Impact of B 2 O 3 and La 2 O 3 addition on structural , mechanical and biological properties of hydroxyapatite

In this study, hydroxyapatite-B2O3-La2O3 composites (with ≤ 20 wt.% B2O3 and ≤ 2 wt.% La2O3) were synthesized via wet precipitation method and calcined at 1100 °C for 1 h. X-ray diffraction (XRD) analysis revealed the existence of the pure hydroxyapatite (HA) phase with high crystallinity. Characteristic absorption bands of HA were also observed in Fourier transform infrared spectra. Furthermore, scanning electron microscopy images demonstrated that the addition of B2O3 and La2O3 into HA enhanced the particle growth. Mechanical properties of the composites were studied by diametral tensile test and the results showed that incorporation of 10 wt.% B2O3 and 2 wt.% La2O3 led to a 39% increase in tensile strength (compared to the pure HA). In vitro cytocompatibility of HA-B2O3-La2O3 composites was investigated using Osteosarcoma Cell Lines (Saos-2). Incorporation of B2O3 and La2O3 into HA had no toxic effect towards the cells. Based on its tensile strength properties and biological response, composite of 88 wt.% HA, 10 wt.% B2O3 and 2 wt.% La2O3 was suggested as a promising composite for bone tissue engineering applications.


I. Introduction
Hydroxyapatite (HA, Ca 10 (PO 4 ) 6 (OH) 2 ) is a biocompatible and bioactive material that is used in different biomedical applications because of its structural similarity to bone mineral.HA has been found to have limited orthopaedics applications due to its brittle nature and low tensile strength.Thus, HA-based composites need an improvement in strength and toughness to be used as implant materials.The addition of metal oxides into HA, such as MgO [1], Al 2 O 3 [2], ZrO 2 [3], Nb 2 O 5 [4], ZnO [5], Bi 2 O 3 [6] and many others, have been done to improve its mechanical or structural characteristics [7].There are a few studies [8][9][10][11] which already revealed that the inclusion of a metal oxide might act as reinforcing agent, but at the same time, it may decompose HA into tricalcium phosphate (TCP).This decomposition has negative influences on densification and mechanical properties of HA [8,9,11].
It has been reported that boron oxide (B 2 O 3 ) augments the wound healing process [12] and has pivotal roles and functions in bone physiology [13].However, doping of HA with B 2 O 3 is scarcely recognised in such reports.Mixing of about 5% B 2 O 3 into HA might increase the flexural strength up to 58 MPa as well as the Vickers micro-hardness which can be around 2.1 GPa [11].In another study, Lee et al. [14] reported that the addition of B 2 O 3 into bioglass structure increased the osteointegration with no sign of toxicity.Yang et al. [15] showed that the addition of B 2 O 3 to SiO 2 -CaO-P 2 O 5 increased its degradation rate and bioactivity.
In addition, lanthanum oxide (La 2 O 3 ) is another vital oxide that can be used for improving properties of HA.It was reported [16] that HA-2.5 La 2 O 3 composite had higher hardness value than the pure HA which was attributed to the low porosity, high grain size and a low glassy phase in the composite.Henceforth, La 2 O 3 was substituted at a low amount and no suppression of bioactivity was observed for this doping [17].Aside, La 2 O 3 added into the silicates resulted in higher elastic modu-lus and hardness [17].Additionally, it improved tensile and bending strengths of HA [18].Indeed, an infiltration of a fibrous tissue with no noteworthy sign of inflammation was observed for La-containing HA discs after implantation into the bones of a rat [18,19].Fernandez et al. reported [18,20] that incorporation of 140-lanthanum into the apatite structure played a vital role in increased resistance of the hard tissues to acid dissolution.
In this study, microstructural and mechanical characteristics of HA-B 2 O 3 -La 2 O 3 synthesized by coprecipitation method were evaluated.To the best of our knowledge, no study yet had been reported for the co-substitution of B 2 O 3 and La 2 O 3 into HA structure.Moreover, in vitro cytocompatibility of doped HA using Saos-2 cell line was also investigated.HA was synthesized by the co-precipitation method using 0.5 M calcium nitrate and 0.3 M di-ammonium hydrogen phosphate which were previously dissolved in distilled water separately and then their pH was adjusted above 10 with NH 4 OH.The Ca/P ratio was kept at 1.67 to produce stoichiometric HA.Di-ammonium hydrogen phosphate solution was added dropwise into continuously stirred calcium nitrate solution.After stirring the HA solution for 2-3 h, it was heated at 90 °C for 1 h, then stirred for 1 day more at room temperature.The solution was filtered through a fine filter paper and washed several times to remove the remaining ammonia.The filtered wet cake was dried at 200 °C in an oven overnight to remove the excess water.Finally, the dried HA particles were ground to fine powder and calcined at 1100 °C for 1 h.

II. Materials and methods
In order to synthesize HA-B 2 O 3 -La 2 O 3 composites, the precursor powders were weighed according to the defined ratio (Table 1), mixed in pure ethanol and left for stirring for 3-5 h to obtain the best composite structure.The obtained mixture was then dried overnight to remove the excess alcohol and milled/mixed in a mor- tar and pestle to achieve the perfect distribution.After that, the composite powders were pressed and sintered at 1100 °C for 1 h.The prepared composites were denoted as xHAyBzLa, where x, y and z are portions of HA, B 2 O 3 and La 2 O 3 in composites, respectively.

Characterization
The phase purity of HA-B 2 O 3 -La 2 O 3 composites was determined by using X-ray diffraction technique (XRD, Bruker D8) operated at 40 kV and 40 mA utilizing Cu Kα radiation between 2θ angles from 20°to 70°with 2.0°/min resolution.The degree of crystallinity was calculated according to the fraction of crystalline phase from X-ray diffraction data by using following equation [21]: where X C is the degree of crystallinity, I 300 is the intensity of 300 reflection and V 112/300 is the intensity of the hollow between 112 and 300 reflection.Lattice parameters were calculated using an UnitCell software [22,23].Fourier transform infrared spectroscopy (FTIR) analysis was conducted to study the functional groups with the infrared spectra measurements in the range of 4000-400 cm -1 (Bruker IFS66/S).Morphology and surface microstructure of HA-B 2 O 3 -La 2 O 3 composites were investigated using scanning electron microscopy (SEM, LEO 1430VP).The average grain size of the sintered samples was determined from the SEM images.Intercept method was used to determine the grain size of the sintered samples [24,25].

Disc shaping and sterilization
Initially, HA-B 2 O 3 -La 2 O 3 powders (670 mg) were uniaxially pressed into discs (14 mm × 2 mm) using automatic hydraulic machine (Carver, Inc.) at 10 kN load for 1 minute, then sintered at 1100 °C for 1 h prior to their use for mechanical analysis.Discs were sterilized at 200 °C for 2 h before performing in vitro cytotoxicity test.

Mechanical analysis
Density (ρ a ) of the sintered composites was determined by measuring their dry weight (m) and calculating the volume (V) obtained from the pellet dimensions using the following formula: Theoretical density of the composites (components a (HA), b (B 2 O 3 ) and c (La 2 O 3 )) was calculated using the densities (ρ), and known weights (W) via the following formula: The relative density of the composites was calculated by dividing the measured density with a theoretical density of each composite.Theoretical density of the pure HA is 3.156 g/cm 3 .The porosity of HA-B 2 O 3 -La 2 O 3 discs was determined by mercury intrusion porosimetry (MIP, Auto-Pore IV 9520, Micyromeritics Instrument).The samples were subjected to a pressure range of 50-60000 psi which can give measurements for pores of radii from 3 nm to 360 µm.
The Vickers micro-hardness (HV) measurements were determined using a Vickers micro-hardness tester (HMV-2, Shimadzu, Japan).Approximately 20 measurements were performed on each sample with a diamond indenter at 1.968 N load for 20 seconds.The average micro-hardness was calculated by the following formula: where HV is the value of micro-hardness, P stands for the amount of load applied to the sample and d is the length of indent diagonal.
The diametral tensile tests of the sintered discs were performed by a universal testing machine (LS 500, Llyod Instruments, UK).The disc was compressed diametrically between two parallel plates.During the compression, a maximum tensile strength was generated across the flat surface keeping diameters of the discs normal to the loading direction.The tensile strength of the samples was calculated with the following formula [24,26]: where F is failure force, D is sample diameter and t is the sample thickness.The diametral tensile test was conducted in triplicates for each experimental group (n = 3).
In order to evaluate the machinability of the specimens, drilling test was performed on the pellets.The machining device was a 3-axis system with a spindle speed of 600 rpm.A 0.05-inch twist drill bit was used to test the samples.Machinability test of the discs was performed in duplicates for each group.

In vitro cytotoxicity evaluation
In vitro cytotoxicity studies were conducted using Saos-2 cells (human osteosarcoma cell line).In this work, the Saos-2 cells were grown in high glucose Dulbecco's modified Eagle medium (DMEM), further supplemented with 10% fetal bovine serum (FBS) and 0.3% Penicillin-Streptomycin.The initial cell seeding density was 5 ×10 4 cells/disc.The cells were incubated on composite discs in a carbon dioxide incubator (5215, Shell Lab, USA) at 37 °C under 5% CO 2 humidified environment.The medium was refreshed every 2 days.The proliferation of the cells on the discs was analysed by PrestoBlue ™ viability assay at different incubation periods (1, 3 and 7 days).At each time point, PrestoBlue ™ viability reagent was added to wells containing cell seeded discs and then the well plates were incubated at 37 °C under 5% CO 2 environment for 20 h.During incubation PrestoBlue ™ reagent was reduced into intensely coloured formazan, as a result of enzymatic activity of the viable cells.After 20 h of incubation, the medium in each well was collected and their absorbance was read by µQuant ™ spectrophotometer (Biotek Instruments Inc., USA) with an excitation wavelength of 570 nm and an emission wavelength of 600 nm.The amount of metabolic reduction caused by cells was then calculated.After collection of the medium, cells were washed with PBS once more and growth medium was added for further culturing of cells.The viability of cells seeded on tissue culture plate was used as positive control.All experiments were performed in quadruplicate.The measured data were expressed as a mean value with corresponding standard deviation.One-way analysis of variance (ANOVA) was carried out with Tukey's Multiple Comparison Test for the post hoc pairwise comparisons using SPSS-15.0Software (SPSS Inc., USA).Differences were considered as significant for p ≤ 0.05.
Morphology of Saos-2 cells on composite discs was investigated using SEM analysis.Two samples for each composition were seeded with cells at an initial seeding density of 5 × 10 4 cells/disc.After 1 and 7 days of incubation, the medium was removed, and the cells were washed with phosphate buffered saline (PBS) and then fixed with 3% glutaraldehyde in PBS for 15 minutes.After fixation, the cells were washed with PBS and dehydrated with graded ethanol-water solution series (30, 50, 70, 80, 90 and 100%) and hexamethyldisilazane (HMDS).The experiments were achieved in duplicates for each group.Prior to SEM analysis, discs were coated with gold-palladium via precision etching coating system (PECS, Gatan 682, USA) to a thickness of 5 nm.

Structural analysis
XRD patterns of the HA-B 2 O 3 -La 2 O 3 composites sintered at 1100 °C are given in Fig. 1.All the patterns showed the characteristic peaks of HA and they were in good agreement with stoichiometric HA (JCDPS No: 09-432).No β-TCP peaks around 31.03°were observed, which confirmed the phase purity of HA.XRD pattern showed the incorporation of B 2 O 3 and La 2 O 3 in the HA structure (90HA10B and 88HA10B2La) and there was no decomposition of HA into biphasic or triphasic calcium phosphate.Peaks located at 27.67°a nd 29.98°belong to (310) and (222) diffraction planes and they are in good agreement with B 2 O 3 structure (JCPDS No: 06-0297).The intensity of XRD peaks indicated an increase in the crystallinity of composites which could be ascribed to the incorporation of La 2 O 3 in HA [27].Similarly, a small peak appearing around 29.91°for 88HA10B2La composite is attributed to the (101) diffraction plane and assigned to the incorporation of La 2 O 3 in the HA sample.This result corresponds to standard La 2 O 3 (JCPDS No: 05-0602).In the XRD pat-tern of 80HA20B composite a small peak observed at 30.57°might belong to the diffraction plane (113) and according to standard (JCDPS No: 26-0347) it refers to calcium borate (Ca 3 B 2 O 6 ).Calcium borate formation can be explained according to the following reactions (Eq.6a-c) [28][29][30][31]: Calcium borate is identified as a potential antiinflammatory agent, therefore, it can be considered as a beneficial additive in the samples [32].Noticeable decrease in intensity of the peaks was observed and this decrease could be interpreted as the lower crystallinity of 78HA20B2La sample compared to other composites (Table 2).The amount of B 2 O 3 and La 2 O 3 incorporated together may be one of the reasons for the decomposition of HA structure into β-TCP and CaO (Eq.7) [33]: However, amount of β-TCP formed was still lower compared to other previous studies conducted on ZrO 2 and Al 2 O 3 -HA composites [8,9].
Lattice parameters of the pure HA and HA-B 2 O 3 -La 2 O 3 composites are summarized in Table 2 and the lattice values were in good agreement with dimensions of hexagonal structure of HA.Few variations in a and c axis of HA-B 2 O 3 -La 2 O 3 composites could be attributed to the dehydroxylation of the hydroxyl group which is very common for the samples calcined at high temperatures [34] or due to the partial substitution of a smaller radius Ca 2+ (0.99 Å) with a larger radius La 3+ ion (1.17Å) [18].However, the addition of both B 2 O 3 and La 2 O 3 into HA did not result in significant changes in XRD patterns of the HA as they were still matched with the standard JCPDS file for HA.
FTIR spectra of the pure HA and HA-B 2 O 3 -La 2 O 3 composites are illustrated in Fig. 2. The pure HA spectrum showed an absorption peak at the position of 474 cm -1 which was attributed to ν 2 (PO 3− 4 ) group.Another two absorption peaks located at 1056 and 1096 cm -1 were attributed to ν 3 (PO 3− 4 ) group, the peak positioned at 962 cm -1 was allocated to ν 1 (PO 3− 4 ) group.Besides these peaks, two absorption peaks located at 571 and 601 cm -1 were assigned to ν 4 (PO 3− 4 ) group.OH -groups were detected in HA structure in two sites (630 and 3572 cm -1 ) corresponding to bending and stretching modes of OH -group, respectively.In addition to the stretching and deformation vibrational modes of PO 3 -  4   groups, a tiny peak with low intensity was observed at 1460 cm -1 which is related to the CO 2 -  3   group.This result may be affected by contaminants during the synthesis process.A gradual decrease in the intensity of the OH -band was observed indicating the partial dehydration of HA as a result of sintering at 1100 °C.The increased sharpness of PO 3 -  4   peaks due to the doping of B 2 O 3 and La 2 O 3 could be associated with the highly crystalline structure as confirmed by the XRD results (Fig. 1).Along with absorption peaks of the pure HA, other peaks were detected at 742 and 780 cm -1 , which could be assigned to the ν 2 (BO 3− 3 ) symmetric bending.Similarly, absorption band in the range of 1205-1245 cm -1 was attributed to the asymmetric stretching ν 3 mode of BO 3 -  3   group [35].Surface morphologies of the pure HA and HA-B 2 O 3 -La 2 O 3 composites were examined by SEM (Fig. 3).SEM images showed that particles of the pure HA were densely packed and the grain size was 192 nm, as illustrated in our previous study [36].Micrograph images confirmed that incorporation of B 2 O 3 resulted in an increase in the porosity as well as in the grain size (342 nm and 364 nm for the 90HA10B and 80HA20B samples, respectively).The results are in good agreement with the previous studies [37,38].Likewise, the introduction of La 2 O 3 into the HA resulted in no big difference in the morphology, however, a larger grain size (615 nm) was observed.Our results are in good agreement with the previous studies [39,40].The crystallinity of the 78HA20B2La composite showed less improvement compared to other groups, furthermore, it showed highly porous structure (Table 3).However, as reported by previous studies, an interconnected porous structure is extremely important for hard tissue applications [41].

Mechanical analysis
Density and relative density of the materials are summarized in Table 3. Relative density of the pure HA was decreased by addition of B 2 O 3 and La 2 O 3 .The presence of weak interphases of B 2 O 3 and La 2 O 3 particles in between HA particles prevented HA particles to sinter.Moreover, this hindered sintering upon increasing amount of dopant led to the reduced relative density [42].Furthermore, the relative density is considered as a vital biomaterial parameter owing to its significant effect on potential for steric hindrances to cell motility [43].It is reported that the material with lower relative density enhanced NR6 fibroblasts cell migration within the scaffold variants [44].
Vickers micro-hardnesses of the pure HA and HA-B 2 O 3 -La 2 O 3 composites are given Fig. 4a.The pure HA exhibited the highest hardness of 5.13 GPa.However, the hardness of the pure HA sharply decreased with an addition of B 2 O 3 and La 2 O 3 .These decreases in micro-hardness value were presumably due to the highly porous nature (Table 3) [45] and increase in the grain size of the HA-B 2 O 3 -La 2 O 3 composites (Fig. 3) [46].
Diametral tensile strength results are illustrated in Fig. 4b.It was observed that the 88HA10B2La and 80HA20B composites exhibited the highest tensile strength in comparison to other compositions which could be due to a high degree of crystallinity.Tensile strength increases with increasing degree of crystallinity [47,48].Similarly, the 78HA20B2La sample showed the lowest tensile strength value which could be due to the  reduction in the degree of crystallinity (Table 2).In addition, it was reported that the presence of a second phase had a considerable effect on the tensile strength that might decrease the mechanical properties of ceramics [24].Moreover, our result was far away from dense HA [49] which could be due to the porous nature of HA-B 2 O 3 -La 2 O 3 composites as observed in SEM images (Fig. 3) [26].The bioceramics used in hard tissue engineering should possess good machining capacity, as they should be frequently subjected to redesign during orthopaedic surgery.Figure 5 presents the samples after drilling test.
None of the composite discs break during the process while HA one broke, as illustrated in our previous study [36].This behaviour could be attributed to the brittle nature of HA, and similar observations were reported in literature [42].Thus, the composites are machinable with no evidence of large-scale cracking or chipping.
The key reason for the enhanced machinability of ceramic composites could be due to many issues such as: i) the uniform distribution of lanthanum [50], ii) the creation of weak interphases of B 2 O 3 and La 2 O 3 particles in between the HA particles [42] (tthese interphases of B 2 O 3 and La 2 O 3 particles induce interfacial de-bonding and crack deflection during drilling, thus led to the inhibition the crack growth [42]), iii) the porous nature of HA-B 2 O 3 -La 2 O 3 composites (it is known that the porosity plays a vital role in the optimal machinability during which pores could have prevented the catastrophic destruction of materials during drilling through crack defection [51]) and iv) the presence of secondary phases.

In vitro cytotoxicity evaluation
HA is well known as a biocompatible material and has no toxic influence on osteosarcoma cells (Saos-2) [52].However, our main interest was to evaluate the impact of B 2 O 3 and La 2 O 3 addition on biological properties of HA.Thus, Saos-2 cell proliferation was evaluated using PrestoBlue ™ viability assay (Fig. 6).The results showed that the cell viability on HA discs after one day was statistically higher than all samples except the composite 88HA10B2La.On days 3 and 7 cell viability on all composite discs was statistically lower than the one observed on HA discs.
After one day the highest cell viability was observed on the HA and 88HA10B2La [53].There was no statistical difference between the HA and 88HA10B2La in terms of cell viability at day 1 which could be due to highly crystalline structure of the composite 88HA10B2La in comparison with other samples and it was assigned to the enhanced crystallinity of the sample (Table 2).Increased crystallinity might have inhibited the degradation rate of HA-B 2 O 3 structure as a result of low ionic exchange with the environment [54].Cells viability observed in other composite groups was statistically lower than observed on HA discs.Incorporation of 10 wt.% and 20 wt.% of B 2 O 3 in HA showed slightly toxic effect.It was reported that incorporation of La 3+ increased the proliferation rate of MC3T3-E1 cells [18,55].For all composites, a time-dependent decrease in the cell viability was observed.This could be due to the initial confluences reached at some areas on discs after cell seeding.However, only on boron doped discs, cell viability significantly decreased with time and cell viability decreased to a level of 20% which indicates that boron ion dissolution caused a toxic effect in cells.Among the composites, cell viability on the 88HA10B2La decreased only 20% after 7 days and is comparable with the TCP sample.Therefore, this group can be considered as cytocompatible [43].
SEM images of cells on HA-B 2 O 3 -La 2 O 3 composite discs after 1 and 7 days of incubation are given in Fig. 7.After 1 day of incubation the cells spread on the surface of the discs, attached to and covered disks sur- face forming layers of cell sheets.Complete coverage indicated that surface properties were ideal for cell attachment and spreading.It was reported that pores and tiny cracks make specific surface area larger and easier to bond with surrounding tissue in vivo, thus providing brackets and passages for the growth of the new bone tissue, hence the better biocompatibility [56].In some SEM images, cell filopodias were broken which was reported to be the result of drying out during the fixation process or a sign of a necrotic cell [57].

IV. Conclusions
Graded HA-B 2 O 3 -La 2 O 3 composites were prepared by wet co-precipitation method with the incorporation of different ratios of

Figure 1 .
Figure 1.XRD patterns of pure HA and HA-B 2 O 3 -La 2 O 3 composites sintered at 1100 °C

Figure 6 .Figure 7 .
Figure 6.The proliferation rate of cells on pure HA and HA-B 2 O 3 -La 2 O 3 discs (error bars represent standard deviation) B 2 O 3 and La 2 O 3 as additives.Incorporation of B 2 O 3 and La 2 O 3 in HA lattice caused an augmentation of the degree of crystallinity and shifting of lattice parameter.The increased sharpness of adsorption peaks in FTIR described the highly crystalline structure as confirmed by the XRD results.SEM images confirmed that incorporation of B 2 O 3 and La 2 O 3 resulted in an increase in the porosity as well as an increase in the grain size up to 615 nm.Density was decreased with incorporation of B 2 O 3 and La 2 O 3 .88HA10B2La and 80HA20B composites had the highest tensile strength among all other composites.Furthermore, the incorporation of B 2 O 3 and La 2 O 3 improved the machinability of HA composites, thus being workable and easy to design.The results from in vitro cytotoxicity tests provided clear evidence that 88HA10B2La composites were cytocompatible.Mechanical and biological evaluations of B and La incorporated HA composites (88HA10B2La) indicate that they can be a promising bioceramics implant for use in bone tissue engineering.

Table 2 .
Lattice parameters and degree of crystallinity of pure HA and HA-B 2 O 3 -La 2 O 3 composites

Table 3 .
Relative density and porosity values of the materials sintered at 1100 °C for 1 h