Accelerated Degradation of Poly-ε-caprolactone Composite Scaffolds for Large Bone Defects

This research investigates the accelerated hydrolytic degradation process of both anatomically designed bone scaffolds with a pore size gradient and a rectangular shape (biomimetically designed scaffolds or bone bricks). The effect of material composition is investigated considering poly-ε-caprolactone (PCL) as the main scaffold material, reinforced with ceramics such as hydroxyapatite (HA), β-tricalcium phosphate (TCP) and bioglass at a concentration of 20 wt%. In the case of rectangular scaffolds, the effect of pore size (200 μm, 300 μm and 500 μm) is also investigated. The degradation process (accelerated degradation) was investigated during a period of 5 days in a sodium hydroxide (NaOH) medium. Degraded bone bricks and rectangular scaffolds were measured each day to evaluate the weight loss of the samples, which were also morphologically, thermally, chemically and mechanically assessed. The results show that the PCL/bioglass bone brick scaffolds exhibited faster degradation kinetics in comparison with the PCL, PCL/HA and PCL/TCP bone bricks. Furthermore, the degradation kinetics of rectangular scaffolds increased by increasing the pore size from 500 μm to 200 μm. The results also indicate that, for the same material composition, bone bricks degrade slower compared with rectangular scaffolds. The scanning electron microscopy (SEM) images show that the degradation process was faster on the external regions of the bone brick scaffolds (600 μm pore size) compared with the internal regions (200 μm pore size). The thermal gravimetric analysis (TGA) results show that the ceramic concentration remained constant throughout the degradation process, while differential scanning calorimetry (DSC) results show that all scaffolds exhibited a reduction in crystallinity (Xc), enthalpy (Δm) and melting temperature (Tm) throughout the degradation process, while the glass transition temperature (Tg) slightly increased. Finally, the compression results show that the mechanical properties decreased during the degradation process, with PCL/bioglass bone bricks and rectangular scaffolds presenting higher mechanical properties with the same design in comparison with the other materials.


Introduction
A major concern related to the design of polymer-based bone tissue engineering scaffolds is related to the degradation process and degradation kinetics of these constructs as well as the rate of tissue regeneration. After generating an in vitro cell culture of a scaffold/tissue system, the degree of remodelling and replacement of the biological implant by the native tissue needs to be considered [1][2][3][4][5][6][7]. Tissue remodelling is important to obtain

Degradation Procedure
Accelerated degradation studies were conducted using sodium hydroxide (NaOH) of 5 mol/L (5 N) in aqueous solution (VWR, Radnor, Pennsylvania, USA) with a density of 1.185 gr/cm 3 (20 °C), a solubility of 20 °C and a pH of 14 (H2O, 20 °C). The degradation period took place for 5 days. On each day, 5 samples were used from each considered case and measured using a high-precision balance. At each time point, the samples were removed from the NaOH and washed three times with the use of deionised water and left to dry overnight. Once completely dry, the samples were measured to determine the weight reduction. The amount of NaOH used for 5 rectangular scaffolds was 15 mL and for the 5 anatomically designed bone bricks, the amount was 50 mL (due to their size). The pH was monitored throughout the experimental work, and no changes were observed (pH of 14).

Degradation Procedure
Accelerated degradation studies were conducted using sodium hydroxide (NaOH) of 5 mol/L (5 N) in aqueous solution (VWR, Radnor, PA, USA) with a density of 1.185 gr/cm 3 (20 • C), a solubility of 20 • C and a pH of 14 (H 2 O, 20 • C). The degradation period took place for 5 days. On each day, 5 samples were used from each considered case and measured using a high-precision balance. At each time point, the samples were removed from the NaOH and washed three times with the use of deionised water and left to dry overnight. Once completely dry, the samples were measured to determine the weight reduction. The amount of NaOH used for 5 rectangular scaffolds was 15 mL and for the 5 anatomically designed bone bricks, the amount was 50 mL (due to their size). The pH was monitored throughout the experimental work, and no changes were observed (pH of 14).

Morphological Characterization
Scanning electron microscopy (SEM) was used to investigate the morphological characteristics of the samples and to determine pore sizes (FEI ESEM Quanta 250, FEI Company, Hillsboro, OR, USA). Scaffolds were coated (platinum coating) with the use of an EMITECH K550X sputter coater (Quorum Technologies, Laughton, East Sussex, UK) before imaging. The SEM images were analysed using ImageJ 1.x (National Institutes of Health, Bethesda, MD, USA) (10 measurements per sample).
The porosity of the scaffolds was calculated with the use of the following equation: where V p is the pore volume, V is the total bulk volume and P t is the porosity. The density of PCL used in these experiments was calculated as follows: where ρ is the density of the unprocessed material, m is the mass measured and V is the volume measured. The calculated density of PCL was 1.124 ± 0.003 g/cm 3 .

Thermal Gravimetric Analysis
Thermal gravimetric analysis, using a TGA Q500 (TA Instruments, New Castle, UK), was used to investigate the thermal degradation and to calculate the ceramic content on the printed scaffolds. Experiments were repeated 4 times per considered scaffold (n = 4).
The tests were conducted in air atmosphere (50 mL/min) with a temperature ranging from 25 • C to 1000 • C at a rate of 10 • C/min. The weight of each sample was 200 mg, and each test was conducted twice.

Differential Scanning Calorimetry
Differential scanning calorimetry (DSC) tests were performed to determine the melting temperature (Tm), the enthalpy (∆Hm), the crystallinity and the glass transition temperature (Tg) (n = 4). Tests were conducted using a TA Q100 (TA Instruments, New Castle, UK) under a nitrogen/air atmosphere (50 mL/min). The heating cycle was as follows: heating from −90 • C to 100 • C at a rate of 10 • C/min and then keeping stable for 2 min. The weight of each considered sample was 20 mg.

Mechanical Testing
Compression tests were performed using an INSTRON 3344 (Instron, High Wycombe, Buckinghamshire, UK) (n = 4), at different degradation time points, according to the ASTM D695-15. Force versus displacement curves obtained using the Bluehill Universal Software (Instron, High Wycombe, Buckinghamshire, UK) were stress-strain curves. The compressive modulus was determined based on the slope of the elastic region of the stress-strain curves.

Data Analysis
Statistical analysis was performed using a one-way analysis of variance (one-way ANOVA) with Tukey's post hoc test (GraphPad Software Inc., San Diego, CA, USA). Weight loss, TGA and DSC data were analysed using Origin 2021 (Origin Lab Corporation, Northampton, MA, USA) and are presented as average values of the obtained results. degradation process, with the PCL bone bricks exhibiting the lowest weight reduction at day 5 (5.07%) and PCL/bioglass the highest weight loss (90%) followed by PCL/HA (81.41%) and PCL/TCP (10.85%). In the case of rectangular scaffolds, the results show that PCL/bioglass samples also degrade faster than their PCL/HA and PCL/TCP counterparts. Moreover, it was possible to observe that the increase in pore size accelerated the degradation process. The PCL/bioglass scaffolds with a 200 µm pore size were fully degraded after day 4, while the scaffolds with 300 µm and 500 µm pore sizes degraded after day 3. For PCL/HA, the results show that all samples (200 µm, 300 µm and 500 µm) were fully degraded after day 4. In the case of the PCL/TCP scaffolds, which were not fully degraded after day 5, the weight loss increased from 24.45% for scaffolds with a 200 µm pore size to 35.25% for scaffolds with a 500 µm pore size. A similar trend was observed for the PCL scaffolds (showing an 11.64% weight loss at day 5 for scaffolds with a 200 µm pore size and a 13.37% weight loss at day 5 for scaffolds with a 500 µm pore size), which exhibited the lowest degradation kinetics. This can be explained by the increase in the surface area exposed to the NaOH that accelerates the hydrolytic degradation and the release into the liquid medium of the ceramic particles previously bonded with the polymeric material [51][52][53][54][55][56][57][58]. Results also indicate that, for the same material composition, bone bricks degrade slower compared with rectangular scaffolds. As the overall porosity of the bone bricks is 52%, while the porosity of the rectangular scaffolds varies between 42% (scaffolds with a 200 µm pore size) and 56% (scaffolds with a 500 µm pore size), the observed differences can be attributed to the pore size gradient created in the bone bricks, which superimposes both the porosity and overall pore size (with an average of 350 µm in the bone bricks). 670 5 of 16 the degradation process, with the PCL bone bricks exhibiting the lowest weight reduction at day 5 (5.07%) and PCL/bioglass the highest weight loss (90%) followed by PCL/HA (81.41%) and PCL/TCP (10.85%). In the case of rectangular scaffolds, the results show that PCL/bioglass samples also degrade faster than their PCL/HA and PCL/TCP counterparts. Moreover, it was possible to observe that the increase in pore size accelerated the degradation process. The PCL/bioglass scaffolds with a 200 μm pore size were fully degraded after day 4, while the scaffolds with 300 μm and 500 μm pore sizes degraded after day 3. For PCL/HA, the results show that all samples (200 μm, 300 μm and 500 μm) were fully degraded after day 4. In the case of the PCL/TCP scaffolds, which were not fully degraded after day 5, the weight loss increased from 24.45% for scaffolds with a 200 μm pore size to 35.25% for scaffolds with a 500 μm pore size. A similar trend was observed for the PCL scaffolds (showing an 11.64% weight loss at day 5 for scaffolds with a 200 μm pore size and a 13.37% weight loss at day 5 for scaffolds with a 500 μm pore size), which exhibited the lowest degradation kinetics. This can be explained by the increase in the surface area exposed to the NaOH that accelerates the hydrolytic degradation and the release into the liquid medium of the ceramic particles previously bonded with the polymeric material [51][52][53][54][55][56][57][58]. Results also indicate that, for the same material composition, bone bricks degrade slower compared with rectangular scaffolds. As the overall porosity of the bone bricks is 52%, while the porosity of the rectangular scaffolds varies between 42% (scaffolds with a 200 μm pore size) and 56% (scaffolds with a 500 μm pore size), the observed differences can be attributed to the pore size gradient created in the bone bricks, which superimposes both the porosity and overall pore size (with an average of 350 μm in the bone bricks). The effects of the scaffold architecture and material composition on the degradation kinetics can also be observed in Figures S1-S15 showing SEM images that, for the different The effects of the scaffold architecture and material composition on the degradation kinetics can also be observed in Figures S1-S15 showing SEM images that, for the different  Tables 1-4 it is possible to observe that both pore size and porosity increase during the 5 days of degradation.   Table 3. Pore size and porosity values of scaffolds (300 µm pore size) before and after the degradation process.

Thermal Gravimetric Analysis
The thermal gravimetric analysis (TGA) results show that all considered samples exhibit degradation temperatures between 304.12 • C and 437.11 • C, after which only the inorganic materials remain (Figures S16 and S17). Moreover, the addition of ceramic particles into the polymer matrix reduces the degradation temperature, which can be observed for all samples throughout the degradation period. Moreover, the results in Table 5 and Figures S16 and S17 show that the overall weight ratio between polymer and ceramic material remains almost constant, while the total amount of ceramic materials significantly decreases, indicating a significant loss of polymer. For example, at day 3, the amount of HA in the samples is 22.4 wt%, and the total amount of HA is 9.4% of the initial HA content, while for the PCL/bioglass samples, the amount of bioglass in the samples is 22.1 wt%, and the total amount of bioglass is 7.1% of the initial bioglass content. These results clearly show a significant loss of polymer in the case of bioglass and confirm the observations discussed in Section 3.2 suggesting that the addition of ceramic materials (bioglass in particular) accelerates the degradation process. Additionally, this behaviour can also be explained by the morphological changes in the polymeric matrix, as discussed in Section 3.2. As samples were printed at 90 • C, the results show that the printing conditions do not induce any material degradation. Additionally, the levels of HA, TCP and bioglass  (Table 1) suggest that the melt blending approach used to prepare the blends is a simple and efficient method.

Chemical Analysis
DSC was used to investigate the crystallinity (Xc), melting temperature (Tm), enthalpy (∆m) and glass transition temperature (Tg) of bone bricks and rectangular scaffolds during the degradation process (Table 6 and Figure S18). The results show that the addition of ceramic materials into the polymer reduces the crystallinity (from 86.91% to 61.08%), the enthalpy (from 89.89 J/g to 65.06 J/g) and the melting temperature (from 64.65 • C to 69.82 • C), while the glass transition temperature slightly increases (from −59.19 • C to −59.1 • C) [58][59][60][61][62][63]. Similar results showing that the addition of ceramic particles constrains the crystallization process, limiting the mobility of PCL chains in the polymer-ceramic matrix and inducing the formation of smaller or thinner regions of crystalline lamellae, were also reported by other groups [64][65][66][67]. The results show that the melting temperature reduced for all the material compositions from day 0 to day 5, with the highest reduction observed for the bioglass scaffolds (there were no significant differences between PCL/bioglass and PCL/HA) and the lowest for TCP scaffolds. This can be explained by the decrease in crystallinity (with the highest reduction in the case of the PCL/HA and PCL/bioglass samples), and the melting temperature can be interpreted as the energy required by a system to destroy ordered regions. The results also show that the degradation occurs mainly through the destruction of the crystalline regions, which is aligned with the variations in both enthalpy and glass transition temperature that decrease and increase during the degradation time, respectively.

Mechanical Analysis
The compressive modulus results for all considered samples at different degradation time points are indicated in Figure 3. As observed, at day 0 and for the same material composition, the printed bone scaffolds exhibited higher compressive moduli than their rectangular scaffold counterparts, indicating the relevance of the scaffold architecture on the mechanical performance. In the case of the rectangular scaffolds, the results show that for the same material composition, scaffolds with large pore sizes (500 µm) present lower compressive moduli than scaffolds with the lowest pore sizes (200 µm), which can be attributed to the increase in porosity. Furthermore, throughout the degradation process it can be observed that for bone bricks, the PCL/bioglass bone bricks present higher compressive moduli, while for rectangular scaffolds, the PCL/TCP scaffolds present higher compressive properties for all the different pore sizes. Moreover, compressive moduli decreased for all samples throughout the degradation process, which is associated with the decrease in crystallinity, decrease in the filament diameters, increase in porosity, filament collapse and limited adhesion between filaments. for the same material composition, scaffolds with large pore sizes (500 μm) present lower compressive moduli than scaffolds with the lowest pore sizes (200 μm), which can be attributed to the increase in porosity. Furthermore, throughout the degradation process it can be observed that for bone bricks, the PCL/bioglass bone bricks present higher compressive moduli, while for rectangular scaffolds, the PCL/TCP scaffolds present higher compressive properties for all the different pore sizes. Moreover, compressive moduli decreased for all samples throughout the degradation process, which is associated with the decrease in crystallinity, decrease in the filament diameters, increase in porosity, filament collapse and limited adhesion between filaments.  (C) rectangular scaffolds with 300 µm pore size; (D) rectangular scaffolds with 500 µm pore size; (E) PCL bone bricks and rectangular scaffolds; (F) HA bone bricks and rectangular scaffolds; (G) TCP bone bricks and rectangular scaffolds; and (H) bioglass bone bricks and rectangular scaffolds. * Statistical evidence (p < 0.05) analysed by one-way ANOVA, and Tukey post hoc test. The * statistical evidence (p < 0.05), **, *** and **** is the one-way analysis of variance (one-way ANOVA) and Tukey's post hoc test with the use of GraphPad Prism software and is used to show the difference between the results. The * is a small difference, while more * are added as the differences between the results increases.

Conclusions
The degradation kinetics of anatomically designed scaffolds with pore size gradients (bone bricks) and rectangular scaffolds with different pore sizes, considering a range of material compositions, was investigated, taking into consideration weight loss and morphological, thermal, chemical and mechanical changes. Due to the long degradation time of PCL, accelerated degradation, using NaOH, was considered. The results show that bone bricks present faster and more controlled degradation in comparison with rectangular scaffolds. The fastest degradation (weight loss) was observed for PCL/bioglass bone bricks and rectangular scaffolds. Moreover, for the same material composition, the degradation is faster in scaffolds presenting large pore sizes. The TGA results show that, in all considered cases, the concentration of the inorganic material remained the same during the degradation process. The DSC results indicate that the crystallinity of all samples at different degradation times decreased, suggesting a faster destruction of the crystalline regions than the amorphous ones. This observation is contrary to other studies suggesting that the degradation of PCL occurs through an initial mass loss that occurs due to the random hydrolytic split of polymeric chains in the amorphous regions, followed by a gradual degradation in the crystalline regions [63][64][65][66]. High levels of crystalline regions on the surfaces of the printed filaments together with the penetration of the degradation medium promoting internal degradation may explain the obtained results. Finally, as expected, compressive moduli decreased throughout the degradation process. However, for the same material content and degradation time point, bone bricks present better mechanical properties than rectangular scaffolds. Moreover, for the same material composition, better mechanical properties were observed for scaffolds with lower pore sizes. Overall, the accelerated degradation process showed that PCL/bioglass bone brick scaffolds present faster and more controlled degradation kinetics, compared with the other material concentrations and scaffold designs, and higher mechanical properties, making it the most suitable physical support for bone tissue applications.