Combination of Na-Ca-phosphate and uorapatite in wollastonite-diopside glass-ceramic: degradation and biocompatibility

In this study, we investigated the effect of introducing sodium calcium phosphate (NCP), uorapatite (FA), or the combination of both phases in the wollastonite-diopside (WD) bioactive glass-ceramic system on the crystalline phases formed, microstructure, degradation, and biocompatibility of those materials. The prepared materials were characterized by DTA, XRD, and SEM/EDX. Moreover, the density was measured via Archimede’s method, and the mechanical properties were measured by Vicker’s microhardness indenter. The in vitro bioactivity test was carried out in the simulated body uid (SBF), and the cell viability test was evaluated using the Vero cells. The results showed that the formed crystalline phases were close to the starting proposed phases. Moreover, NCP-containing WD glass-ceramic was showed the lowest density value due to its low densication, and accordingly, it showed the lowest Vicker`s microhardness value due to the same reason. Furthermore, combining sodium calcium phosphate in WD glass-ceramic was increased cell viability better than that included uorapatite, whereas, the combination of both crystalline phases in WD glass-ceramic led to an increase in the cytotoxicity to the highest value. In conclusion, different properties of wollastonite-diopside glass-ceramics can be tailored by the combination of NCP or FA, and hence, these glass-ceramic materials can be modied effectively according to the purpose for which it is intended to be applied. The obtained results indicated that different properties of WD glass-ceramic materials can be tailored by the combination of NCP and/or FA. Hence, these glass-ceramics are expected to be useful materials in promising biomedical applications, such as orthopedics and dentistry.


Introduction
Glass-ceramic materials have to pay more attention to the biomedical uses meanwhile the innovation of bioglass in 1971 by Hench, et al [1] . They are being used broadly for biomedical requests as orthopedic implant and bone ller materials, as a result of their biocompatibility and tight bonding with bone [2,3] .
The bioactive glass-ceramics have a signi cant place in the biomedical application, due to their higher mechanical strength compared to the derived glass thanks to their unique microstructure [4][5][6] . Moreover, they are capable to bond with the living bone directly by the growth of healthy tissue onto their surface through a biologically active hydroxycarbonate apatite (HCA) layer without the formation of surrounding brous tissue. The HCA phase that forms on the bioactive glass-ceramic surfaces is chemically and structurally equivalent to the mineral phase of the bone. This correspondence has an effective role in the interfacial connecting between the bioactive material and bone [7] .
Among various silicate materials, pyroxene glass-ceramics had created a global interest for several applications recently [8][9][10][11] in material science and technology owing to a brilliant set of physical properties. Diopside (CaMgSi 2 O 6 ) is a signi cant member of the pyroxenes class. Glass-ceramic materials based on diopside have high mechanical and excellent chemical properties with more distinct biological activity. These properties make them interested in biomedical applications as bone tissue engineering [10,[12][13][14][15][16] . Glass-ceramics containing wollastonite (CaSiO 3 ) crystals are promising candidates for bone tissue regeneration as a result of their great mechanical properties, good bioactivity, and biocompatibility [17,18] . The presence of wollastonite crystals in glass-ceramics was reported to increase the strength and toughness of those materials [19] . Expectedly, combining diopside and wollastonite crystals is likely to produce glass-ceramics with excessive mechanical strength and outstanding bioactive properties. Accordingly, there have been several works that studied diopside-wollastonite glass-ceramics for biomedical applications. Therefore, wollastonite-diopside glass-ceramics have been used widely for orthopedics and bone regeneration applications [20][21][22][23][24] . Moreover, this glass-ceramic system was modi ed by the addition of other crystalline phases, such as tricalcium phosphate, uorapatite, and sodium calcium phosphate.
Fluorapatite glass-ceramics are superior materials for bone grafting and dental applications thanks to the antibacterial properties of the uorine ion [25] . Kokubo, et al. were the rst authors' combined apatite with wollastonite to prepare the binary apatite-wollastonite glass-ceramics with better mechanical properties [26] . The phase formation in the uorapatite-diopside binary system was studied as a new type of bioglass-ceramics possessed high chemical and mechanical properties with great biocompatibility [27] .
Furthermore, uorapatite has been introduced to wollastonite-diopside glass-ceramic to impart more potential properties for different biomedical applications [28,29] . For developing biomaterials with talent properties, Salinas, et al. [2] were succeeded to prepare bioactive glass-ceramics depending on the ternary Ca 3 (PO4) 2 -CaSiO 3 -CaMgSi 2 O 6 system. Salman, et al. [30] prepared glass-ceramic based on CaMgSi 2 O 6 -CaSiO 3 -Ca 5 (PO 4 ) 3 F-Na 2 SiO 3 system with TiO 2 or ZnO additives. They stated that, by in vitro results, the glass-ceramics containing Na-Ca silicate showed higher bioactivity to be suitable for restorative dental and bone-implant materials. Herein, we aimed to study the effect of introducing of Na 4 Ca(PO 3 ) 6 and/or Ca 5 (PO 4 ) 3 F phases in the wollastonite-diopside bioactive glass-ceramic system on the physicochemical properties, such as density, degradation, and biocompatibility, as well as, the mechanical properties (microhardness) of the resulting glass-ceramics for biomedical uses as bone replacement materials. The crystalline phases formed and microstructure characterization are also studied.

Glass synthesis
The batch compositions were designed depending on the weight proportions of diopside-wollastonitesodium calcium phosphate-uoroapatite phases as presented in Table 1. The glass samples were prepared via a conventional melt-quenching technique. The calculated weights of high purity (> 99.9 %) grade reagent powders calcium carbonate (CaCO 3 ), magnesium carbonate (MgCO 3 ), sodium carbonate (Na 2 CO 3 ), ammonium di-hydrogen phosphate (NH 4 H 2 PO 4 ), calcium uoride (CaF 2 ), and silica (SiO 2 ) were well thoroughly mixed. The mixed powders were melted in alumina crucible at 1400-1450°C for 3 h in an electric furnace (Carbolite HTC 1600, Vechstar). The melting process was continued with stirring until clear homogeneity. Then, the melts were cast on a preheated stainless steel mold in the designed shapes.
The cast glass species were moved to an annealing furnace at 450°C in an electric furnace. Then, the furnace turns off after 1 h and allowed to cool naturally to room temperature for preparing hard glass samples free from strain and stress.

Differential Thermal Analysis (DTA)
All prepared glass samples were analyzed by Differential Thermal Analysis (DTA-SDTQ600-TA Instruments, USA) to determine their critical thermal characteristics such as the glass transition temperatures (Tg), crystallization temperatures (Tc), and any other endotherms up to 1100°C. A very ner 30 mg powder of each glass sample (< 38 µm) were analyzed in alumina crucible using analytical grade alumina powder as a reference with a heating rate of 10°C/min from room temperature up to 1100°C in a owing nitrogen atmosphere.

Glass-ceramic preparation
Double stage heat-treatment schedules were applied to investigate the glass-ceramic materials via the controlled crystallization process of glasses. According to the DTA data, each glass composition was heat-treated for 3 h near the glass transition temperature, in the rst stage, to provide adequate nucleation sites which occur nearly in the endothermic temperature. Then, the temperature was raised to the exothermic temperature for 6 h, in which the crystal growth was expected (the second stage) at a constant heating rate of 10°C/min.

X-ray diffraction analysis (XRD)
The developed crystalline phases of the glass-ceramic specimens were identi ed employing the X-ray powder diffractometer (XRD, PANalytical PW3040/60, Netherlands). Cu-Kα radiation source (λ = 1.5401 Å), was used with generated at 30 mA and 45 kV with a scan speed of 2°/min over a 2θ-range of 5-75°.
The diffraction Peaks of the investigated crystalline phase are analyzed with JCPDS les and noticeable with different symbols.

Microstructure study (SEM)
The microstructure of the investigated glass-ceramic samples was analyzed for studying the internal morphology of the crystal phases formed. The fractured surfaces for the selected samples were immersed in a 2 vol% HF as an etching solution for one minute to remove the residual glassy phase. Then the samples were washed with distilled water and their fractured surfaces were coated with a thin lm of Au, by SEM Coating unit, to reduce any charging effects. Finally, they were observed by using scanning electron microscopy (SEM/EDS, Quanta FEG 250, Netherlands) that operated at an acceleration voltage of 20 kV.

The density test
The apparent density of the bulk glass-ceramic samples was measured, to an accuracy of (± 0.0002) via the Archimedes method by immersion all specimens in distilled water as buoyant liquid. The mean and the standard deviation values are shown for density have been achieved from at least 6 different specimens for each sample. Molar volume was calculated using the obtained density data for all measured samples of the bulk glass-ceramics.

The Mechanical properties (Vicker's microhardness)
The microhardness of the investigated glass-ceramics was measured to study the mechanical properties using a Vicker's microhardness indenter (ModelHV-100 microhardness tester) with an applied load of 100 g xed for all samples for 15 s. For indentation testing, all glass-ceramic specimens were cut and polished carefully to nd at and smooth parallel surfaces for each specimen. At least 7 Vickers impressions were made on the polished surface of each sample with an estimated accuracy of (± 0.5 µm). The nal hardness value was calculated from Eq. (1): A is a constant equal to 1854.5, P is the applied load (g) and d is the average diagonal length (mm). The microhardness values are changed from kg/mm2 to MPa by multiplying with a constant value of 9.8.

In vitro bioactivity test
The glass-ceramic degradation was carried out in the simulated body uid (SBF), prepared as described previously [31] . The bulk glass samples of 0.2 g were immersed in 15 ml SBF for 1, 3, and 7, 14, and 22 d at 37°C. At the predetermined period, the pH of the soaking solution was measured by pH meter (3040 Ion Analyzer), and the sample was removed from the solution and dried at 70°C, and then weighed by the electronic balance of four decimals. The weight loss (%) was calculated under Eq. (2).
Wt. loss % = (W i -W d /W i ) × 100 (2) Where, Wi and Wd are the initial weight and weight after soaking in SBF, respectively. The whole immersing solution was collected and replaced by a fresh one, and the concentrations of ions released

Microscopic Studies
After the end of the treatment, the wells were washed three times with 100 µl of phosphate-buffered saline (pH 7.2), and then the cells were xed to the plate with 10 % formalin for 15 min. The xed cells were then stained with crystal violet solution. After drying, the cellular morphology was investigated using an inverted microscope (CKX41; Olympus, Japan).

DTAanalysis
The thermal behavior of the studied glasses was explored by the DTA technique to identify the glass transition (Tg) and glass crystallization temperatures (Tc) which are useful to determine the most appropriate crystallization temperatures for each composition. Figure 1 and Table 2 represent the characteristic transition temperatures (Tg) and crystallization temperatures (Tc), for the obtained glasses as detected from DTA traces. The DTA analysis of the base composition (sample WD1) showed that an endothermic peak effect appeared at 747 oC. This effect indicative for the Tg, at which the atoms begin to arrange themselves in preliminary structure elements formerly the crystallization step. On the other hand, an exothermic peak was detected at 1038 oC represented the Tc, at which the growth of crystals has occurred completely. The recorded data revealed that additions of 10 wt.% of sodium calcium phosphate-Na 4 Ca(PO 3 ) 6 (sample WD2) or ourapatite-Ca 5 (PO 4 ) 3 F (sample WD3) at the expense of diopsideCaMgSi 2 O 6 phase in WD1 composition led to a decrease in the Tg to lower temperature from 747 o C to 733 o C (Fig. 1). Introducing of both Na 4 Ca(PO 3 ) 6 and Ca 5 (PO 4 ) 3 F phases instead of 20 wt.% CaMgSi 2 O 6 phase (sample WD4) in the base glass composition led to a reduction of both Tg and Tc to lower temperatures (721°C and 1027°C, respectively). This can be attributed to the effect of Na 2 O to decline the glass viscosity and accelerating the nucleation stage. Denry and Holloway [32] reported that the sodium acts as a glass network modi er capable of lowering the viscosity of silicate glasses by increasing the number of non-bridging oxygen in the glass composition. As a result of the decrease in the viscosity, the mobility and diffusion rates of the different ions and ionic complexes forming glasses during the crystallization process will be markedly increased, leading to higher crystallization tendency.
Introducing of Na 2 O in the glass composition enhanced its crystallization tendency, leading to displacing of Tg and Tc towards lower temperatures. Similarly, P 2 O 5 additions to the silicate glass composition led to accelerating volume nucleation and support glass-ceramic formation [33] .

XRD analysis
The existence of various phases formed in the studied glasses was a function of the synthesized glass compositions and the crystallization parameters used. The XRD patterns of the investigated glassceramics (WD1, WD2, WD3, and WD4) are displayed in Fig. 2 and the crystalline phases developed by the reheating glasses are given in Table 2. It can be noticed from the gure that the diffraction pattern of the base glass sample (Fig. 2, pattern I), WD1 heat-treated at 745 o C/3h-1040 o C/6h, was crystallized into pyroxene members of diopsidic-type (CaMgSi 2 O 6 ) as a major phase (PDF # 17-0318) beside wollastonite-CaSiO 3 phase (PDF # 27-1064). Diopside and wollastonite phases are proper varieties of pyroxene minerals in which the developed glass-ceramics have high mechanical strength and great chemical durability [34] . Glass-ceramics with high Ca contents possess a high ability of good apatite mineralization with excellent bioactive properties [35] .
The addition of 10 wt.% sodium calcium phosphate at the expense of 10 wt.% diopside phase in the base composition, i.e. WD2, heat-treated at 730 o C/3h-1065 o C/6h, led to the crystallization of sodium calcium phosphate phase with a new formula -Na 4 Ca(PO 3 ) 3 (PDF # 23-0669) in addition to diopside (as a major phase) and wollastonite phases as proved by the X-ray diffraction analysis (Fig. 2, Pattern II). Whereas, introducing of 10 wt.% uoroapatite phase at the expense of the diopside phase (glass sample WD3 heated at 735 o C/3h-1075 o C/6h) resulted in the development of uoroapatite (PDF # 74-4390) as a new phase together with the major diopside phase. The existence of the wollastonite phase as a secondary phase was also detected as indicated from XRD patterns (Fig. 2, pattern III). More Ca 2+ ions are included in the media with P 5+ ions which have a higher ionic eld strength, and hence a greater tendency to shielding. It was also reported that P 2 O 5 had a high tendency to react with CaO [36] . Thus, uoroapatite-Ca 5 (PO 4 ) 3 F crystallized from the droplet phase with highly concentrated Ca and P in the presence of CaF 2 in the media [37] . On further decreasing the diopside components by replacing 20 wt.% of CaMgSi 2 O 6 phase with 10 wt.% of Na 4 Ca(PO 3 ) 6 and 10 wt.% of Ca 5 (PO 4 ) 3 F in the base composition (glass sample WD4 heat-treated at 720 o C/3h-1025 o C/6h) resulted in a formation of various crystalline phases, which were diopside, wollastonite, uoroapatite and sodium calcium phosphate phases as detected by the X-ray diffraction analysis (Fig. 2, pattern IV). The high content of CaO and P 2 O 5 , by introducing both Na 4 Ca(PO 3 ) 6 , Ca 5 (PO 4 ) 3 F phases, leads to intense phase separation in the glass. Therefore, phase separation is likely to occur, and the phosphate-rich and silicate-rich phases are separated from the parent glass [37] .

SEM analysis
The SEM images for the WD1, WD2, WD3, and WD4 glass-ceramic samples are displayed in Fig. 3 which presents different morphological shapes as a function of oxide content and crystalline phases formed. It was clear from the gure that the volume crystallization of tiny aggregates-like growths was the distinctive microstructure formed in the WD1 sample (Fig. 3a). The introducing of 10 wt.% Na 4 Ca(PO 3 ) 6 was caused a change in the crystals microstructure to become dendritic like-growths with some glassy matrix in-between as seen in the micrograph of the WD2 sample (Fig. 3b). Whereas, the volume crystallization of globular grains were the distinctive shape in the WD3 sample with 10 wt.% of Ca 5 (PO 4 ) 3 F (Fig. 3c). The replacement of diopside by both Na 4 Ca(PO 3 ) 6 (10 wt.%) and Ca 5 (PO4) 3 F phases (10 wt.%) in the base composition (sample WD4) led to developing bundles of oriented long rods with aggregates of spherical crystals (Fig. 3d).

Density
The obtained average values of the measuring density in the laboratory by Archimedes method for the bulk glass-ceramic species heat-treated at different temperatures are presented in Table 2 and Fig. 4a. The general trend of increasing the density for bulk glass-ceramic than that in the corresponding glass was caused by an increase in the temperature during the heat treatment process, which led to the elimination of the remaining pores or due to the growth of more amounts of denser crystalline phases [11] .
The density of WD1 glass-ceramic was 2.76 g/cm 3 , whereas, it was 2.50 g/cm 3 , 2.92 g/cm 3 , and 2.90 g/cm 3 for WD2, WD3, and WD4 glass-ceramic samples, respectively. The obtained density data revealed that the WD2 glass-ceramic was had the lowest value of the density due to less densi cation with increasing the free volume of the material and decreased its density [38] as a result of the presence of some glassy matrix in-between crystals of the formed microstructure as seen in Fig. 3(b). Generally, the degree of crystallinity of the glass-ceramic is more effective in their corresponding densities [39] . So, the density values of the WD3, and WD4 glass-ceramic samples were more than that of WD1, P 2 O 5 -free sample, which improved the crystallization process and increase the crystallinity of the investigated glass-ceramics and led to form denser microstructure, as shown in SEM micrographs ( Fig. 3c and d).

The mechanical properties (Microhardness)
Bones in the body are constantly subject to changing mechanical stresses due to the variety of body movements. So the mechanical feature is very important and must be taken into account when designing glass-ceramic materials for use in bone tissue in bioengineering applications [11] . The mechanical properties of glass-ceramics mainly depend on the mechanical characteristics of the crystalline phases formed, including the internal microstructure of their crystals, the interfacial bonding among the crystals, and vitreous phases formed in-between, as well as, the ability of the crystals to condense. The microhardness property for the resulting glass-ceramics was determined by Vicker's microhardness to reach the suitable data which are comparable with those of human bone (680-8000 MPa) [40] . The average hardness data measured for the prepared glass-ceramic species are displayed in Table 2 and Fig. 4b. The results showed that the synthesized glass-ceramics were presented good mechanical properties with high microhardness values of around 4560-5015 MPa. This may be attributed to the presence of high mechanical properties of pyroxene members [34] , diopside, and wollastonite, as the major crystalline phases. The replacement of sodium calcium phosphate at the expense of diopside components explaining the decrease in the microhardness value observed for the resultant WD2 sample (4560 MPa) compared to the WD1 sample (4850 MPa) due to the formation of considerable volume fractions of a mechanically weaker NaCa(PO 3 ) 3 phase instead of a mechanically stronger diopside phase [12] . While the hardness values were increased signi cantly for WD3 and WD4 samples, reaching 5045 MPa and 5015 MPa, respectively. Kansal, et al. [13] reported that the combination of high mechanically bioactive material of pyroxene members as diopside CaMgSi 2 O 6 and wollastonite-CaSiO 3 phases with resorbable composite such as uorapatite-Ca 5 (PO 4 ) 3 F led to the development of glass-ceramic materials with good mechanical strength and distinct biologically active properties.

In vitro bioactivity test
The degradation and in vitro bioactivity tests of the glass-ceramic samples under investigation were carried out in SBF. Accordingly, the possibility of the formation of a new layer of hydroxyapatite (HAp) was examined by using SEM coupled with EDX analysis. Furthermore, the pH, weight loss %, and ion concentrations (calcium and phosphate ions) were also monitored as a function of time. Figure 5 shows SEM/EDX analysis of WD1, WD2, WD3, and WD4 glass-ceramic samples after immersion in SBF for 22 d.
It can be observed from the gure that newly very small spherical aggregates were formed on the glassceramic surfaces, which likely assigned to precipitation of a new layer of Ca-phosphate -almost HApcrystals. This suggestion was con rmed by EDX analysis of the glass-ceramic surfaces, where the analysis showed that the atomic percentages of Ca and P became lower than the theoretical starting percentages of those atoms. Table 2 presents the comparison between the atomic Ca/P ratio of the original composition (theoretical ratio) and the ratio calculated from EDX analysis. Where, the starting Ca/P ratio was 6.70, 12.94, and 4.62 became 1.91, 2.46, and 1.91 for WD2, WD3, and WD4, respectively. While, there was no phosphate in the original composition of the WD1 sample, but, the measured Ca/P ratio was 3.97. These results indicated that the Ca/P ratio of the newly formed layer after immersion in SBF was close to the ratio in HAp (Ca/P = 1.67). Meanwhile, WD2 and WD4 samples were possessed the same Ca/P (1.91) and they were the closest ratios to that of HAp.
The in vitro degradation and dissolution study of bioactive glass-ceramic materials in the physiological uids is important to give a simulated picture of the behavior of those materials inside the body. In this study, the change of pH, weight loss, and ion concentrations released in SBF was assessed. Figure 6a presents a change of pH of SBF solution as a function of time. It can be noted from the gure that the pH was abruptly increased to maximum values for all samples from 7.40 to 8.09, 7.95, 7.9, and 7.82 for WD1, WD2, WD3, and WD4, respectively, after 3 d of incubation. The pH value of the WD1 sample was signi cantly (p < 0.02) higher than other samples, whereas, WD4 was signi cant (p < 0.05) lower than the other samples at this time. After this increase was close to steady-state for WD2, WD3, and WD4 samples, while, the pH value of the solution incubated WD1 was slightly decreased to reach 7.83 at the end of incubation time, and it became signi cantly (p < 0.003) lower than the other glass-ceramic samples at this time. This difference in the pH values of incubated uid can be explained by the weight loss and ions released from the glass-ceramic samples in the immersing solution. Figure 6b shows the cumulative wt.
loss percent of WD1, WD2, WD3, and WD4 glass-ceramics as a function of time. As displayed in the gure, the weight loss % of all samples, mostly, were degraded linearly with time.
Furthermore, the WD1 sample showed the highest weight loss %, while, WD4 sample presented the lowest one. However, at the initial periods of incubation, the differences in weight loss % among all samples were insigni cant (p > 0.05), but, they became signi cant at the end of incubation time (p < 0.05). Accordingly, the introduction of Na 4 Ca(PO 3 ) 6 or Ca 5 (PO 4 ) 3 F into the glass-ceramic (sample WD2 and WD3, respectively) was decreased the degradation, and consequently, the reaction with SBF of the resulted glass-ceramics, and this effect become obvious when both phases were introduced in the glass composition. It has been reported that the uoride ions decrease the glass reactivity in the solution.
Accordingly, the addition of uoride, as a uorapatite apatite phase, into the glass composition was acted as a corrosion inhibitor [41] . And so, the existence of uorapatite crystals in the glass-ceramic was decreased its degradation in SBF.  Vero cells using the MTT assay [42] . Figure 7a shows the cell viability % of WD1, WD2, WD3, and WD4 was showed a lower degradation rate and higher released Ca 2+ ion concentration than WD2 and WD3 samples. This released high Ca 2+ ion was increased the calcium ion concentration in the cell culture medium which may result in cell death, as reported previously [43,44] .
The photos seen under an inverted microscope for the Vero cells after incubation with all samples for 24 h con rmed the results of MTT assay in which different morphological changes and loss of adhesion to a substratum were observed at a higher concentration for WD1, WD2, WD3, and WD4. Unlike, the control cells showed adherent growth and a regular polygonal shape, While at lower concentration the adherent ability increased with normal cell morphology and few round cells representing cellular outgrowth.
Studies were found that suitable divalent ions such as Ca 2+ and Mg 2+ ions can promote cell attachment, growth, and differentiation due to cell adhesion molecules on cell surfaces [45][46][47][48] .

Conclusion
In the current work, considerable combinations between bioactive multiphase glass-ceramic had been proposed depending on wollastonite-diopside/ sodium calcium phosphate and/or uorapatite aiming to synthesize bioactive glass-ceramic materials with talent mechanical, chemical, and biological properties required for biomedical application. The crystallization characteristics and resultant microstructure were described by DTA, XRD, and SEM. The biological activity was studied through in vitro test by simulated body uid (SBF), whereas, the cell viability test was estimated with the Vero cells. As well as the density and microhardness of the prepared materials were also evaluated. A bioactive phases including  6 ) in wollastonite-diopside glass-ceramic was obviously presented less cytotoxic effect than that included uorapatite (Ca 5 (PO 4 ) 3 F), and a combining of both crystalline phases was increased the cytotoxicity to the highest value. It could be concluded therefore that the produced glass-ceramics are promising candidate materials for biomedical application and can be used for natural bone replacements in human medicine.

Con icts of interest
The authors declare that they have no con ict of interest.