Strong and osteoconductive poly(lactic acid) biocomposites by high-shear liquid dispersion of hydroxyapatite nanowhiskers

Abstract Controlled incorporation of bioactive hydroxyapatite (HA) into poly(lactic acid) (PLA) signifies a promising approach to design and development of biomedical-adaptive materials. Here we unravel a microwave-assisted biomineralization approach to synthesis of HA nanowhiskers (HANWs), which were characterized by well-controlled diameter (∼30 nm) and length (80 − 120 nm), combined with a desirable calcium − phosphorus ratio (Ca/P) of 1.67. A high-shear liquid dispersion (HSLD) method that provided a combination of high pressure (up to 50 kPa) and high shear rate approaching 10000 s −1 was established to fabricate homogeneous PLA/HANWs composites. In particular, upon incorporation of 30 wt % HANWs the tensile strength and elastic modulus of PLA-HA30 (76.7 MPa and 3.3 GPa) were elevated by 48% and 84% compared to those of pure PLA, respectively, as accompanied by a nearly 2-fold increase in the cell viability. This work paves a facile yet effective roadway to strong and osteoconductive PLA composites appealing for bone tissue repairing. Graphical Abstract Synopsis: Biomineralization concepts are abstracted to synthesize well-defined and bioactive HANWs, which could be uniformly dispersed in PLA biofilms by high-shear liquid exfoliation.


Introduction
During the past decades, biodegradable poly(lactic acid) (PLA) has been recognized as an representative biomedical polymers in terms of good in vivo absorbability, high mechanical strength and ease of processing [1,2]. It stimulates extensive research into PLA-based biomedical-adaptive candidates ranging from tissue engineering to drug delivery devices [3,4]. This perspective has been well elaborated by recent progress in PLA-enabled improvements of biological features that are unattainable with traditional polymer or metal-based implants [5,6]. As a good example, the osteoblastic cells were well attached and spread on nanofibrous PLA, showing great promise as a bone cell supporting matrix in the area of tissue regeneration [7].
Despite the contributions demonstrating biomedical application promise of PLA, it is pertinent to point out that the currently available PLA resins are generally characterized by poor hydrophilicity that prejudices cell attachment or adhesion, as well as relatively low mechanical compatibility with surrounding tissue [8]. Effective approaches of promoting the hydrophilicity and mechanical properties are mainly associated with incorporation of functional and bioactive fillers such as bioactive glass, chitosan, gelatin, graphene and hydroxyapatite (HA) [9][10][11][12][13][14]. Among the existing filler candidates, HA nanoparticles, featuring excellent osteoconductivity and bone-bonding properties due to the chemical and biological similarity to the mineral phase of native bone, have been widely used to enhance the mechanical properties and biological characters of PLA composites [15,16]. By incorporation of nanoscale HA spindles with a diameter of 10 À 30 nm and a length of 60 À 100 nm, Yan et al. reported that the biodegradability and bioactivity of PLA composites were significantly improved during the evaluation in simulated body fluid (SBF) solutions [17]. Within this context, HA-filled PLA composites have been increasingly developed for their efficacy and versatility at overcoming bottlenecks not readily surmounted by the PLA matrix.
A challenging aspect of engendering HA-modified PLA nanocomposites is to sufficiently exfoliate and properly disperse the nanoscale HA individuals due to the intrinsically high surface energy [18]. To meet the essential criteria for uniform dispersion in PLA, sonocoating and electrospraying of HA suspensions were used by Higuchi et al., which may show distinct limitations related to high time consumption and complex processing conditions [19,20]. Here, we report an unprecedented approach to effective exfoliation of HA nanowhiskers (HANWs) using a high-shear liquid dispersion (HSLD) method. In particular, the microwaveassisted biomineralization in SBF solution was used to synthesize HANWs with surfactant molecules serving as the "soft template". Moreover, the as-synthesized HANWs were subjected to an extremely strong shear flow (shear rate up to 10000 s À1 ) in PLA solutions, enabled by a cyclic shear homogenizer. Once the HANWs are homogeneously incorporated into PLA composites, a combination of mechanical robustness and cytocompatibility is highly anticipated, holding great promise in the biomedical field (e.g. fast bone tissue regeneration). The materials design principles are of interest to engender dental and orthopedic implant formulations with sufficient mechanical strength and improved osteointegration performances.

Materials
A nonionic surfactant in laboratory grade (Triton X-100, CAS No. 9002-93-1), that is tert-Octylphenoxy poly(oxyethylene)ethanol, having a molecular formula of t-Oct-C 6 H 4 -(OCH 2 CH 2 ) n OH (n ¼ 9 À 10) was purchased from Sigma-Aldrich under the product name of TX-100. Deionized water, dichloromethane, ethanol, Ca(NO 3 ) 2 · 4H 2 O, NH 4 H 2 PO 4 , NH 3 · H 2 O and other chemical reagents in analytical grade were supplied by Macklin Inc., and were used as received. Taking into considerations the commercial availability and excellent biocompatibility, poly(lactic acid) (PLA) comprising $2% D -LA with a weight-average molecular weight of 2.23 Â 10 5 g/mol and polydispersity index (PDI) of 2.1 was procured from NatureWorks under the trade name of 4032 D.

Preparation of hydroxyapatite nanowhiskers (HANWs)
The synthetic method of HANWs is schematically depicted in Figure 1. A microwave-assisted biomineralization processing method was used to synthesize HANWs (Figure 1(a,b)). Specifically, 0.1 g TX-100, 0.25 mol Ca(NO 3 ) 2 · 4H 2 O and 0.15 mol NH 4 H 2 PO 4 were successively dispersed in 100 mL water under mechanical stirring for 0.5 h. The pH value was adjusted to 12 by adding NH 3 ·H 2 O. The mixed solution was subjected to 15-min microwave irradiation with an output power of 600 W (ATPIO-6TF, Xianou Co. Ltd., China), allowing the nucleation of HA with TX-100 molecules serving as the "soft template", followed by stepwise crystal growth into HANWs. The resultant product was washed with water three times to completely remove the surfactant and unreacted compounds, followed drying in an oven at 80 C for 12 h.

Preparation of HANW-Modified PLA composites
As depicted in Figure 1(c-i), a high-shear liquid dispersion (HSLD) method was developed to incorporate the as-prepared HANWs into PLA matrix. In specific, PLA was dissolved in dichloromethane (0.1 g/mL) and forced to cyclic flow in a liquid cyclic shear homogenizer (Model GXR, Zhitongyuanfeng Co., Ltd., China) operating at a rotation speed of 10000 rpm with a flux of 5 L/min. Using the finite element analysis method ( Figure  1(f-h)), the homogenizer was stimulated to trigger a combination of high pressure (up to 50 kPa) and high shear rate approaching 10000 s À1 . Once the steady flow of PLA solution was obtained, the HAWN powder was gradually added into the shearing solution with the preset filler loadings (i.e. 10 and 20 wt %). After cyclic shearing for 30 min, a homogeneous solution comprising PLA and HAWNs could be obtained (Figure 1(i)), followed by solution casting and vacuum drying at 40 C for 12 h and 80 C for 6 h. This allowed formation of PLA composite films loaded 10, 20 and 30 wt % of HANWs with a thickness of $100 mm (Figure 1(j)), which were denoted as PLA/HA10, PLA/HA20 and PLA/HA30, respectively. Pure PLA films without addition of HANWs were prepared by the same HSLD processing method to make a control sample.

Scanning electronic microscopy (SEM) observation
The morphological features of HANWs, HANWmodified PLA composites and fracture surfaces after tensile failure were directly imaged on an SE-4800 SEM (Hitachi, Japan) at an accelerated voltage of 10 kV. Prior to the SEM observations, the samples were dried in a vacuum oven at 40 C for 8 h, followed by sputter coating with gold.

Transmission electron microscopy (TEM) observation
TEM and high-resolution TEM (HR-TEM) observations were conducted on an HT7700 TEM (Hitachi, Japan) at an accelerated voltage of 100 kV. Prior to TEM imaging, HANWs dispersed in ethanol were dropped onto the carbon film 400 mesh copper TEM grids (Ted Pella, Inc.) and allowed to dry in ambient conditions.

Energy dispersive X-ray spectrometry (EDS) microanalysis
The elemental composition for HANWs was directly mapped on an EDS detector (X-Max N , Oxford Instruments, UK) linked to the SE-4800 SEM. The powder sample was directly scanned without coating treatment, and the operating voltage for EDS mapping was 15 kV.

Fourier transform infrared spectroscopy (FTIR)
The FTIR spectra were recorded on a PerkinElmer Spectrum 2000 spectrometer (PerkinElmer Instruments, Inc.) with 16 scans at a resolution of 4 cm À1 .

Mechanical testing
The tensile properties of PLA/HANW composite films were evaluated at 23 C on an Instron 5944 universal test instrument, equipped with a load cell of 500 N (ASTM standard D638-14). The gauge length and crosshead speed were set at 20 mm and 5 mm/min, respectively.

Contact angle (CA) testing
Water contact angle measurements were conducted on a JY-82A goniometer (Dingsheng Instrument, Chengde, China). Deionized water drops were used as the probing liquid on the film surfaces at room temperature, and five replicates were performed to obtain the average results.

In vitro cytocompatibility evaluation
MG-63 cells (American Type Culture Collection) were cultured in DMEM medium supplemented with 10% fetal bovine serum (FBS). Cells were seeded into the 12-well plates containing the composite film samples (10 mm Â 10 mm). To image the cell morphology, the cells were fixed with 4% paraformaldehyde (PFA) in PBS for 15 min at 23 C before which the cells were permeabilized with 0.5% Triton X-100 made in PBS solution for 15 min at 23 C. 500 mL of 100 nM rhodamine phalloidin (Cytoskeleton) was then added onto the cells, and was incubated at 23 C in the dark for 0.5 h. The cells were washed in PBS for three times and counterstained DNA with 500 mL DAPI (1 lg/mL) for 5 min. After being rinsed in PBS, the films were inverted on a drop of antifade mounting media on a glass slide. Subsequently, all the samples were imaged on an Olympus FV-1000 confocal microscope.
To quantitatively measure the cell viability by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) colorimetric assay, the tested films were transferred to the new 12-well plates after incubation of 6, 12 and 24 h. 500 mL MTT (500lg/mL) was added into each well for 4-hour incubation. After removing the supernate, 500 mL dimethyl sulfoxide (DMSO) was added into each well to dissolve the formazan crystals for 10 min at 23 C. 100 mL of each sample was transferred to a 96-well plate and the OD values at 490 nm was measured in a plate reader (Thermo Fisher Scientific, Inc.). At least five replicates were used for each sample, and the mean values with standard deviations were reported.

Statistical analysis
All numerical results were presented as the mean value ± standard deviations for at least three or five samples. One-way analysis of variance was used to implement statistical analysis. The value of p < 0.05 was identified as statistically significant.

Results and discussion
We first examined the hypothesis on the use of microwave-assisted biomineralization approach toward effective synthesis of HANWs, as appraised from the combination of morphological features and structural basis (Figure 2). It is apparent from direct SEM observations that rod-like nanosized whiskers were formed in the appearance of dense accumulation and local aggregation (Figure 2(a,b)), due presumably to high surface activity arising from the low dimensions [21]. Moreover, the high surface roughness of HANWs were of important significance, which favor positive cell responses by stimulating protein interaction and subsequent cell adhesion [16,22]. Figure 2(c) offers evidence for high structural uniformity of the HANWs, mainly featuring a diameter of approximately 30 nm and a length of 80 À 120 nm. It is worth noting that the calcium À phosphorus ratio (Ca/P) of HANWs was determined to be 1.67 (Figure 2(d)), on a par with that of apatite entities formed in naturally occurring bones [17]. This suggests that the as-synthesized HANWs fall into the category of highly desired HA structure having the highest osteoconductive and osteoinductive capability to regenerate bone tissue [23,24]. This was in line with the FTIR spectrum (Figure 2(e)), showing the prominent absorption peaks at 1089, 1019 and 962 cm À1 attributed to the stretching and bending of phosphate [20,25,26].
Based on the above results, the biomineralization approach to creation of well-organized HANWs was tentatively verified. We further characterized and imaged the crystalline structure in the HANWs. As verified by the XRD intensity profile, Figure 2 [27,28]. The crystalline nanodomains were directly imaged by HR-TEM (Figure 2(g)), in which the lattice spacing was unraveled to be approximately 0.34 nm assigned to the (002) plane of HA hexagonal structure [29]. It implies that upon microwave-assisted biomineralization proceeding the HANWs preferentially grew along the c-axis direction of HA, with the surrounding TX-100 molecules serving as the "soft template" that dynamically varied the conformations [30]. This mechanism is likely shared by the scenario of cetyltrimethylammonium bromide (CTAB)-induced formation of HA nanorods, in which the CTAB may assemble into rodlike structures under specific hydrothermal synthesis [29]. Using polyvinylpyrrolidone (PVP) as a surfaceregulating agent, Chu et al. have demonstrated that both the nucleation and crystal growth of HA could be well controlled to generate one-dimensional nanorods with tunable diameter and length [31]. Given the structural features in terms of high crystallinity and rich oxygen functional groups, the HANWs consisting of well-organized nanocrystals were hypothesized to positively stimulate osteoblastic proliferation [15,32], holding great promise as bioengineering materials to promote the osteointegration of dental/orthopaedic implants [33].
Despite tremendous promise of HA-based nanocomposites in the application of bone tissue repairing [34], it remains a great challenge to properly exfoliate and uniformly disperse the nanoscale HA particles in the polymeric matrices, especially during large-scale processing routes. In the proposed HSLD approach, the HANWs were hypothesized to be sufficiently sheared, oriented and exfoliated in PLA solution under the combined conditions of high pressure (up to 50 kPa) and high shear rate (up to 10000 s À1 ), as provided by the cyclic shear homogenizer ( Figure 1(f-h)). This hypothesis was examined by direct SEM observations of cryogenically fractured composites showing the dispersion morphology of HANWs in PLA matrix ( Figure 3). It is apparent that proper dispersion of HANWs were achieved for all composites, regardless of filler concentration. Though at the highest filler loading of 30 wt %, individual HANW entities featuring high exfoliation degree and homogeneous dispersion were observed for PLA-HA30 without evident agglomeration. More importantly, it is worth stressing that the HANWs were intimately embedded into the PLA matrix, primarily attributed to the enhanced interfacial interactions during the HSLD processing [35]. The sufficient control over the dispersion morphology and interfacial interactions are of great significance for the development of biologyadaptive materials containing biologically beneficial nanoenhancers like nanodiamonds and bioactive glass [36,37].
The effective exfoliation and proper dispersion of HANWs in PLA composites, combined with favorable interfacial interactions, are highly anticipated to confer large improvements for the mechanical properties. This was evidenced by the tensile testing of HANW-modified PLA composites ( Figure  4). As revealed by the stress À strain curves in Figure 4(a), the mechanical response upon external deformation of PLA composites was significantly promoted in the presence of HANWs. The lowest yield strength (51.8 ± 0.7 MPa) and elastic modulus (1.8 ± 0.06 GPa) were observed for pure PLA ( Figure  4(b,c)), in line with the property database established on the basis of common PLA resins [38][39][40]. Upon incorporation of well-exfoliated HANWs, both the strength and modulus were dramatically promoted with increase of filler content, gradually climbing up to 76.7 ± 0.8 MPa and 3.3 ± 0.16 GPa for PLA-HA30 (increase of 48% and 84% compared to those of pure PLA), respectively. This is in clear contrast to the normal scenarios revealing the limited promotion or even undesired decline in strength of PLA composites after addition of HA particles [40,41]. Moreover, it is of significance to observe that the ductility of HANW-modified PLA composites was not evidently sacrificed, as evidenced by the measured values of elongation at break in the range of 4%À6%, on a par with pure PLA (Figure 4(d)). The largely increased resistance to stress penetration endowed the PLA/HANWs composites with important prerequisites to remove the application constraints under changing intracorporeal environment [33,42].
In addition to notable promotion of mechanical properties, the HANWs enabled significant increase in the surface wettability and osteoblast growth and proliferation, both of which are important criteria in biomedical applications [43]. As illustrated by the water contact angle measurements ( Figure 5(a)), pure PLA was characterized by the highest contact angle of 127.4 due mainly to the relatively high hydrophilicity [13]. Upon incorporation of well-dispersed HANWs, the merits of high wettability were directly inherited by PLA composites. The contact angle was decreased by over 14 in the presence of 30 wt % HANWs, falling to the lowest value of 113.2 for PLA-HA30. It was assumed that the increased wettability was closely related to the rich oxygen functional groups like hydroxyl carried by the HANWs [44], which could be further enhanced by the gains of surface roughness with the existence of nanoscale whiskers [45].
To determine the HANW-enabled improvements of cell viability, MTT colorimetric assay was conducted to evaluate the MG-63 cell viability on the PLA composite films. As revealed in Figure 5(b), osteoblast cells proliferated on pure PLA were characterized by the lowest viability regardless of proliferation time, arising from the intrinsically inferior capability to stimulate cell growth and rare bioactive sites for accommodation of cell attachment. As increasing the proliferation time from 6 to 12 and 24 h, OD values at 490 nm for the cells cultured on pure PLA obtained stepwise increase from 0.22 to 0.38 and 0.52, respectively. With the existence of HANWs featuring high surface energy and bioactivity, the OD values witnessed moderate increase from 0.32 for PLA-HA10 to 0.43 for PLA-HA30 after proliferation for 6 h. The promotion was prominently enlarged with increasing time, achieving the highest OD value of 1.65 for the cells cultured on PLA-HA30 after 24 h. It meant a nearly 2-fold increase of cell viability compared to the pure PLA counterpart. Overlapping the contact angle measurements and the cell viability evaluation, the PLA/HANW composites could be identified as bioactive materials with improved surface hydrophilicity and cytocompatibility.
In the target applications of bone tissue engineering, the osteogenic performance with regards to osteoblast adhesion and stretching on the substrates represents an important criterion for cytocompatibility evaluation. Figure 6 shows the direct morphological observation for the biological response of the osteosarcoma cell line MG-63. It is evident that the osteoblast growth and proliferation on the PLA/ HANW composites were remarkably enhanced. Compared to the pure PLA counterpart ( Figure  6(a)), an increased density of cells with the lamellipoial protrusion were observed for HANW-modified composites, showing direct relation to the filler concentration ( Figure 6(b-d)). To illustrate the spatiotemporal coupling with the cell adhesion, the distribution of F-actin in the cells were discerned by the intensity plots of the phalloidin. It is apparent that a typical cytoplasmic distribution of F-actin was observed for pure PLA, in contrast to the regioselective distribution at the cell edges for HANW-modified composites. It primarily resulted from the HANW-enabled multiple improvements in surface roughness, hydrophilicity, cell attachment site and affinity to the seeded cells. In essence, the well-dispersed bioactive HANWs were ready to facilitate the anchoring interactions and accommodate more osteoblast cells, conferring the formation of elongated lamellipodia and pseudopodia [46]. Additional benefits may be associated with the HANWenhanced degradability of PLA, conferring proper and controlled degradation during the bone tissue regeneration.

Conclusions
The microwave-assisted biomineralization in SBF solution containing the TX-100 surfactant was used to synthesize HANWs, being preferentially nucleated and grown along the "soft template". The highly crystalline HANWs exhibited a well-controlled diameter ($30 nm) and length (80 À 120 nm), as well as a desirable Ca/P molar ratio of 1.67 on a par with that of naturally occurring bones. Aiming at competent translation of the high strength and excellent bioactivity of HANWs to the PLA composites, an HSLD method featuring a combination of high pressure (up to 50 kPa) and high shear rate approaching 10000 s À1 was developed to enable sufficient exfoliation and proper dispersion of HANWs. Both the mechanical strength and stiffness were largely improved for the HANW-modified PLA composites, as exemplified by notable elevation of 48% and 84% for the tensile strength and elastic modulus of PLA-HA30 compared to those of pure PLA (51.8 MPa and 1804 MPa), respectively. This was favorably accompanied by prominent improvements in cytocompatibility and osteoconductive capability, showing a nearly 2-fold increase of the cell viability after 24-h proliferation. The combined use of biomineralization and HSLD signifies an useful approach to controlling the intrinsic structure and dispersion morphology of biologically beneficial particles, thus empowering high-performance and highly bioactive composites that hold great promise for bone tissue repairing. University, in 2020. Her research interest focuses on biomedical materials and nanomaterials for healthcare.
Lv Ke received his B.Eng. degree from the School of Mechanical and Electrical Engineering, Xuzhou University of Technology in 2020. His recent research has been focused on PLA-based multifunctional composites for environmental and biomedical applications.
Zi-Lin Zhang is an undergraduate student in Xuhai College, China University of Mining and Technology (CUMT). His research interests are in sustainable functional nanofillers and PLA-based composites.
Kai-Zhe Zhang is a prospective Master student in Xuhai College, CUMT. His research focuses on development of environmentally friendly nanofillers and PLA-based functional composites.
Shenghui Zhang is an associate professor in CUMT. His research interests are in high-performance and lightweight composites for aerospace construction and transportation.
Yanqing Wang is an associate professor in CUMT. His research interests are in advanced alloys and high-performance composites for mining and architectural applications.