Controlling the Ductile/Fragile Behavior of a 3D-Printed PLA-BaTiO₃ Biocomposite by PBS Addition

Abstract

The demand for patient-specific medicine is steadily increasing, particularly with the need for innovative materials capable not only of supporting tissue regeneration but also accelerating it. The aim of this study was to develop a new printable composite material exhibiting ductile behavior, in contrast to brittle failure, in order to support cell growth even under structural compromise. PBS was selected as a blending component with PLA due to its enhanced biocompatibility for bone tissue regeneration. Both neat PLA and the PLA/PBS blend were subsequently combined with BaTiO₃, processed into filaments, 3D printed, and subjected to mechanical testing. PLA-based composites demonstrated higher stiffness under compression, with up to a 6.5% increase in Young’s modulus compared to the blended samples. However, the incorporation of PBS resulted in a more ductile response, as evidenced by three-point bending tests, even at BaTiO₃ concentrations of 10 wt%. This improved ductility is expected to provide safer conditions for cell growth and enable elastic recovery following mechanical loading.

1. Introduction

For decades, bone healing strategies have relied primarily on alloy-based prosthetics to replace hard tissues [1]. Numerous efforts have been devoted to overcoming the inherent limitations of rigid prosthetic implants, such as fretting [2] and stress shielding [3], often by employing structural design strategies that preserve a stiff core [4]. However, recent advances in tissue engineering have shifted the focus from replacement toward the regeneration of tissues.

Despite the remarkable resilience of bone, external support becomes essential when the fracture gap exceeds one millimeter [5]. Current bone tissue engineering approaches include the use of bulk phosphocalcic ceramics or injectable composites that mimic the natural bone environment to promote regeneration [6,7]. In this process, bone-forming cells deposit extracellular matrix on the scaffold surface, and healing should be as rapid and complete as possible, ultimately rendering the regenerated tissue indistinguishable from the original. Although these approaches are superior to simple replacement, they still face significant limitations.

While bulk ceramics provide high compressive strength and mechanical properties comparable to natural bone, they offer a limited surface area for interaction with biological fluids, which can prolong healing. Combined with the growing demand for patient-specific treatments, this limitation has driven research toward additive manufacturing techniques, which enable the fabrication of customized, porous scaffolds tailored to individual anatomical needs.

The most widely used artificial bioresorbable polymer is poly (lactide) (PLA), already employed in surgical devices such as screws [8]. To enhance its bioactivity, PLA has been combined with hydroxyapatite (HAp) to better mimic the natural bone environment. For instance, Danoux et al. reported a more than 200% increase in alkaline phosphatase (ALP) activity after 14 days of culture when comparing neat PLA to a 50 wt% PLA/HAp composite [9]. More recently, composites incorporating barium titanate (BaTiO₃) have been developed to reproduce one of bone’s key physical properties: piezoelectricity. As summarized by Khare et al., many studies have demonstrated the positive impact of electrical stimulation on the rate of bone regeneration [10]. Rather than relying on external devices, researchers have sought to embed electrical activity within scaffolds using piezoelectric materials. For example, Yang et al. showed that the incorporation of BaTiO₃ and graphene into a scaffold increased the short-circuit current from nearly 0 to 5.1 nA for PLA/BaTiO₃, and up to 10 nA with the addition of graphene. This enhancement coincided with a 40% increase in ALP activity after 7 days [11]. Importantly, this effect was observed only under mechanical stimulation, highlighting the potential of piezoelectricity in vivo. However, as Guillot-Ferriols et al. noted, it remains challenging to precisely determine which aspects of electrical stimulation most influence osteogenesis, as studies use different stimulation sources [12].

Beyond bone repair, piezoelectric materials are being explored for a variety of biomedical applications, including nerve repair [13] and cardiac tissue regeneration [14]. To leverage multiple benefits, researchers have combined ceramics such as barium titanate with hydroxyapatite [15,16] or calcium silicate [17]. A comparative study by Huang et al. found that hydroxyapatite outperformed β-tricalcium phosphate in biological assays but was less favorable in mechanical testing, suggesting that an ideal scaffold material might combine both [18].

Scaffold architecture also plays a critical role. Teams such as Bittner’s have attempted to recreate different bone structures by designing scaffolds with multiple levels of porosity [19]. While porosity enhances biological integration, it also weakens mechanical integrity. In the event of failure, scaffolds must still provide structural support and avoid fragmentation into mobile debris that could compromise healing.

To address these limitations, ductile polymers are often blended with PLA to improve toughness and prevent brittle failure. The most widely studied are poly (caprolactone) (PCL) [20,21] and poly (butylene succinate) (PBS) [22]. PLA/PBS blends, in particular, have been extensively studied to produce tougher materials [22,23,24], promoting ductile fracture behavior and enhancing the mechanical resilience of bioresorbable scaffolds. A study by Ojansivu et al. (2018) also demonstrated that PLA/PBS and PLA/PCL blends (95/5 wt%) significantly outperformed neat PLA scaffolds in terms of osteogenesis, with cell proliferation increased by 220% after 14 days [25].

In the present work, we investigate the influence of incorporating PBS into a PLA matrix, together with the addition of barium titanate filler. PBS is expected to improve flexibility, while BaTiO₃ has been shown to enhance piezoelectric properties [15,16]. This study focuses on the mechanical behavior of 3D-printed PLA-based composites and its relationship with thermal properties and microstructure. Compression and three-point bending tests were performed, as these loading modes closely replicate physiological strain conditions. Mechanical analyses were conducted beyond the elastic limit to characterize the materials’ behavior under failure, thereby assessing their potential to support tissue regeneration post-yield. In addition, the effect of BaTiO₃ content was evaluated to identify a balance between its osteogenic benefits and its influence on the ductile-to-brittle transition.

2. Materials and Methods

2.1. Materials

Poly (lactide) (PLA) was supplied by NatureWorks (Minneapolis, MN, USA) under the reference Ingeo™ Biopolymer 2003D. Poly (butylene succinate) (PBS) was obtained from NaturePlast (Ifs, France) under the reference PBI 003. Barium titanate (mBT) was purchased from ThermoScientific (Watham, MA, USA) as Barium Titanium Oxide, with a purity of 99% and with 99.9% of the particles finer than 325 mesh.

2.2. Methods

Processing: Preparation of Compounds, Filaments, and Samples

PLA and PBS were blended at an 80/20 wt% ratio using a co-rotating twin-screw extruder to obtain a homogeneous PLA/PBS blend, as illustrated in Figure 1. Neat PLA and PLA/PBS blends were subsequently compounded with BaTiO₃ at concentrations of 5, 10, and 20 wt% using the same extrusion process. All formulations were then processed into filaments with a controlled diameter of 1.70 ± 0.1 mm using a 3Devo Filament Maker.

These filaments were then used for 3D printing with a Creality CR10-V3 FFF 3D printer. Three-dimensional models were created using 3D Builder software (v20.0.4.0) and sliced with Creality Slicer 4.8.2. Compression specimens were produced as 1 cm³ cubes in accordance with ISO 604 [26]. Rectangular rods measuring 80 × 10 × 4 mm³ (according to ISO 178 standards [27]) were printed for flexural and impact fracture testing. All mechanical test specimens were printed with 100% infill using alternating 45° and −45° raster orientations, in order to prevent void formation due to filament slippage and to reduce specimen flattening during testing. Before testing, all specimens were dried at 50 °C for three days to remove residual moisture.

2.3. Characterization Methods

To characterize both the raw materials and the fabricated products, various techniques were employed to determine thermal properties, mechanical properties, microstructure, and morphology.

2.3.1. Mechanical Analysis

As the intended application of the studied material is implant manufacturing for bone tissue engineering, the characterization of mechanical properties was focused on compression and bending tests. Compression tests were carried out using an Instron 3366 universal testing machine at a constant crosshead speed of 5 mm/min. Flexural tests were conducted on the same system at a constant speed of 2 mm/min, up to a vertical deflection of 30 mm.

2.3.2. Thermal and Microstructural Analysis

The microstructures of the 3D-printed blends and composite specimens were analyzed using differential scanning calorimetry (DSC) with a TA Q20 analyzer, at a heating rate of 2 K·min⁻¹ under a nitrogen atmosphere. Analyses were conducted during the first heating cycle, as processing-induced thermal effects cannot be excluded in 3D-printed scaffolds. Post-processing treatments such as annealing were avoided to prevent uncontrolled shrinkage that could alter scaffold architecture. Polymer and ceramic contents in the various formulations were assessed throughout the process using thermogravimetric analysis (TGA). These measurements were performed with a Netzsch TG 209 F3 Tarsus at a heating rate of 3 K·min⁻¹ under a nitrogen atmosphere.

2.3.3. Morphological Analysis

Filament diameters were continuously monitored using the integrated optical sensor, and data were recorded with Devovision software (v0.2.0). Partial or complete fractures resulting from flexural testing were analyzed using X-ray microtomography with a High-Resolution 3D Micro Computed Tomography & Digital Radioscopy Compact System (DeskTom). Scanning parameters included a tube voltage of 100 kV, a current of 100 μA, a voxel size of 10 μm, and an acquisition rate of 3 frames per second. Three-dimensional reconstruction of the samples was performed using X-ACT64, and visualization with VGStudio version 3.1.2.

3. Results and Discussion

Following the compounding of the different materials, the samples were successfully fabricated and subsequently analyzed through both thermal and mechanical testing, complemented by X-ray-assisted microstructural observations. These analyses enabled us to monitor the processing conditions and assess their influence on the properties of the printed parts.

3.1. Thermal Analysis

The targeted proportions of the different composite constituents were achieved, although a slight overlap was observed due to the proximity of the degradation temperatures of PLA and PBS (Figure 2). PLA-based composites exhibited a single degradation stage, with an onset temperature of approximately 310 °C. In contrast, PLA/PBS blend composites displayed two distinct degradation stages, consistent with the higher thermal stability of PBS, which degraded at around 340 °C.

Barium titanate remained stable within the tested temperature range, and its residual mass after matrix degradation was consistent with the expected profiles. The different formulations contained the anticipated amounts of barium titanate, namely 0%, 10 ± 0.3%, and 20 ± 0.4%, with a slight deviation to 12 ± 0.3% for the PLA/PBSmBT10 blend (

Table 1. Control of ceramic content in each compound formulation by TGA.

Compound	Desired Ceramic Percentage	Measured Ceramic Percentage
PLAmBT10	10%	10 ± 0.3%
PLAmBT20	20%	20 ± 0.4%
PLA/PBSmBT10	10%	12 ± 0.3%
PLA/PBSmBT20	20%	20 ± 0.4%

). Among all compositions, the PLA/PBS blend exhibited the greatest thermal stability, with a slight delay in PLA degradation and an overall upward shift in degradation temperatures.

Differential Scanning Calorimetry (DSC) analysis revealed distinct transition temperatures, and the degree of crystallinity of both PLA and PBS was calculated for the PLA and PLA/PBS composites (

Table 2. Transition temperatures and crystallinity in PLA- and PLA/PBS-based composites.

Material	Tg PLA (°C)	Tcc PLA (°C)	Tm PLA (°C)	χc PLA (%)	Tm PBS (°C)	χc PBS (%)
PLA	52.5 ± 0.1	95.7 ± 0.1	153.1 ± 0.1	3.4 ± 0.8	-	-
PLAmBT10	55.1 ± 0.1	92.3 ± 0.2	154.7 ± 0.1	3.6 ± 1.2	-	-
PLAmBT20	56.8 ± 0.2	92.8 ± 0.3	153.2 ± 0.2	3.4 ± 0.5	-	-
PLA/PBS	49.9 ± 0.2	89.4 ± 0.8	152.7 ± 1.1	10.8 ± 1.4	113.3 ± 0.2	33.6 ± 1.5
PLA/PBSmBT10	51.9 ± 0.1	88.8 ± 0.2	153.4 ± 0.7	11.8 ± 1.2	113.5 ± 0.6	33 ± 1.2
PLA/PBSmBT20	53.6 ± 0.2	83.5 ± 0.7	153.2 ± 0.6	11.5 ± 1.2	114.5 ± 0.6	36.7 ± 1.5

). PLA crystallinity, for example, was determined by subtracting the cold-crystallization enthalpy from the melting enthalpy. Using this method, PLA crystallinity increased by more than 220% in PLA/PBS blends compared with neat PLA. However, the cold crystallization of PLA and the melting of PBS occur in close proximity, leading to overlapping thermal signals. This overlap, also reported by Lin et al. (2022) [28], was attributed to the nucleating effect of PBS, which shifts the cold-crystallization temperature of PLA.

While this explanation is plausible, it is also possible that the proximity of these two thermal events interferes with the resolution of the analysis, without necessarily implying direct interaction. Such signal overlap could artificially inflate the apparent crystallinity of PLA while reducing the measured crystallinity of PBS (Figure 3). Lin et al. also reported a slight decrease in the glass transition temperature of PLA/PBS blends, which they interpreted as a mild plasticizing effect [28].

In addition, ceramics such as barium titanate may act as nucleating agents, as suggested by the reduced cold-crystallization temperature observed in ceramic-filled composites. This phenomenon is consistent with the findings of Mystiridou et al. (2021) [29]. However, despite the nucleation effect, no significant increase in PLA crystallinity was observed. Moreover, the observed rise in glass transition temperature in the presence of the barium titanate, from 52.5 °C to 56.8 °C, may suggest a restriction in the mobility of PLA macromolecular chains, indicating interaction at the molecular level that may influence processing and performance.

3.2. Compression Tests

All specimens were subjected to uniaxial compression testing to determine their Young’s modulus. The samples were printed with 100% infill in order to obtain values as close as possible to the intrinsic properties of the materials, thereby minimizing the influence of structural design and porosity.

The resulting stress–strain curves are shown in Figure 4. It should be noted that the curves do not represent material failure, as the tests were stopped before reaching the load cell capacity. For this reason, only the elastic properties were analyzed in compression. The curve profiles are consistent with those typically observed in standard injection-molded polymers [30].

From these curves, key elastic properties were extracted, including Young’s modulus, maximum elastic strain, and the elastic limit. However, plastic properties could not be determined since no specimen failed under the test conditions. This confirms that all formulations exhibit high resistance to compression.

The elastic properties are summarized in

Table 3. Elastic properties of the composites in compression tests.

Material	Young Modulus (MPa)	Elastic Limit (MPa)	Elastic Deformation (%)
PLA	1673 ± 29	64 ± 0.7	4.93 ± 0.43
PLAmBT10	1864 ± 35	80 ± 0.5	5.99 ± 0.17
PLAmBT20	1906 ± 109	72 ± 2.4	4.86 ± 0.24
PLA/PBS	1625 ± 38	71 ± 1.8	6.57 ± 0.24
PLA/PBSmBT10	1744 ± 20	67 ± 0.8	5.61 ± 0.10
PLA/PBSmBT20	1630 ± 12	69 ± 1.3	6.58 ± 0.3

. As expected, PLA-based composites showed greater stiffness than PLA/PBS blend-based composites, reflecting the inherently lower modulus of PBS compared to PLA. For all PLA formulations, increasing BaTiO₃ content led to a corresponding rise in Young’s modulus. This result is consistent with the thermal analyses, as the presence of barium titanate restricts the mobility of polymer chains. The reduction in stiffness due to PBS incorporation is also in line with findings by Zhang et al. (2018) [31], who reported a 16% decrease in tensile modulus with the addition of 20 wt% PBS.

In our study, however, the reduction observed under compression was less pronounced, remaining below 5%. Notably, unlike in neat PLA, the addition of BaTiO₃ to the PLA/PBS blend did not significantly increase stiffness. This discrepancy may result from the dispersion of ceramic particles within the interphase between PLA and PBS, potentially affecting stress transfer mechanisms in the composite. Despite these variations, all formulations exhibited relatively similar elastic properties, with stiffness values remaining within a narrow range.

3.3. Three-Point Bending Tests

Three-point bending tests, which involve both tensile and compressive stresses, yielded trends consistent with the compression tests. PLA and its composites exhibited higher stiffness than their PLA/PBS blend counterparts, as summarized in

Table 4. Elastic properties of the composites in 3-point bending tests.

Material	Flexural Modulus (MPa)	Elastic Limit (MPa)	Elastic Deformation (%)
PLA	3211 ± 66	100 ± 1	4.85 ± 0.12
PLAmBT10	2930 ± 141	90 ± 3	4.41 ± 0.13
PLAmBT20	3178 ± 201	90 ± 3	4.17 ± 0.14
PLA/PBS	2472 ± 60	83 ± 1	5.19 ± 0.19
PLA/PBSmBT10	2266 ± 306	73 ± 9	5.18 ± 0.34
PLA/PBSmBT20	2208 ± 126	63 ± 1	4.42 ± 0.59

.

However, unlike in compression testing, the tests are designed to achieve fracture, so the analyses focus on elastic and failure properties. In this sense, the bending tests revealed distinct differences in plastic behavior. PLA-based specimens fractured in a brittle manner, failing shortly after surpassing the elastic limit. In contrast, the PLA/PBS blend and its composite containing 10 wt% BaTiO₃ exhibited ductile behavior, deforming significantly beyond the elastic regime with a marked drop in stress, while still maintaining overall integrity (Figure 5). However, this ductile behavior was no longer observed at 20 wt% BaTiO₃.

The addition of PBS decreased overall mechanical performance, with global stiffness losses of −23%, −23%, and −31% for ceramic loadings of 0%, 10%, and 20%, respectively, as well as reductions in the elastic limit of −17%, −19%, and −30%. Nevertheless, the PLA/PBS blend could sustain greater deformation than neat PLA and PLA-based composites, with elastic strain values increasing by 7% and 17% for 0 wt% and 10 wt% BaTiO₃, respectively. This increase may be attributed to the plasticizing effect of PBS, also suggested by DSC results and supported by Lin et al. (2022) [28].

These findings are consistent with previous reports on the ductility-enhancing role of PBS in PLA-based materials, such as the review by Su et al. (2019) [22], which highlighted increased elongation at break in PLA/PBS systems. Our results confirm that this effect is preserved at moderate ceramic loadings, despite ceramics generally being associated with increased brittleness.

Nevertheless, although the PLA/PBSmBT10 composite displayed ductile fracture behavior, it remained more brittle than the pure blend. Tomography confirmed this, revealing a higher proportion of torn filament structures in the composite compared to the blend alone (Figure 6).

3.4. Fracture Morphology After Bending Tests

Representative fracture morphologies from bending tests are shown in Figure 6a,d. PLA-based composites exhibited clean, brittle fractures (Figure 6a), whereas PLA/PBS blend-based composites displayed incomplete fractures, characteristic of ductile failure (Figure 6d). X-ray microtomographic reconstructions from the viewing angles indicated in Figure 6a,d supported these observations. Consistent with macroscopic observations, PLA-based composites displayed a smooth fracture (Figure 6b,c) with minimal distinction between individual printed strings. In contrast, PLA/PBS blend-based composites exhibited plastic elongated individual strings (Figure 6e,f), with the blend keeping the most unbroken strings.

While most composites displayed smooth fracture surfaces similar to neat PLA, PLA/PBSmBT10 demonstrated a spiked fracture surface with irreversibly deformed filaments. After unloading, PLA/PBS and PLA/PBSmBT10 specimens even exhibited partial elastic recovery. The spiked fracture pattern observed in PLA/PBS blends, compared with the smooth fracture of PLA, is consistent with the findings of Lin et al. (2022) [28] and SEM observations of specimens after three-point bending tests.

3.5. Comparison with PLA/PCL/HAp Biomaterials

Comparing the results of this study with those of Pitjamit et al. [20] and Hassanajili et al. [21], the results consistently highlight the trade-off between stiffness and ductility in bioresorbable polymer-based composites for bone tissue engineering. In the present study, PLA/PBS blends reinforced with BaTiO₃ exhibited improved ductility compared to neat PLA, with compression modulus values around 1.6–1.9 GPa and flexural modulus values of 2.2–3.2 GPa, while still maintaining structural integrity. The addition of PBS reduced stiffness slightly (<8% under compression).

In contrast, the PLA/PCL/HAp systems reported by Hassanajili et al. [21] showed lower stiffness in compression (between 0.3 and 1.3 GPa) due to a higher proportion of flexible polymer (PCL). The review and experimental results presented by Pitjamit et al. [20] were similar to our own, with compressive stiffnesses of between 1 and 1.2 GPa and flexural stiffnesses of around 2.4 GPa. The small differences come from the formulations with slightly more flexible polymers (30% PCL whereas we only have 20% PBS) for similar levels of mineral filler (15% in our study, 20% in ours). These systems demonstrated compressive and flexural properties in the same range as the present study, confirming that blending strategies can mitigate brittle failure while preserving sufficient stiffness for scaffold applications.

Taken together, these results show that PLA/PBS/BaTiO₃ composites occupy an intermediate position: they are less stiff than highly ceramic-filled PLA composites (PLA/HAp), but significantly more ductile, which is advantageous for temporary bone scaffolds where resilience and structural integrity under strain are critical. Moreover, the potential piezoelectric contribution of BaTiO₃ provides an additional functional advantage not addressed in PLA/PCL/HAp systems.

4. Conclusions

This study successfully demonstrated the fabrication of novel composites using Fused Filament Fabrication (FFF) additive manufacturing, with the objective of reducing the risk of brittle failure in bone scaffolds. The PLA/PBS blend matrix and its composites exhibited ductile failure behavior, whereas neat PLA-based materials displayed brittle fracture. This ductility is of particular importance, as it ensures that the scaffolds can preserve their structural role and maintain integrity under mechanical loading.

Although the incorporation of PBS led to a reduction in stiffness compared with neat PLA composites, this decrease in Young’s modulus was modest—less than 8% under compression. Such a reduction is negligible when compared with the significant stiffness gap between polymer-based composites and natural bone tissue, which can reach up to 200 GPa [8]. Since these materials are bioresorbable and are not designed for permanent in vivo implantation, replicating the exact stiffness of bone is not required. Instead, their combination of mechanical resilience and ductility positions them as promising candidates for safe, temporary support during bone regeneration.

Comparable bioresorbable systems used in bone surgery, such as PLA/PCL blends or PLA-based composites reinforced with hydroxyapatite (HAp) [20], exhibit stress and stiffness ranges similar to those reported here, further validating the relevance of PLA/PBS/BaTiO₃ systems as temporary scaffolding materials that combine mechanical support with biological compatibility.

Future work will focus on evaluating the piezoelectric behavior of these composites, in order to assess their potential influence on osteogenesis. Particular attention will be given to the role of PBS in modulating ductility, the associated reduction in stiffness of 3D-printed parts, and its possible effect on charge generation during mechanical loading.