Physical and biological properties of electrospun poly(d,l‐lactide)/nanoclay and poly(d,l‐lactide)/nanosilica nanofibrous scaffold for bone tissue engineering

Abstract Electrospun scaffolds exhibiting high physical performances with the ability to support cell attachment and proliferation are attracting more and more scientific interest for tissue engineering applications. The inclusion of inorganic nanoparticles such as nanosilica and nanoclay into electrospun biopolymeric matrices can meet these challenging requirements. The silica and clay incorporation into polymeric nanofibers has been reported to enhance and improve the mechanical properties as well as the osteogenic properties of the scaffolds. In this work, for the first time, the physical and biological properties of polylactic acid (PLA) electrospun mats filled with different concentrations of nanosilica and nanoclay were evaluated and compared. The inclusion of the particles was evaluated through morphological investigations and Fourier transform infrared spectroscopy. The morphology of nanofibers was differently affected by the amount and kind of fillers and it was correlated to the viscosity of the polymeric suspensions. The wettability of the scaffolds, evaluated through wet contact angle measurements, slightly increased for both the nanocomposites. The crystallinity of the systems was investigated by differential scanning calorimetry highlighting the nucleating action of both nanosilica and nanoclay on PLA. Scaffolds were mechanically characterized with tensile tests to evaluate the reinforcing action of the fillers. Finally, cell culture assays with pre‐osteoblastic cells were conducted on a selected composite scaffold in order to compare the cell proliferation and morphology with that of neat PLA scaffolds. Based on the results, we can convince that nanosilica and nanoclay can be both considered great potential fillers for electrospun systems engineered for bone tissue regeneration.


| INTRODUCTION
Over the recent years, biopolymers find a wide range of advanced applications that include biodegradable food packaging, 1,2 bioremediation, 3 biosensing, 4 controlled drug release, 5 and tissue engineering. 6 In this context, there is a growing interest in functional biopolymeric porous scaffolds exhibiting high mechanical properties with the ability to support cell attachment and proliferation. [7][8][9] Electrospinning is becoming more and more attractive for its simplicity and flexibility if compared with the other scaffold fabrication techniques. 10,11 In fact, the electrospinning technique permits easy control of the fiber diameter, porosity, and mechanical properties of the final scaffolds by changing the processing parameters and the materials used. [12][13][14] Electrospun membranes are often engineered to be involved in advanced applications including catalysis, 15,16 controlled drug release, 17,18 bioprocess intensification, 12,19 biosensing, 20 food packaging, 21 but a still-open challenge is the preparation of mats with adequate properties for tissue engineering purposes. 22 Beyond its versatility in material selection, which can include both natural and synthetic polymer, electrospinning also provides the possibility to include nanoparticles into the polymeric fibers. 23 By considering the unique properties related to nanometric size and high specific surface, the incorporation of functional nanoparticles into an electrospun polymer matrix can provide substantial properties enhancements, even at low nanoparticles content. 24,25 This feature can be used to modify specific properties of nanofibrous polymers by increasing their mechanical strength, 26 their bioactivity, 27 and/or endowing them with additional features including electrical conductivity. 28 The chemical composition and the particle dimension strongly affect the matrix/particles interaction and, as a consequence, the final properties of polymer-based composites. 29,30 For this reason, several inorganic nano-sized particles, such as metals, carbon-based materials, silica, and clays have been extensively studied in the field of polymer-based nanocomposites in order to enhance their performances for specific applications such as environmental remediation, electromagnetic interference shielding, sensing, 31 supercapacitors, 32 packaging, automotive, and solar energy fields. 25 In tissue engineering applications, various types of inorganic nanofiller such as carbon nanotubes, 33 graphene, 26 and hydroxyapatite 27 have been used to produce polymer/inorganic biocomposite fibers, leading to high-performance nanofibrous mats. In this context, nanosilica and nanoclay are one of the most investigated nanofillers for polymer matrix due to several interesting properties including low thermal conductivity, chemically inertness, and non-toxicity 34-45 that can be exploited for tissue engineering purposes and other biomedical applications. [46][47][48][49][50][51][52][53][54][55] Although several research articles report the physical characterization of non-bioresorbable polymer/silica and polymer/clay composite electrospun fibers, 36,37,[39][40][41][42][43][44][45]56 relatively a few articles deal with biopolymers-based electrospun nanocomposites [46][47][48][49][50][51]57 and even less with their interaction with living cells for tissue engineering applications. [52][53][54][55]58 Singh et al. developed multifunctional mesoporous silica-shelled polycaprolactone hybrid nanofiber scaffolds for bone regeneration exhibiting excellent mechanical functionality. 52 Mehrasa et al. prepared aligned nanofibrous composites made of poly(lactide-co-glycolide)/gelatin/mesoporous silica by electrospinning for nerve tissue engineering applications. 53 Koosha et al. produced chitosan/polyvinyl alcohol/montmorillonite (MMT) nanofibrous scaffolds highlighting that the mechanical properties of the nanocomposite were highly improved by the addition of only 1 and 3 wt% of MMT. Furthermore, these nanocomposite scaffolds were found to be biocompatible showing no adverse cytotoxic effect on human fibroblast cells. 54 Gaharwar et al. fabricated nanoclayenriched poly(ε-caprolactone) electrospun scaffolds for osteogenic differentiation of human mesenchymal stem cells (hMSCs) able to enhance the attachment, proliferation, and differentiation of hMSCs if compared to the neat PCL electrospun scaffolds. 55 To the best of our knowledge, there is not any available work focusing on the direct comparison of polylactic acid (PLA)-based electrospun scaffold containing either silica (AS) or clay (CLO) nanoparticles for potential tissue engineering purposes. PLA is a commonly used material for tissue engineering applications due, in part, to its ability to degrade into the naturally occurring lactic acid under physiological conditions. Other exceptional features are the low immunogenicity and the interesting mechanical properties. 59,60 However, PLA has some drawbacks such as biological inertness and low cell adhesion that need to be addressed. 61 Therefore, this work aim is to compare the physical properties and the in vitro cytotoxicity and cell attachment of pre-osteoblastic cells on PLA/AS and PLA/CLO electrospun scaffolds. The inclusion of the nanoparticles into the electrospun mats was investigated with Fourier transform infrared spectroscopy in attenuated total reflectance (FTIR-ATR). The rheological measurements conducted on the polymeric solution/suspensions were correlated with the fiber morphology that was analyzed through scanning electron microscopy (SEM) and image processing. Uniaxial tensile tests were carried out in order to evaluate the reinforcing effect of both AS and CLO for three different compositions (1, 3, and 5 wt%). Differential scanning calorimetry (DSC) permitted the investigation of the crystallinity of the composites. Finally, pre-osteoblastic cells were seeded on the composite scaffolds in order to compare their vitality and morphology on the PLA/AS and PLA/CLO bionanocomposites with that of neat PLA scaffolds.

| Materials
PLA (2002D, NatureWorks) was used in this work. Acetone (Ac) and chloroforms (TCM) were purchased from Sigma-Aldrich. The commercial nanosilica is Aerosil R812 fumed silica supplied by Evonik Industries AG (Evonik Degussa) with a declared specific surface of 230-290 m 2 /g, modified with hexamethyldisilazane. The commercial clay is Cloisite 30B, supplied by Southern Clay Products. The clay is a MMT modified by 90 meq/100 g of bis(2-hydroxyethyl)methyl tallow alkylammonium cations. Before processing, the polymer and the fillers were dried overnight at 90 C under vacuum in order to avoid PLA hydrolytic degradation during processing. 62 All the reactants were ACS grade (purity >99%).

| Electrospinning processing
PLA nanofibers, PLA/AS, and PLA/CLO electrospun composites preparation followed a fabrication route similar to those described in a previous work. 27 In brief, PLA (10 wt%) was dissolved in TCM:Ac (2:1 vol) at room temperature under continuous magnetic stirring overnight. PLA/AS and PLA/CLO suspensions were prepared by adding AS or CLO particles to the solvent system that was then subjected to ultrasonication for a total of 1 hr, and finally adding the PLA with a weight concentration of 10 wt % with respect to TCM:Ac mixture. AS and CLO were added in the solvent system in order to achieve 1, 3, and 5 wt% with respect to the polymer phase, according to scientific literature. 54 The PLA, PLA/CLO, and PLA/AS scaffolds were prepared by using semi-industrial electrospinning equipment (NF-103, MECC CO., LTD., Japan) equipped with a cylindrical grounded rotary drum (diameter = 10 cm). The polymeric solution/suspensions were filled to a 5 ml syringe fitted with 19-gauge stainless steel. The following constant parameters were set: flow rate, 1 ml/hr; needle tip-collector distance, 13 cm; high voltage, 15 kV; temperature, 25 C, collector angular speed, 10 rpm; processing time, 120 min; According to the above-mentioned parameters, approximately 70 μm thick membranes were obtained. In order to remove any residual solvents, the collected scaffolds were dried for 48 hr under fume hood.

| Morphological analysis
The morphology of the scaffolds was evaluated by scanning electron microscopy, (Phenom ProX, Phenom-World) and by transmission electron microscopy by using the Versa 3D Dual Beam Scanning Electronic Microscope (Thermo Fisher, FEI) equipped with a retractable STEM detector. For SEM analysis, circular samples (diameter = 10 mm) were attached by using adhesive carbon tape on the aluminum stub. Before the analysis, the samples were sputter-coated with gold for 60 s under argon atmosphere by using a Sputtering Scancoat Six (Edwards) in order to avoid electrostatic discharge during the test. 63 The SEM was set with an accelerated voltage equal to 10 kV. The samples for TEM investigation were prepared by the direct deposition of the electrospun nanofibers onto the carbon-coated copper grid. The samples were analyzed by using an accelerated voltage equal to 30 kV.

| CLO and AS specific surface
Surface area measurements were performed by using an autosorb iQ-MP/XR (Quantachrome) instrument. Before measurement, each sample was outgassed under vacuum at 120 C for 3 hr. The surface area was determined by physical adsorption of N 2 at the liquid nitrogen temperature, using the Brunauer-Emmett-Teller (BET) equation. 64

| Particles size and fiber diameter distributions
A dedicated image processing software was used to investigate the fiber diameter and the particle size distribution of the electrospun mats. ImageJ on SEM images of AS and CLO particles was used to determine the particle size distribution while a plugin for ImageJ (DiameterJ) was used to investigate the fiber diameter distribution. 65 2.7 | Differential scanning calorimetry DSC (Setaram, model DSC131) was used to investigate the calorimetric properties of the scaffolds. The analysis was carried out with two cycles of heating from room temperature to 190 C at 10 C/min heating rate under nitrogen flow on electrospun samples with approximately the same weight (~5 mg) sealed in aluminum pans.
PLA and PLA-based composites crystallinity degree (χ) were calculated according to the following equation 66 : where ΔH cc and ΔH m are the cold crystallization and melting enthalpy of the samples, respectively. X PLA is the weight fraction of PLA and ΔH 0 m is the melting enthalpy of 100% crystalline PLA equal to 93.7 J/g. 66

| Polymeric solutions/suspensions complex viscosity
A plate-plate rotational rheometer Mars (Thermofisher Rheological) with 25 mm parallel-plate geometry at 25 C was used to perform the characterization of the polymeric solutions/suspensions complex viscosity. Oscillatory frequency sweep tests were performed at a constant stress of 1 Pa with an increase of angular frequency from 1 to 100 rad/s.

| Water contact angle measurements
FTA 1000 (First Ten Ångstroms, UK) instrument was used to perform the static contact angles measured by using distilled water (DW) as fluids. In particular, a droplet of DW (~4 μl) was dropped on the scaffold and the images were taken after 10 s from the DW deposition. At least seven spots of each composite nanofiber mat were tested and the average value was taken.

| Mechanical properties
A laboratory dynamometer (Instron model 3365) equipped with a 1 kN load cell was used to perform the tensile mechanical measurements on rectangular-shaped specimens (10 × 90 mm).
The electrospun mats were cut off from along the radial direction of the cylindrical collector. Due to the high elongation of the samples, the tensile tests were carried by using a double crosshead speed: 1 mm/min up to 10% strain and 50 mm/min until fracture.
The initial length of the samples was 20 mm while the thickness of each sample was measured before the test. From the nominal stress-strain curves, the following mechanical parameters were obtained: elastic modulus (E), tensile strength (TS), and deformation at break (ε b ). Seven samples were tested for each material and the average values of the mechanical parameters were reported with their standard deviations.
The scaffolds were then sterilized under UV for 2 hr and soaked for 12 hr in a complete culture medium. Twenty microliters of cellular suspension were inoculated onto each scaffold in order to reach a seeding concentration of 5 × 10 4 cells/cm 2 . After 90 min of incubation at 37 C and 5% CO 2 (to promote cell adhesion), each scaffold was transferred into a medium-filled well of a 24 multiwell plate.

| Cell viability assay
AlamarBlue™ Cell Viability Reagent (Invitrogen) was used to evaluate cell proliferation. The samples were transferred into clean wells and F I G U R E 1 Scanning electron microscopy (SEM) micrographs of (a) clay (CLO) and (b) silica (AS) particles. Particle size distribution of (c) CLO and (d) AS particles. Characterizations of both particles were carried out after 1 hr sonication in TCM:Ac (2:1 vol) each scaffold was incubated at 37 C and 5% CO 2 for 3 hr with 500 μl of an Alamar Blue reagent (10×) diluted (1:10) in Medium.
The fluorescence values were read on a plate reader; excitation wavelength was 530/25 (peak excitation is 570 nm) whereas emission wavelength was 590/35 (peak of emission is 585 nm). The number of living cells is directly proportional to the fluorescence value. The assays were carried out at 0, 4, 7, and 11 days of culture in triplicate for each time. Scaffolds without cells were used as blank for each measurement.

| Statistical analysis
Statistical analyses of the data were performed through one-way analysis of variance, and when applicable, data were compared using the Student's t test. p-value <.05 was considered statistically significant. The micrograph of PLA fibers shows the typical morphology of an electrospun material with smooth and randomly oriented fibers in the nanoscale (PLA mean diameter = 1.07 ± 0.16 μm). PLA fibers displayed a rather homogeneous diameter as confirmed by the narrow peak of the fiber diameter distribution. All PLA/AS composites produced in this work exhibited higher mean fiber diameter and more spread peak of the fiber diameter distribution than neat electrospun PLA. Specifically, the mean fiber diameter of PLA/AS 1% and 3% were approximately 10% higher than PLA fibers while PLA/AS 5% mean fiber diameter was even 56% higher than PLA. SEM images clearly reveal that PLA/AS fibers diameter were not homogeneous but character- On the other hand, all PLA/CLO fibers showed a lower mean fiber diameter than neat electrospun PLA as already observed for different electrospun matrices filled with clays. 54,57 More in detail, PLA/CLO 1, 3, and 5% mean fiber diameters were approximately 2, 11, and 18% lower than PLA fibers, respectively. Furthermore, the fiber diameter distribution peaks of PLA/CLO systems were narrower than that of PLA fibers, as quantitatively described by the lower standard deviation evaluated from the graphs shown in Figure 2. CLO particles are brighter than PLA thus easily visible also when they are embedded in the fibers. Interestingly, only some CLO aggregates around 5 μm are visible in PLA/CLO 5% that is much lower than that observed for the sonicated CLO particles. About the filler dispersion, from SEM images it can be noticed that PLA/AS nanocomposites displayed a higher concentration of particles in the bead-like regions of the fibers, in particular at the higher AS concentrations.

| Wettability
The surface wettability of the bionanocomposites electrospun mats was analyzed to evaluate the hydrophilic/hydrophobic character of the scaffolds through water contact angle (WCA) measurements  Table 1.
Electrospun PLA showed the T g -related endothermic peak during the first and second heating scan at 61.5 and 59.6 C, respectively.
The T g of PLA seemed to be slightly affected by the presence of both Results in Table 1 highlighted that during the first heating scan, the crystallinity of electrospun PLA was 9.41% and it remained almost constant for the composites containing 1 and  All PLA/AS composites were found to be more brittle than PLA as highlighted by the strong reduction of the elongation at break. On the other hand, PLA/CLO 1%, and partially also PLA/CLO 3%, mats showed  In fact, E value of PLA/AS 1, 3, and 5% were 118, 278, and 438% higher than that of electrospun PLA, respectively, while the elastic modulus of PLA/CLO 1, 3, and 5% were 222, 438, and 653% higher than that of PLA, respectively.
As qualitatively observed in the stress-strain curves, the ten-

| Cell culture
The

| DISCUSSION
In this work, it was assessed the physical and biological properties of PLA electrospun bionanocomposites filled with two different commercial nanofillers that is, nanosilica (Aerosil ® R 812) and nanoclay (Cloisite ® 30B). Aerosil R812 has a silicon dioxide content of over 99.8% thus it is an excellent candidate for application in tissue engineering due to its chemical inertness that has no known health implications. 74 Silica and its derivatives were introduced as bone substitutes 75 demonstrating clinical success rates in terms of promotion of new vital bone and as a bio-mimetic coating for implant surfaces. 52,76 In this context, silica nanoparticles can form a tighter interface with the polymer matrix in composites due to their large specific surface area then this filler can not only endow polymer scaffolds with biomineralization capability but also increases the stiffness of polymer materials. 77 On the other hand, Cloisite 30B, an organically modified montmorillonite, is one of the most used nanoparticles because of its capacity to improve the mechanical and thermal properties of the polymer matrix. 78 Furthermore, the quaternary ammonium moiety present in Cloisite 30B has been shown to have antimicrobial properties. 79 SEM and BET analysis conducted on the fillers highlighted that AS particles are smaller than CLO. This feature can affect several physical and biological properties of the nanocomposite mats such as fiber diameter, roughness, and, as a consequence, its biocompatibility.
The FTIR-ATR spectra confirmed the successful inclusion of the fillers in the PLA nanofibers. In fact, PLA/AS and PLA/CLO systems showed the characteristic peaks of both the polymer matrix and the respective nanoparticles. Also, the presence of peaks at 2,925 and 2,853 cm −1 due to alkylammonium ions in Cloisite 30B in PLA/CLO nanofibrous composites reveals that the organo-modifier in Cloisite 30B was still present after processing. The alkylammonium ions preservation after 1 hr sonication of CLO in TCM:Ac (2:1 vol) can be likely ascribed to the packed morphology of these particles. In fact, the organo-modifier is intercalated among the MMT layers that, as clearly visible from SEM and STEM images, maintained their stuck structure also after the electrospinning process thus preserving the salt from dispersion in the solvent system. 70 The fiber morphology and diameter of PLA-based electrospun scaffolds were affected by both the filler weight concentration and type, and both features are furtherly related to the solution viscosity.
More in detail, the PLA/AS fiber diameter was found to be higher than that of neat electrospun PLA. These results can be likely explained by considering that the electrospinning process is based on the uniaxial stretching of a charged droplet of polymeric solution. The stretching degree can be strongly affected by the polymeric solution/suspension viscosity since the higher the shear viscosity, the higher the resistance to flow. For this reason, the electrospun fiber diameter increase upon increasing the solution viscosity, as already observed in other works. 26,[80][81][82] It is well known that solid particles can increase the polymeric suspensions complex viscosity except for nanoparticles presenting pro-degradative effects or acting as plasticizers. [83][84][85] In this work, the introduction of AS induced an increase of the complex viscosity in the polymer solution that can explain the increase of the electrospun fiber diameter.
The spread diameter size distribution of the PLA/AS system can be related to the presence of AS agglomerates observed by STEM analysis but also to the increased viscosity, which was found to be more accentuated at the low frequencies. The WCA measurements revealed that both the nanofillers cause a slight increase of the scaffolds wettability and that it increased upon increasing the filler loading. In order to explain these results, it may be taken into account that the wettability performance of nanocomposites is strongly dependent on the surface topographical properties but also on the chemical properties of the filler. 86 Although the increased diameter of the PLA/AS fibers can induce an increase of the WCA by reducing the specific surface of the mat, the slight increase in hydrophilicity observed for the PLA/AS systems can be ascribed to the presence of silicon dioxide on the nanosilica surface. 87 On the other hand, the more pronounced WCA reduction observed for PLA/CLO systems can be likely ascribed to the thinner fiber diameter observed by SEM and to the presence of the organo-modified clays at the surface of the fibers, as highlighted by FTIR-ATR measurements. 88,89 Electrospun PLA usually exhibits much lower crystallinity than that of PLA in pellets because of the high speed of solidification during the process. 90 Figure 9a,b and the elongation data in Figure 9e.  On the other hand, the slight decrease in cell growth observed on PLA/AS 5% starting from Day 4 can be likely ascribed to potential metabolic activity changes induced by AS particles. In fact, in other electrospun nanocomposite filled with nanosilica, it was observed that the higher the filler concentration the higher the number of nanoparticles that can adhere to the cell surface and then be internalized thus restricting the cellular functionality. 94 Moreover, high concentration of silicate nanoparticles can also interact with the media proteins and results in the formation of aggregates that cannot be engulfed by the cells. This might also contribute to decreasing the growth trend at higher silicate concentrations. 94 The effect of electrospun polycaprolactone filled by clay nanoparticles on protein absorption was investigated by Gaharwar et al. 55 Their results highlighted the low protein adsorption induced by these particles. Furthermore, the relatively high dimension of CLO particles reduces the possibility to be internalized by cells. Therefore, the cell growth plateau observed in PLA/CLO 5% scaffolds after 4 days could be related to the release of the alkylammonium salt intercalated into the clays into the medium that can thus affect the cell growth. In fact, it is well known that the alkyl ammonium surfactant present in Cloisite 30B can be released in a liquid medium also when incorporated in polymer matrices and it is primarily responsible for its microbicidal properties. 38 Coherently with viability assay results, SEM micrographs carried out on the electrospun mats after 4 days of culture revealed that upon increasing both AS or CLO particles filled into the polymer matrix, the number and size of clusters of aggregated cells increased.
The only system likely able to exhibit a biofilm-like structure after 11 days of culture is PLA/CLO 5%. Also, the spread morphology of the cells cultured on the composites mats revealed that both AS and CLO particles were likely able to enhance the adhesion of the preosteoblastic cells on the electrospun systems.
These results can be presumably ascribed to the enhanced elastic modulus and wettability of the electrospun mats filled with AS and CLO particles. These data suggest that both AS and CLO nanoparticles embedded in PLA electrospun matrix provided a favorable cell proliferation environment that can be ascribed to the increased cell affinity for the substrate. Further tests will be carried out in order to better understand the influence of these particles on cell proliferation and differentiation and will be reported in a separate paper since a detailed analysis goes far beyond the scope of the present paper.

| CONCLUSIONS
In this work, the physical and biological properties of electrospun PLA, PLA/AS, and PLA/CLO bionanocomposites were evaluated.
The successful inclusion of both AS and CLO particles in PLA fibers was confirmed by SEM images and FTIR-ATR that demonstrated also the preservation of the Cloisite 30B organo-modifier.
The AS loading led to an increase of the mean fiber diameter and a less homogenous fiber size distribution of PLA/AS systems probably due to the high viscosity of the processing suspensions observed at the low frequencies. On the contrary, CLO induced a decrease of the mean fiber diameter and a more homogenous fiber size distribution likely ascribed to the presence of the quaternary ammonium salt in the MMT particles. The organo-modifier of Cloisite 30B was also able to reduce the WCA of the PLA/CLO electrospun composites.
DSC analysis revealed that both AS and CLO nanoparticles increased the crystallinity of PLA in both the heating scans performed acting as a nucleating agent for the polymer matrix.
CLO reinforcing effect was higher than that of AS particles. In fact, PLA/AS scaffolds highlighted a more brittle behavior than electrospun PLA or PLA/CLO nanocomposites, probably because of the bead-like structures formed in the fibers that acted as a local defect. The more intense increment of the elastic modulus due to

CONFLICT OF INTEREST
The authors declare no potential conflict of interest. Carrubba.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.