Calcium phosphate stability on melt electrowritten PCL scaffolds

Calcium phosphate (CaP) coating on melt electrowritten (MEW) substrates is a potential candidate for bone regeneration influencing the interaction of osteoblasts with implanted scaffolds. Pretreatment to improve hydrophilicity of the hydrophobic polymer fibres affects subsequent coating with bioactive compounds like CaP. Therefore, this study evaluated the subsequent stability and structural properties of CaP coated MEW Poly-ε-caprolactone (PCL) scaffolds following pre-treatment with either argon-oxygen plasma or sodium hydroxide (NaOH). Scanning electron microscopy and m-CT showed uniform CaP coating after one hour immersion in simulated body fluid following plasma pretreatment. Moreover, fourier transform infrared spectroscopy, energy dispersive spectrometry and X-ray diffraction analysis confirmed the presence of hydroxyapatite, tetracalcium phosphate and halite structures on the coated scaffolds. Contact angle measurement showed that the plasma pretreatment and CaP coating improved the hydrophilicity of the scaffold. However, the mechanical properties of the scaffolds were degraded after both plasma and NaOH treatments. The tensile stability was significantly improved following mineralization in plasma-treated scaffolds due to the smaller crystal size formed on the surface resulting in a dense CaP layer. The results obtained by thermogravimetric analysis also confirmed higher deposition of CaP particles on coated scaffolds following plasma modification. The results of this study show that plasma pre-treated mineralized MEW PCL scaffolds are sufficiently stable to be useful for further development in bone regeneration applications. © 2020 The Authors. Publishing services by Elsevier B.V. on behalf of Vietnam National University, Hanoi. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).


Introduction
The remodeling of the bone tissue around implanted materials is influenced by the surface charge and chemistry of the implanted materials [1]. PCL is a biodegradable polyester widely used as an implantable biomaterial [2]. However for tissue engineering purposes, PCL has some significant shortcomings such as slow degradation rate, hydrophobic properties and low cell adhesion [3]. The incorporation of CaP into PCL has yielded a class of hybrid biomaterials with remarkably improved mechanical properties, controllable degradation rates, and enhanced bioactivity as calcium and phosphate ions are essential for skeletal mineralization where mineral crystals are deposited in an organized fashion onto the organic ECM [4]. Moreover CaP coating imparts an increased surface roughness to coated scaffolds. Rough implant surfaces enhance the contact between the implant and the bone tissue improving subsequent integration [5]. Coating biocompatible substrates with these inorganic crystals has subsequently shown the significant bone growth and vascularization [6] including CaP coated electrospun poly (ethylene oxide terephthalate)Àpoly(buthylene terephthalate) scaffolds in vivo [7].
Bone calcification and maturation can be stimulated by releasing calcium ions [8]. Calcium and phosphorus ions released from coated scaffolds can adjust the ion concentration and local pH of the environment, affecting protein adhesion, attachment of the osteoblasts and their activation which has an impact on bone regeneration [9]. Following coating, CaP crystal structure, surface area and particle size as well as the temperature, acidity and fluid movement within a coated scaffold can all affect the dissolution process [10,11]. Furthermore, changes to the pore size and pore number in CaP particles will enhance body fluid convection due to better contact between the CaP crystal surface area and body fluids [12]. On the other hand, greater porosity also results in poor mechanical properties and CaP coated layers displayed a weak load-bearing capacity [13].
Surface activation by pretreating the substrate material has been reported to affect the rate of coating formation [14]. Various approaches have been tried to improve subsequent CaP deposition onto PCL scaffolds [15] including O 2 plasma treatment [16], chemical modification [17], film deposition [18], thermal and lipase dependent surface modification [19] and etching in alkaline and acidic solutions [20]. Similar methods of activation have been used with electrospun fibrous scaffolds pior to CaP coating e.g. gelatin treated poly lacticglycolic acid (PLGA) scaffolds to produce positively charged groups [21] and ethanol treatment on electrospun PCL, poly(3hydroxybutyrate) (PHB) and polyaniline (PANi) polymers [22].
Although other studies demonstrated the production of CaP on solution electrospun scaffolds with nanometer scale fibres (300 nme1 mm) [7,23,24], no quantitative studies are available comparing the stability characteristics of CaP coated MEW scaffolds with micrometer scale fibres (2e50 mm) following plasma and NaOH pre-treatment. This study shows the great potential of evaluating the CaP stability on the scaffold constructs with largersized fibre dimension. Accordingly, our study characterized the effects of NaOH and argon-oxygen (AreO 2 ) plasma pre-treatment on the CaP coated MEW PCL scaffolds using scanning electron microscopy (SEM), fourier transform infrared spectroscopy (FTIR), energy dispersive spectrometry (EDS), micro-CT (m-CT), thermogravimetric analysis (TGA), X-ray diffraction (XRD), mechanical tests and contact angle as the CaP stability is critically important for later potential bone engineering applications.

Experimental
The MEW printer used in this study contained a high voltage source (DX250R, EMCO, Hallein, Austria) controlled by a voltage divider (Digit Multimeter 2100, Keithley, Cleveland, USA), a pneumatically regulated melt feeding system (FESTO, Berkheim, Germany) and a planar movable aluminium collector plate (XSlide, Velmex, New York, USA) controlled by G-code (MACH 3 Computerized Numerical Control (CNC) software, ARTSOFT, Livermore Falls, USA). A proportional-integral-derivative controller was used to regulate the electrical heating system (TR400, Delta-t, Bielefeld, Germany) to assure a stable melt temperature profile.
Two grams of medical-grade 80 kDa PCL pellets (Corbion, Australia) was placed in a 2 mL syringe with a 21G nozzle, and heated to 80 C for 30 min to melt before insertion into the MEW heated head. The feed rate was 20 mL/h, which was controlled via compressed air. A threshold voltage between 5 and 7 kV was applied to create the charged polymer and to form a Taylor cone. The XeY movement of the collector platform was controlled using programmable software (G-code) that places the deposited polymer fibres in the desired pattern. From our previous studies and other reports [25,26], an optimal scaffold pore size for bone regeneration is in the range of 100e400 mm. In this study, the averagepore size of 250 mm was designed and printed.
MEW PCL scaffolds (2 Â 2 cm) were placed in 100% ethanol for 15 min under a vacuum to remove any residual contamination before allocation into one of five treatment groups: (1) Control group (nC) e non coated; (2) NaOH treatment (Na-nC) e scaffolds immersed in pre-warmed 1 M NaOH at 37 C for 30 min then washed with Milli Q water until the pH was neutralized; (3) Plasma treatment (Plas-nC) e Ar and O 2 plasma cleaned at 10.15 W for 7 min each side under vacuum (PDC-002-HP, Harrick Plasma, USA); (4) NaOH treatment þ CaP coating (NaeC) e NaOH treatment of scaffold as (2) above followed by immersion in highly saturated SBF (10x) solution [27] at 37 C for 0.5, 1, 3 and 6 h. The SBF was replaced every 30 min. After washing the scaffolds in Milli Q water, they were immersed in 0.5 M NaOH at 37 C for 30 min. Finally, the scaffolds were rinsed with distilled water then collected for freeze drying overnight; (5) Plasma treatment þ CaP coating (Plas-C) e Plasma treatment of scaffold as (3) above followed by SBF as (4) above.
To characterize the surface morphology of the MEW scaffolds, the samples were coated with gold and examined with a scanning electron microscope (Jeol JCM-5000) operating at 15 kV accelerating voltage.
Scaffolds were cut into 6 mm discs using a tissue biopsy punch (kai Europe GmbH, Solingen, Germany) and coated with gold. The elemental analysis was performed by JSM-7800 scanning electron microscope (Japan), equipped with energy dispersive X-ray spectroscopy (INCA, Oxford Instruments, UK).
The scaffold hydrophilicity was assessed by measuring the water contact angle using a Contact Angle and Surface Tension instrument (FTA200, Poly-Instruments Pty. Ltd., Australia) running with the following parameters; pump speed 2 ml/s, needle diameter 0.279 mm, water droplet diameter 1.0 mm. Three different locations on the sample were selected to measure the angle between the surface and a liquid droplet. Images were captured via a CCD video camera running in real time and saved for further analysis.
Tensile strength tests were performed on all five groups of coated and non-coated PCL scaffolds using an electromechanical Micro-Tester (Instron 5848, Norwood, Ma) with a 500 N load cell and a gauge length of 15 mm (5 samples/group). Samples 45 Â 10 mm and 1 mm thick were prepared and stretched at a speed of 15 mm/min until breakage. The subsequent slope of each stressestrain curve was analysed.
X-ray diffraction of the scaffolds was recorded using a Cu-K a1 operating at 40 kV, 40 mA. The scanswere performed on powder from 10 to 40 scanning range, a step size of 0.04 and irradiation time of 0.96 s per step. The mean crystallite size was determined using the system software (DIFFRAC SUITE EVA). FTIR spectroscopy (Bruker Vertex 70 spectrometer) was used to characterize the functional groups on the scaffolds. Four different points on each sample were analysed. The diamond anvil cell (DAC) was placed on the aligned orientation of the sample and screwed until touch the sample. The scan test samples was analysed for chemical properties.
Thermal behaviour of 20 mg of each of the CaP coated PCL scaffolds were examined at a temperature range of 25e600 C with a heating rate 10 K min À1 (Netzsch Jupiter Simultaneous Thermal Analyser, Germany).
The distribution of CaP in the scaffolds was examined by m-CT.
A 6-mm disc of each scaffold was placed inside the X-ray tube of a micro-CT scanner (mCT40, SCANCO Medical AG, Brüttisellen, Switzerland) and exposed to 55 kV of X-rays with a current of 120 mA. Analysis was performed using a greyscale threshold of 10 and resolution of 6 mm. The m-CT software package was used for 3D visualization of the scaffolds reconstructed from the 2D scanned slices. The fibres showing in grayscale images were eliminated by selecting a suitable threshold corresponding to the CaP particle distribution. The volume of mineralisation in the test constructs (NaeC and Plas-C) was approximated by subtracting the mean volume of the control (nC) scaffold using CTAn program.
All data were expressed as mean ± standard deviation. Comparisons between groups were analysed by analysis of variance (ANOVA, post hoc test: Tukey). The statistical software SPSS 17.0 for windows was used for calculations and p < 0.05 was considered to be statistically significant.

Morphological characterization of scaffolds (SEM)
SEM images of the scaffold structures showed that the scaffolds retained their porous nature after CaP coating ( Figure S1). 0.5 h SBF treatment did not fully cover the whole fibre surface ( Figure S1-a), while immersion for 1 h provided uniform coating of the structures in both NaeC and Plas-C groups ( Figure S1-b). Morphologically, the CaP clusters formed were more spherical in arrangement on the NaeC scaffold (Figure S1-b2) in comparison with Plas-C scaffold, where they were distributed smoothly (Figure S1-b4). After 3 and 6 h immersion in SBF, there was an increase in crystalline deposition and a thick layer of CaP particles encased the fibres which reduced the scaffold pore size ( Figure S1-c, d). Fig. 1-a showed the morphology of PCL surface scaffold before CaP coating (nC). The coating structure in NaeC scaffolds showed some cracks and separation of the coated layer from the fibres after immersion for 1 h ( Fig. 1-b1, yellow arrows) whereas a dense evenly coated layer appeared on the Plas-C scaffolds ( Fig. 1-d1). NaOH or AreO 2 plasma treatment alone did not appear to have any significant impact on the fibre diameter ( Fig. 1-c, e) although some degradation and peeling of the outer layer of the Na-nC scaffold was apparent ( Fig. 1-c2). Also, the surface of Plas-nC scaffold displayed a relatively rough morphology with nanometre features on the surface of the fibres (Fig. 1-e2).

Elemental characterization (EDS)
EDS analysis identified the proportion of elements found on the scaffold areas through percentage in weight. As expected EDS analysis showed the presence of calcium on the surface of both NaeC and Plas-C scaffolds ( Figure S2, Table 1). Pre-treatment with AreO 2 plasma however increased the level of Ca to 6.7% in Plas-C compared to 2.7% in the NaeC group (was treated with NaOH). Phosphorous however was not detected in NaeC scaffold while it was 1.7% in the Plas-C scaffolds suggesting pre-treatment with AreO 2 plasma may influence the Ca/P ratio. Sodium as expected was higher in NaeC (7.0%) than Plas-C scaffolds (2.2%). Also, the Plas-C scaffold showed the presence of K and Mg ions which were not found on the other scaffolds. The presence of Copper was observed in both Plas-nC and Plas-C scaffolds.

Surface evaluation by contact angle (CA)
The hydrophilicity of the treated and untreated scaffolds was assessed by contact angle measurement (Fig. 2). We observed that nC scaffolds showed the hydrophobic nature of PCL with an average contact angle of 135 ± 4.9 ( Fig. 2-a). CaP coating significantly increased hydrophilicity of the scaffold surface (CA ¼ 0 ) in both NaeC and Plas-C groups ( Fig. 2-b, d). Treatment with 1M NaOH only slightly decreased the contact angle (91 ± 12.4 ) in Na-nC scaffolds (Fig. 2-c) wheras Plasma treatment alone also significantly increased hydrophilicity of the Plas-nC scaffolds surface (CA ¼ 0 ) (Fig. 2-e).

Mechanical properties
Assessment of the mechanical performance of the PCL scaffolds was carried out and the mechanical properties were calculated from the curve (Fig. 3, Table S1). Apparent stress-strain relationships were recorded ( Fig. 3-a) and the Young's modulus of nC scaffold (1.93 ± 0.23 kPa) was shown to be the highest of the scaffold groups. Plas-nC and Na-nC scaffolds both markedly reduced Young's modulus (0.57 ± 0.27 and 0.85 ± 0.47 kPa respectively). Subsequent CaP coating of the plasma treated samples however almost restored the Young's modulus to pretreatment levels (Plas-C 1.67 ± 0.76 kPa) in contrast to NaeC samples where the Youngs modulus was only increased minimally after coating ( Fig. 3-b).
The nC scaffold also showed the highest elongation failure value (1088.2 ± 121.4%) and ultimate tensile strength (29.66 ± 1.37 kPa) compared to the other scaffolds (Fig. 3-c, d, Table S1). Similar to the Youngs modulus results, plasma and NaOH treatments again decreased elongation failure values and ultimate tensile strength, but these indicators of tensile strength were able to be partially restored by coating with CaP. Overall, the nC and Plas-nC scaffolds showed the highest and the lowest potential to tolerate tensile loading (p 0.002), respectively.

X-ray diffraction (XRD) analysis
XRD spectra of the scaffolds are shown in Fig. 4. The diffraction peaks at 2q ¼ 21.60 and 23.95 (asterisks) attributed to PCL were seen in all groups. The absence of crystalline CaP revealed that no coating materials were found in nC specimens. Major pattern peaks at 2q ¼ 31.73 , 66.34 and 75.12 (triangle) could be assigned to the halite structure of NaCl while diffraction peaks at 2q ¼ 11.92 , 29.74 and 34.01 (dot) corresponded to the formation of the Halite and brushite crystalline forms were distinguished in Plas-C and NaeC scaffolds by their difference in crystal orientation (200 for NaeC and 220 for Plas-C scaffolds). In addition, the crystals looked sharp in shape and larger in size for the NaeC scaffolds. Approximately 36.77% halite and 63.23% brushite structures were found in Plas-C scaffolds. The crystal sizes of both coated scaffolds (NaeC and Plas-C) were determined by Scherrer's equation and the average crystal sizes were; PCL (11.72 ± 0.23 nm), halite (44.53 ± 2.76 nm), brushite (50.93 ± 14.44 nm for Plas-C scaffolds and PCL (11.60 ± 0.12 nm), halite (237.27 ± 121.67 nm), brushite (41.63 ± 6.14 nm) for NaeC scaffolds.
Although some small crystalline structures were observed in Na-nC and Plas-nC scaffolds, the absence of crystalline CaP in Na-nC and Plas-nC scaffolds demonstrated this was not due to the coating materials. Fig. 5 presents the FTIR spectra of the PCL scaffold groups. In nC scaffold, peaks associated with CeOeC at 1161 cm À1 ,CeOeC at 1239 cm À1 ,CeO and CeC at 1293 cm À1 , carbonyle stretching at 1721 cm À1 , CH 2 stretching at 2946 cm À1 and CH 2 stretching at 2866 cm À1 were identified.

FTIR analysis
The FTIR spectra of CaP on the surface of NaeC scaffold group, showed bands corresponding to OH À stretching at 3341 cm À1 , asymmetric PO 4 3bending at 558 and 603 cm À1 , asymmetric PO 4 3stretching at 1026 cm À1 and symmetric PO 4 3stretching at 959 cm À1 .
Hydrophilic groups at 2942 cm À1 were identified on the surface of the Na-nC scaffold while for the Plas-C scaffolds, the following absorption bands were identified; OH À stretch at 3366 cm À1 , asymmetric PO 4 3bend at 564 cm À1 , symmetric PeO stretch at 960 cm À1 and asymmetric PO 4 3stretch at 1045 cm À1 . The hydrophilic OH bands at 2943 cm À1 corresponded to the surface of the Plas-nC scaffold.

Thermal analysis (TGA)
The TGA-DSC curves of the CaP coated scaffolds (na-C and Plas-C) were obtained under N 2 atmosphere (Fig. 6). Weight loss occured over three temperature ranges as detailed in Table 2 and Figure S3. The first temperature range (25e193.4 C) was associated with a mass loss of 0.45% for the NaeC scaffold and 0.86% for the Plas-C scaffold at the endothermic peak of 64.9 C. The preliminary decomposition occurred in the range 193.4e431 C with a weight loss of 41.66% and 39.15% for the NaeC and Plas-C scaffolds respectively at the maximum peak temperature of 390.3 C and 393.3 C respectively. Further decomposition occurred between 431.5 and 600 C with the highest exothermic peak of 517.9 and 514 C for NaeC and Plas-C scaffolds respectively. Following the complete degradation of the material, the residual CaP particles was 58.23% and 56.38% for Plas-C and NaeC scaffolds respectively at 600 C.

Characterization of PCL scaffolds coated with CaP (m-CT)
The effect of the different coating treatments was evaluated using m-CT to determine the distribution of CaP particles within the scaffolds (Fig. 7). No CaP coating was identified in the control (nC), Na-nC and Plas-nC scaffolds ( Fig. 7-a, c & e respectively).
The NaeC scaffold showed a heterogeneous distribution of CaP where some areas contained more concentrated CaP, which was aggregated creating some large CaP clusters, whilst other regions were empty of CaP particles ( Fig. 7-b). In contrast, the Plas-C scaffolds indicated an even distribution of CaP coating on the surface of the PCL scaffold struts that were spread throughout the inner and outer of scaffold structure (Fig. 7-d).
The total coated mineral volume was significantly increased in Plas-C scaffolds compared to the NaeC scaffolds (40.21 mm 3 in comparison with 31.54 mm 3 ) (Fig. 7-f).

Discussion
CaP is a biomimetic compound widely used in bone tissue engineered applications [19]. Because of the nucleation potential of phosphate and calcium ions, there is a demand for a firm and uniform coating of them onto the scaffold surface. To achieve this, polycaprolactone, a widely used scaffold material, can be surface activated by alkali, acid or plasma pre-treatment [28,29].
This study examined both NaOH and AreO 2 plasma treatment of PCL which have been shown to markedly improve the hydrophilicity of PCL. While Ar plasma alone may be preferable for surface cleaning compared to O 2 plasma, as it has less impact on the substrate's material properties [30], an AreO 2 plasma mixture has been shown to be more efficient at increasing sample roughness than Ar exclusively [31], which also enhances the adsorption of CaP [32].  However the processes of plasma spraying and etching by NaOH, may significantly alter the material properties of PCL before the coating with CaP. Our SEM data showed that AreO 2 plasma treatment significantly increases the surface roughness of the PCL fibres. This is in agreement with previously reported similar studies which showed the surface roughness of a graphite/polymer composite and a porous PCL scaffold also increased with O 2 plasma modification [33]. Also, the study of Lin et al. showed the rough surface of porous PCL scaffolds due to plasma treatment [17].
Mild alkaline conditions in contrast to acidic treatment has been shown to be more beneficial as a substrate pretreatment process, generating less undesirable by-products [34]. While, a dilute NaOH solution was used in this study, its corrosive nature still resulted in the surface layer of the PCL fibres peeling off to create a rough surface, similar to Luickx et al. who also used 1M NaOH on electrospun PCL scaffolds [35]. However, FDM PCL/graphene 3D printed scaffolds were shown to be resistant to 3 h exposure to 5M NaOH [36]. This suggests that NaOH treatment of electrospun mesh-like structures are more vulnerable to cleavage of carboxyl and hydroxyl chains in PCL polymers compared with the fibres of FDM scaffolds.
Following CaP coating, the SEM data showed the Plas-C scaffolds had a uniform coating density without any cracks or fractures compared to the brittle coated layers seen in the NaeC scaffolds and other reported studies where non-uniform surface activation with NaOH treatment resulted in subsequent uneven CaP deposition [17,33]. This was also confirmed by the m-CT evaluation where large aggregated non-uniform CaP crystals were detected on the surface of NaeC scaffolds compared to the smooth coated Plas-C scaffold.
It is not fully understood what are the main factors i.e. physical (Van der Waals), chemical or mechanical interaction which influences coating adhesion to its substrate [37]. The surface of PCL polymers become super-active by the action of carbonyl (eCOe), carboxyl (eCOOÀ) and hydroxyl (eOH) anions and these negatively charged groups are then ready to attract the soluble positive calcium ions of the SBF solution [38]. The uneven CaP coated layer achieved from NaeC may be due to scaffold pores which were already filled with air and thus not able to take up the aqueous NaOH solution. In contrast, plasma pre-treatment could overcome this limitation by stimulating a homogeneous activation on MEW PCL scaffolds prior to immersion in SBF solution.
Following NaOH and AreO 2 plasma treatment, the samples had a negatively charged surface potential able to interact with the positively charged Ca 2þ ions in the SBF solution. EDS analysis clearly showed the Ca 2þ content was enhanced in both NaeC and Plas-C scaffolds. However, the percentage of Ca 2þ ions was higher in Plas-C scaffolds compared to NaeC scaffolds. The positive charge as a result of accumulation of Ca 2þ ions is then ready to interact to the oppositely charged PO 4 3À ions which was observed in Plas-C group. Poorly crystallized CaP and no phosphorus element however was detected by EDS analyses of NaeC scaffolds. Furthermore, EDS results demonstrated higher concentrations of Na þ (7.0%) in NaeC scaffolds in contrast to Plas-C scaffolds (2.2%).
Rather the most intense peaks of the XRD analysis identified large deposits of halite structures. Our results showed the molar ratio of Na/Cl was 0.50 for NaeC scaffolds and 0.17 for Plas-C groups. The molar ratio of Na/Cl of less than 1 indicated the removal of Na þ ions in both scaffold groups. According to the XRD graphs, the main crystal structures of the Plas-C samples were brushite (CaHPO 4 , 2H 2 0), and hydroxyapatite (HAP). While an equal ratio of calcium and phosphate ions represents the brushite structure, the Ca/P ratio in this study was 3.94 for Plas-C scaffolds, which was close to biphasic combinations of HAP (Ca/P: 1.67) and tetracalcium phosphate (TTCP) (Ca/P: 2.00). The mixture of phases of particles resulted in different morphologies of coated pieces that  influenced the solubility depending on the crystallinity and its size. The changing structures can be associated with the different pH and temperature conditions. Previous studies showed the stability of the crystal particles of coated scaffold can be modified according to the pH and temperature of the implanted site [2]. The brushite structure of the Plas-C scaffolds tends to form to HAP or TTCP, by changing the acidity or basicity of the environment. For example, in a neutral state, TCP can convert to HAP. But under acidic pH, TCP will change to a brushite structure [39]. However, the most stable phase of CaP at neutral pH in the human body is HAP which is the main constituent of bone tissue [40]. Among the different crystal structures, crystals with larger sizes have lower solubility because of the reduction in surface area [41]. Although both Plas-C and NaeC scaffolds contained halite and brushite crystal structures, the smallest crystal size was found for Plas-C scaffolds. Previous studies have shown low crystallinity and a finer crystal size can increase solubility [42] suggesting Plas-C scaffolds might have higher solubility in contrast to NaeC scaffolds. The stability and solubility of the CaP minerals reduce in order from brushite to TTCP and HAP [43]. However, Jang et al. reported higher stability of brushite in an acidic media compared to a neutral environment or alkaline as it transforms into apatitic calcium phosphate (Ap-CaP) [44]. Previous studies demonstrated the advantages of brushite within b-tricalcium phosphate (b-TCP) and monocalcium phosphate monohydrate (MCPM) for dental paste formulations and injectable orthopedics because of the high solubility of brushite. Also, the combination of brushite matrix and b-TCP granule microstructures confirmed rapid bone formation in contrast to HAP cements on the market [45]. In addition, the cubic halite crystals were detected in both Plas-C and NaeC scaffolds but in larger sizes for NaeC scaffolds. However, the higher percentage of brushite crystals enriched in Ca and P elements was confirmed in Plas-C scaffolds by EDS and XRD.
Published studies showed that larger halite crystals which results from lower temperature, reduction of free energy of the solution or the high concentration of solute can lower solubility due to less surface area contact with the solvent [46]. Release of calcium and phosphate minerals decrease with greater crystal size and higher crystallinity [47]. Therefore, there would be higher solubility of the CaP coating for Plas-C scaffolds in comparison to NaeC due to smaller crystal size.
Since the average crystal size has an impressive effect on the mechanical properties of the coated scaffolds, the tensile strength of Plas-C scaffolds showed higher Youngs modulus, elongation break% and tensile strength values between all the treated samples. Previous reports demonstrated a weak compressive strength of brushite in comparison to HAP because of more resorbable properties [47]. Both Plas-C and NaeC scaffolds showed an increase in tensile modulus compared to Plas-nC and Na-nC scaffolds similar to the study of Al-Munajjed et al. where higher mechanical properties with collagen/calcium-phosphate composite scaffolds were noted when compared to the collagen only scaffold [48]. Although Plas-C samples formed brushite crystal structures, their smaller average crystal size created a dense layer of CaP which lead to better overall mechanical properties. This is in agreement with the study of Obayi et al. who reported the mechanical tensile strength of samples increased with decreasing crystal size due to the Hall-Petch relationship [49].

Conclusion
Following O 2 eAr plasma and NaOH surface modification, an apatite mineral layer was precipitated onto the surface of MEW PCL scaffolds by immersing them in simulated body fluid. The study showed that plasma pre-treatment affords a uniform and homogeneous CaP coating with a thin layer of mineral deposition. Although XRD analysis showed that both Plas-C and NaeC scaffolds include brushite and halite structures, the halite structure was found primarily in NaeC scaffolds whereas a mixture of brushite and biphasic combinations of HAP and TTCP were found in Plas-C scaffolds. Furthermore the crystal size was also smaller in Plas-C scaffolds. Mechanical characterization indicated Plas-C scaffolds were stronger compared to the other treated scaffolds, but did lose some integrity compared to the untreated control scaffold. Plas-C scaffolds may have more stability of the CaP minerals due to the higher percentage of brushite and HAP suggesting that the plasma treatment is the most suitable for further development of MEW fibres for bone regeneration applications.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.