size of the sample and the suitability of the JCAL model for the porous material of interest should be considered.

**Figure 8.** Mean relative differences of inferred non-acoustic parameters.

#### **4. Conclusions**

This work applied a Bayesian approach on polyester fibrous materials to inversely estimate some non-acoustic parameters. Meanwhile, the homogeneity of polyester nonwoven panels was assessed by comparing the inferred non-acoustic parameters from direct and reverse orientations. The Johnson-Champoux-Allard-Lafarge model and Markov chain Monte Carlo optimization technique were chosen to implement the Bayesian approach. Polyester samples with density ranging from 16.93–45.56 kg/m<sup>3</sup> were selected in this study. Measurements of reflection coefficient and transmission coefficient were carried out in a four-microphone impedance tube. The mean relative differences of inferred parameters between two orientations were used to analyze the homogeneity. The results indicated that the estimated porosity and tortuosity are reasonable. Moreover, other assessed parameters have generally lower contrast and standard deviation by looking at the qualities from two directions, while a sizable difference of thermal characteristic length was found. In addition, the increase in density decreases estimates of viscous characteristic length and increases estimates of airflow resistivity. While density does not show clear connections with thermal characteristic length and thermal permeability. Measured airflow resistivity and inferred values are very close. The inverse method can accurately estimate the non-acoustic parameters of denser polyester fibrous panel material (i.e., density > 28 kg/m3). Slight inhomogeneity could significantly affect the determination of thermal characteristic length. The mean relative differences of inferred tortuosity, porosity and airflow resistivity are, respectively 0, 0.004 and 0.019 which means the polyester material is nearly homogeneous. The applied acoustic method is an efficient way to characterize homogeneity of porous materials by comparing with computerized tomography technology. Hence, the proposed Bayesian inversion based on impedance tube measurement is valuable for the study on characterization of the homogeneity of porous material.

**Author Contributions:** Conceptualization, Conceptualization, T.Y. and J.-P.G.; Methodology, T.Y. and J.-P.G.; Software, J.-P.G. and T.Y.; Validation, T.Y., F.S., J.-P.G. and S.M.; Formal analysis, T.Y., J.-P.G. and F.S.; Investigation, T.Y. and J.-P.G.; Resources, T.Y., M.P. and J.M.; Data curation, T.Y., J.-P.G. and X.X.; Writing—original draft preparation, T.Y.; Writing—review and editing, T.Y. and X.X.; Visualization, T.Y. and J.M.; Supervision, J.-P.G., S.M., M.P., R.M. and J.M.; Project administration, M.P.; Funding acquisition, M.P. and T.Y. All authors have read and agreed to the published version of the manuscript.

**Funding:** This article is based upon work from COST Action DENORMS CA15125, supported by COST (European Cooperation in Science and Technology) and the European Union (European Structural and Investment Funds-Operational Program Research, Development and Education) in the frames of the project "Modular platform for autonomous chassis of specialized electric vehicles for freight and equipment transportation", Reg. No. CZ.02.1.01/0.0/0.0/16\_025/0007293.

**Acknowledgments:** The authors would like to gratefully acknowledge the support from the Laboratoire d'Acoustique de L'Université du Mans (LAUM, UMR CNRS 6613). The authors would like to thank Kirill V Horoshenkov, Alistair I. Hurrell and Mohan Jiao for their help throughout part of the airflow resistivity measurements.

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

#### **References**


© 2020 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 (http://creativecommons.org/licenses/by/4.0/).

*Article*
