Synthesis of Bioactive Glasses SiO 2 -Al 2 O 3 -MgO-K 2 CO 3 -CaO-MgF 2 -CNT: Structural, Mechanical and Biological Properties

: Six glass compositions were synthesized using a melt quenching technique reinforced with different concentrations of carbon nanotubes (CNTs) from 0.1 to 0.7% in the glassy system SiO 2 -Al 2 O 3 -MgO-K 2 CO 3 -CaO-MgF 2 . Density and molar volumes were estimated by employing a liquid displacement method. In the present study, the reinforcement effect of CNTs was explained using several spectroscopic techniques i.e. Fourier transform infrared (FTIR), ultraviolet-visible (UV-Vis), Raman, and nuclear magnetic resonance (NMR) spectroscopy respectively. Based on Tauc plots of the UV-Vis spectra, the energy band gap was determined and their values decreased from 4.33 to 3.9 eV. Contact angle measurements were performed to check the wettability of the glasses. 29 Si-MAS-NMR spectroscopic study showed the random distribution of two dissimilar Ca 2+ and Mg 2+ ions within these glasses which lead to structural and topological frustration. To check the cell viability, MTT and alkaline phosphatase (ALP) assay were also performed. Owing to outstanding stability in various fluids like saline water, distilled water, and hydrochloric acid, the synthesized glasses exhibited functional activities with an adequate proliferation of rat calverail mechanical, tribological, and biological activities, the fabricated bioactive glassy material can be used for biomedical and multifunctional applications.

Lower the network connectivity of a glass, lower its glass transition temperature and enhance its reactivity along with solubility simultaneously. Consequently, network connectivity (NC) is very supportive for the fabrication of a new bioactive glassy compound. It has already been reported that the natural bone mainly consists water content of (10%), organic material (20%) and other mineral matter (70%) [34]. In case of certain injury, if the bone or its part is lost then bone grafting can be done as it serves for both biological as well as mechanical applications [35]. Hench and Wilson were reported two main features that should be presented in a bioglass: (1) a stable interface should be formed among the material and natural tissues and (2) the mechanical properties should be the same or better as the replaced tissues [36]. Hence, bioglass is suitable ceramic material for bone regeneration as the mineral ingredients of the bone is approximately the same as the chemical composition of the apatite layer [8].
Further, carbon is a naturally occurring material found in various forms and suitable for many technological applications due to its wide range of properties [37]. Carbon nanotubes (CNTs) have small dimensions, superior mechanical strength, aspect ratio and stiffness [35].
Nowadays, this material was used for various applications like aerospace, energy generation, electronics, etc [38]. Sumo Ijima et al. in 1991 have been reported the various forms of carbon structure as CNT and revealed its significant electrical conductivities, high aspect ratio, flexibility including adequate mechanical strength [39,40]. Moreover, its derivative is called graphene which has a single layer carbon nanosheet and exhibits excellent mechanical, thermal, optical and biological properties. Thus, CNT can be formed by rolling a graphene nanosheet in cylindrical shape (1-D) [41]. Webster and its co-workers had firstly reported in 2002 that nano-carbon material can be used for biomedical osteoblast differentiation [37]. CNT has an ordered structure material due to which has excellent antimicrobial and mechanical properties. Besides its excellent electrical conductivities, high aspect ratio, flexibility and adequate mechanical strength, CNT has several other applications for the fabrication of many devices such as optical devices, superconducting, molecular switches including biomedical devices [42]. So, owing to its large surface area, excellent biocompatibility and adequate mechanical strength, CNT can be utilized for several biomedical applications [43,44]. Researchers found to be enough interaction among CNTs and cells and reported that carbon nanostructures of nanofibers, diamond, and fullerene, etc. can also be used for biomedical applications [45]. Meng et al. were studied a ternary ceramic composite system HAp-ZrO2-CNT by using a hot-press sintering technique. In this study, they reported good biocompatibility and enhanced flexural strength and fracture toughness by ~126% and ~124%, respectively than that of unmodified HAp [46]. It has been reported that the addition of multi walled carbon nanotubes (MWCNTs) to the 3-D bioglass, enhanced the compressive strength and bioactivity significantly [47,48]. Further, Armentano et al. have been studied in vitro biomineralization processes on human alveolar bone-derived cells using carboxylated and fluorinated single wall nanotube (SWNT) films. In this study, they demonstrated an encouraging biocompatibility and gene expression with cell phenotypes that can be regulated easily [49].
However, Zanello et al. have compared the osteoblast cell formation and bone creation using functionalized and non-functionalized CNTs [50]. Khang et al. had studied the articular cartilage tissue regeneration using a carbon nanotube/polycarbonate urethane film [51]. Further, Yu Xiao et al. have studied the functionalized MWCNTs via in situ deposition of HAp to enhance their hydrophilicity as well as biocompatibility [52]. Ya-Ping Guo et al. synthesized carbonated HAp/CNT coatings with mesoporous structures, and then macro powders and CNTs were electrophoretically deposited on a substrate using a microwave irradiation technique [53]. The achieved coatings comprising mesoporous structure (pore sizes ~3.9 nm), and reveal an excellent 6 in vitro bone developing activity. In this research paper our aim is to explore the structural networking between different glass former and network modifier due to the variation of CNT concentration on structural, mechanical, and biological properties of the synthesized glassy system SiO2-Al2O3-MgO-K2CO3-CaO-MgF2-CNT. The glass sample GCNT0.7 can be used for various biomedical applications.

Glass sample preparation
Various glass compositions were melted from their raw materials especially using high purity chemicals i.e. SiO2, Al2O3, MgO, K2CO3, CaO, MgF2 (Merck USA) and CNT (Sigma Aldrich, USA) in batches of 20 to 40 gm in a highly pure alumina crucible. The total mole amount of the glass composition is set to be 100% for superior comparison. The batch compositions, density, molar volume, optical band gap and contact angle of the base glass samples are listed in Table 1. All the glasses were reinforced with different content of CNT (0.1 -0.7 mol %) and melts were kept for 2 h @ 1450 o C to improve its homogeneity. The glass melt was cast into a brass block mould (2x4 cm); pressed by another brass plate and then quickly replaced into a preheated programmable electric muffle furnace for annealing @ 450 o C for 4 h to release residual internal stress and then annealed glass samples taken out from this furnace at room temperature.
Eventually, the prepared glass samples were cut and polished for further characterizations.

XRD measurements
For XRD measurements, glass samples were ground with the help of a mortar pestle to achieve a fine powder. Further, XRD patterns have been recorded on powdered glass samples using a Rigaku Miniflex-II X-ray diffractometer employing Cu-Kα radiation comprising the wavelength of ~1.54056 Å for the confirmation of amorphous nature over a 2Ɵ range from 20 o to 80 o and X-ray tube was operated at 40 kV and 40 mA.

Density and molar volume
Density of glass samples depends upon varied concentrations of CNT as well as on compositional changes. Thus, the density of all glasses was calculated using Archimedes' principle [54]. Doubled distilled water was utilized and density was measured using a weighing digital balance possessing accuracy up to 0.0001 mg. The measurements of density were carried out thrice for each glass sample. The average values of the density measurements are shown in Table 1. The maximum random variation in the density measurements was 0.001 g/cm 3 . Therefore, the densities (ρ) and molar volumes (VM) of the CNT doped glass samples were calculated using the following formulae [55,56] (1) where, ρ is the density (g/cm 3 ), ρw is the density of distilled water (1 g/cm 3 ), W1 = weight of empty specific gravity bottle (g), W2 = weight of specific gravity bottle with sample (g), W3 = weight of specific gravity bottle with sample and distilled water (g), and W4 = weight of specific gravity bottle with distilled water (g), and: (2) where, Mi is the molecular weight of i th factor and Xi is the molar portion of the i th component.

Fourier Transform Infrared (FTIR) Spectroscopy
FTIR absorption spectra of the annealed glass samples were obtained at room temperature within the wavenumber range of 4000 -600 cm -1 using a JASCO FT/IR-5300. Samples were grinded into fine powder (<180 μm) and then mixed with KBr powder in the ratio of 1:9 precisely.
Subsequently, the powdered mixtures were pressed uniaxially using a hydraulic press machine to form clear homogeneous discs. The prepared discs of the various glass samples were dried at 150 o C for 1 h in an oven to remove the moisture.

UV-vis spectroscopy
UV-vis spectroscopy of the all glass samples was also carried out using UV-Vis   29 Si magic angle spinning (MAS) Solid State NMR experiments were performed on JEOL ECX 500 NMR Spectrometer (housed at Dr. Harisingh Gour Central University, Sagar, India).

Nuclear Magnetic Resonance (NMR) Measurements
The spectrometer was decorated by a 3.2 mm JEOL double resonance MAS probe. 29 Si MAS-NMR tests were carried out using a sample spinning speed of 10 kHz. Reference for 29 Si chemical shifts measurements were done by using 2,2-dimethul-2-silapentane-5-sulfonate sodium salt (DSS). For getting 29 Si MAS NMR spectra a 90 0 pulse width of 3 s was also used by a relaxation delay of 30 s. A drop of deionized water is carefully placed on the polished surface of the glass sample using a hypodermic syringe needle. The photographs of drops were taken quickly by a high-resolution digital camera just after putting the drop of water on the sample. The drop volume was controlled within the range from 10~20 µL [59]. Contact angles of the all polished glass samples were measured by using "ImageJ" software and their values are listed in Table 1.

Dynamical Mechanical Analysis (DMA) Measurements
DMA measurements were performed to determine storage modulus and strain properties of three selected glass samples. DMA accepts several different forms of the samples. Equipment available for this study required small rectangular samples approximately dimensions of 1.3 mm x 5 mm x 8 mm. Dimensions of all sample were measured by DMA and further confirmed using a digital caliper. Dynamic mechanical testing was carried out in compression mode using TA Instruments Q800 DMA at room temperature. Parameters were chosen based on review of literature and discussions with experienced DMA researchers.

Scanning Electron Microscopy
The surface morphology of the fine powdered glass samples was observed using a scanning electron microscope (model: JEOL JSM-6400). Glass samples were ground to make the fine powder using a mortar and pestle. For the recording of SEM images, the tiny amount of fine powdered glass samples was taken on to a copper stub and then coated with Ag-Pd thin films using a sputtering machine 'JEOL, JEC-3000FC Auto Fine Coater' to avoid the charge build-up on them.
Finally, the achieved samples along with stubs were then mounted in a sample holder by a conducting carbon tape and hence, SEM images were recorded at desired magnifications.

MTT Assay for cell viability
To evaluate the capability of the osteoblast cells in presence of the fine powdered samples GCNT0.1 and GCNT0.7 (containing the lowest and higher amount of CNT), rat calverail osteoblast (ROB) cells were extracted from mouse pups calvaria culture and on 90% confluency, cells were reseeded in 96 well plates. After 24 h, different concentrations of these samples were given to the cells and then incubated for 48 h at 37 °C in a moistened atmosphere containing 5% CO2 and 95% air. After the incubation period, MTT reagent (5 mg/ml) was added and the plate was incubated further at 37 °C for 3 -4 h to form the formazan crystals. After the end of incubation, dimethyl sulfoxide (DMSO) was added to dissolve the crystals and optical density (OD) was estimated at 570 nm [60].

Alkaline phosphatase (ALP) assay for osteoblast differentiation
For evaluation of the osteogenic activity of this compound, alkaline phosphatases assay was being carried out on the osteoblast cells extracted from mouse pups calvarial tissue. Osteoblast cells at 90% confluence were trypsinized and reseeded in the 96 well plate (2×10 3 cells per well) in 10% alpha-MEM medium. After 24 h, cells were treated with varying contents of glass samples GCNT0.1 and GCNT0.7 in osteoblast differential medium (α-MEM supplemented with 5% FBS, 10 nM b-glycerophosphate, 50 mg/ml ascorbic acid and 1% penicillin/streptomycin) and plates were incubated for 48 h at 37°C in a moistened atmosphere of 5% CO2 and 95% air. After incubation, entire ALP activity was studied using the substrate p-nitrophenylphosphate (PNPP) and then quantified calorimetrically at 405 nm [61,62].

Stability in saline water, distilled water and hydrochloric acid (HCL)
To check the stability in physiological conditions, weight loss tests were also performed on the fabricated glass samples at room temperature using different solvent i.e. the saline water (Sodium Chloride Injection IP (0.9% w/v) made by Tara Bioscience Pvt. Ltd. (India), distilled water and hydrochloric acid (MERCK). Glass samples were then dipped into this solution for 8 h, 8h and 3h soaking time respectively. Moreover, the weight differences were measured before and after the treatment with saline water, distilled water and hydrochloric acid respectively by a weighing digital balance that contained an accuracy up to 0.0001 mg and thus obtained data was analyzed successfully.

XRD analysis
XRD patterns of the glass samples GCNT0.1, GCNT0.2, GCNT0.3, GCNT0.4, GCNT0.5 and GCNT0.7 are shown in Figure 1 respectively. These XRD patterns depicted a broad diffuse peak instead of sharp crystalline peak which endorsing a typical long-range structural disorder of an amorphous material. The XRD patterns of these glass samples showed almost similar behaviors and confirmed the pure unstructured nature of the fabricated glasses.

Density and molar volume analysis of the glass samples
Density of the fabricated annealed glass samples was calculated and listed in Table 1. From ions are arbitrarily distributed in the glassy matrix. Hence, this leads to the creation of non-bonding oxygens (NBOs) which is attributed to an increase in the density of the fabricated glass samples but herein, the amount of CaO is decreasing as increasing the reinforcement of the CNTs. Therefore, density of the investigated glasses decreases with the rising content of CNT.
Density of the glasses against reinforcement concentration of CNT depicted in Figure 2
For a particular peak corresponding to their different wavenumber starting from higher to lower wavenumber side including their respective band assignments are presented in Table 2 Further, broad transmission bands were observed within the wavenumber range of 3600-3750 cm -1 which are assigned due to the O-H group vibrations [64][65][66]. However, the transmission bands in the wavenumber region from 2911-2925 cm -1 are owing to the growth of hydrogen bonding in the glassy matrix [67,68]. Two distinct peaks were also observed at 2920 and 2923 cm -1 which are assigned due to the formation of C-H stretching vibrations; probably originated from the surface of the CNT and nicely dispersed in a glassy matrix [69,70]. The broad transmission band 6 lies in the region from 2346 -2354 cm -1 and recognized to the formation of O-H bond which probably due to the hygroscopic nature of K2CO3 and CaO. [67]. Moreover, all the O-H bonding groups observed at various wavenumbers that were formed at the non-bridging oxygen (NBO) sites in the glassy matrix [67]. Further, the transmission band lies in the wavenumber 1690-1697 cm -1 revealing the stretching vibrations of hydroxyl groups and the existence of water molecules [71].
However, the intense and sharp band lies within wavenumber range from 1525-1530 cm -1 which are attributed to the stretching vibrations of C=O group on the exterior surface of the CNT [69].
Additionally, as rising the reinforcement of CNT, the centers of the bands within their wavenumber range are shifted towards low wavenumber side. It might be owing to the increase in the ratio of oxygen to CaO atoms that are associated with NBOs [72].
The observed bands in the region 1200-1250 cm -1 are attributed to the stretching vibrations of O-Si-O in the glassy network [73]. The various counts of NBO in the SiO4 tetrahedral were observed due to the wavenumbers in the region 840-1250 cm -1 [74]. The band corresponds to the wavenumber 901 cm -1 representing the elongating vibrations of the Al-O --alumina-silica-oxygen bridge [75][76][77][78]. However, the band noticed at 915 cm -1 revealed the vibrations of Si-O - [75,76]. A low intensity transmission band was also observed at ~840 cm -1 that assigned to the bending vibrations of H-Si-O [73]. Moreover, it is gradually lifted near to the higher wavenumber side as increasing the content of CNT. As the presence of Al in these glasses exist, a transmission band lies in the wavenumber range of 700 -750 cm -1 and shows the tetrahedral coordinate of Al as center of gravity at ~695 cm -1 [79,80]. The formation of pseudo wollastonite confirmed the wavenumber 714 cm -1 [75].

UV-Vis spectroscopic analysis
The UV-Vis absorption spectroscopy analysis helped to observe the absorption of UV  Table 1. and these values were observed within the wavelength region of 200-1000 nm. The oxygen bonding strength is essential to calculate the distinct values of optical band gap in the glassy matrix [79]. As the reinforcement of CNT gradually increases from 0.1-0.7 and CaO decreases simultaneously, the values of band gap don't show any discrete sequence. The glass composition similar to this composition a slight change in the dopant content, change its optical band gap values from 3.59-3.65 eV [79]. However, the observed band gap in the present glassy system SiO2-Al2O3-MgO-K2CO3-CaO-MgF2-CNT was found to be more i.e. 3.89-4.34 eV as compared to the reported values [79]. As no sharp absorbance peak is observed in Figure 4 (a), it can be concluded that the glass samples are UV-inactive glasses due to a higher optical band gap.

Raman spectroscopic analysis
In order to give more insight into the structure of glasses, the Raman spectra of various bioactive glass samples GCNT0.1, GCNT 0.2, GCNT0.3, GCNT0.4, GCNT0.5 and GCNT0.7 were recorded and they containing seven contrast peaks in the range of 30 -2000 cm -1 . All recorded spectra are shown in Figure 5 (a-f) respectively and their respective band assignments associated with their peak positions are enlisted in Table 3. The Raman bands in the region 1323-1326 cm -1 correspond to the occurrence of some defects in the CNT structures which are probably due to the melt-quenching during the synthesis of glasses [85]. However, the peaks occur in the range of 1200-1300 cm -1 (broad bands) are associated with the asymmetric vibrations of SiO4 tetrahedral molecules [86,87]. The peak at 650 cm -1 is due to the bending vibrations and revealed the bridging oxygen (BO) of SiO4 molecules [86,88]. The bands in the region of 600-700 cm -1 are assigned due to the bending or stretching vibrations of Si-O-Si bonds [89]. The bands at different wavenumbers i.e. 341, 344 and 348 cm -1 are assigned due to the mixed stretching and bending vibrations of Si-O-Si bonds [86,90]. Further, the Raman bands reveal the wavenumbers 173, 178 and 180 cm -1 which are associated with the Si-O bonds of stretching vibrations [91]. An intense band at ~141 cm -1 is occurred due to the enlarging vibrations of O-H bond [92]. Hence, FTIR and Raman spectroscopic results revealed that glasses were formed through glass forming networks of silicate, alumina and network modifiers as cations of alkaline earth atoms. The silicate networks are modified by using several oxides such as Al2O3, MgO, K2CO3, CaO, MgF2 in the glassy matrix.

Nuclear magnetic resonance (NMR) study
Solid State Nuclear Magnetic Resonance (SSNMR) spectroscopy is an indispensable microscopic tool to discriminate the same type of atoms in different chemical environments,  (1) , Q (2) , Q (3) , and Q (4) . The silicon tetrahedral without the presence of another tetrahedral silicon neighbor is denoted by Q (0) . When one end of a silicon tetrahedral is connected with another silicon tetrahedral via an oxygen bridge, then it is denoted by Q (1) .
Whereas, Si-tetrahedral connected with two and three neighboring silicon tetrahedral via oxygen are designated as Q (2) and Q (3) respectively. When all four corners oxygen atoms of the silicon tetrahedral are shared by the other four adjacent silicon tetrahedral, then this type of fully crosslinked framework is designated as Q (4) And 2 (0) = 2 (1) + ( ) 2 (5) Where ( ) 2 represents oxygen, which is not assured by the silicon tetrahedra and these are referred as free-oxygen. However, the equilibrium constant is represented by the following equation:  indicates the distribution of ( ) species within the silicate-network is random [32,97,100,101].
Within the silicate network, Ca 2+ and Mg 2+ ions are randomly distributed. This leads to the generation of non-bonding oxygens within the network. As the ionic radius of Ca 2+ ions (1.00-1.12

Contact angle analysis
Contact angle study showed the hydrophilic nature and chemical diversity of the solid surface of CNT doped glass samples. It is reported in the literature that if the value of contact angle less than 90• then the material is hydrophilic while greater than 90•, then the material nature is hydrophobic [103,104]. Hence, the hydrophilic nature of CNT doped glassy samples was determined using contact angle measurements. The glass sample GCNT0.1 reveals less contact angle, 18.14 o that is advantageous for hydrophilic nature of glassy materials whereas glass sample GCNT0.7 (0.7 % of CNT) shows hydrophobic nature [105]. The contact angle of the all synthesized bioactive glasses GCNT0.1, GCNT0.3, GCNT0.4, GCNT0.5 and GCNT0.7 were calculated and depicted in Figure 7 (a-f) respectively. Moreover, the calculated values of contact angles are determined by using 'ImageJ' software and are enlisted in Table 1 including the highest value of density, 2.98 gm/cc.
As increasing doping concentration of CNT, the contact angle has to be increased up to 77.8 • which is responsible to a decrease in hydrophilic nature by the creation of hydrophobic layers on the surface of these bioactive glass samples. The variation of contact angle with increasing doping percentage of carbon nano tube is revealed in Figure 7 (g). Consequently, it reveals that the value of contact angle is directly proportional to the increasing doping percentage of CNT.
Additionally, the hydrophilic behavior is mainly owing to the tiny pores and rough surface of the glass sample which is attributed for the lowest density of the GCNT0.1. Therefore, the lowest value of density and contact angle may offer the transportation of the body fluid and probably enhance the tissue regeneration on the surface of this synthesized bioactive glass sample GCNT0.1 [106].

Mechanical properties of the glasses
The study of mechanical properties with various parameters such as storage modulus, length, stiffness, and loss modulus were plotted against strain for different glass samples i.e. Herein, the higher value of storage modulus is a signature by the glass sample to store deformation energy in an elastic limit which is directly associated to the extent of cross-linking within the glassy matrix. Moreover, the higher degree of cross-linking the greater the storage modulus. So, it is concluded that as increasing the content of CNT the mechanical strength and resists deformation in response to an applied force of this glass sample object is enhanced significantly.
These fine particles are well interrelated to each other and appearing similar to a layered material.
Further as increasing the doping amount from 0.1 to 0.2 the morphology of the glass particle was completely changed including their shape and size (Figure 9 b and c). The presence of the tiny CNT was clearly observed in these SEM images. Moreover, as increasing the doping concentration of CNT in these glassy compositions, the various CNT can be seen easily. The magnified portion of the SEM images are shown as the insets of Figure 9 (d and e). The average length to diameter ratio was found to be the order of ~450 to ~50 nm. The CNTs are well dispersed into the glassy matrix and their shape and size are probably enhanced in comparison to their original shape and size. The enhancement in their size might be owing to the dispersion in the glassy matrix during the synthesis of the glasses. It has already been reported by Aria et al. that as-grown CNT are hydrophobic in nature, and they become more super hydrophobic when passing through vacuum pyrolysis treatment [107]. Therefore, the minor reinforcement of CNT into a glassy matrix can tune the hydrophobicity significantly. Moreover, as the glass sample concentration was augmented, the cell viability repeatedly decreased in comparison to the control. So, based on these results, the glass sample GCNT0.7 was found to be better than sample GCNT0.1. Hence, these outcomes verify that the fabricated glass sample GCNT0.7, reinforced with 0.7 % CNT is highly biocompatible, nontoxic in nature, and can be used for various potential biomedical applications.

Effect of GCNT0.1 and GCNT0.7 on osteoblast differentiation
For investigating osteogenic potential of the bioactive glass samples GCNT0.1 and GCNT0.7, ALP assay was carried out where the production of alkaline phosphatase serves as a marker of osteoblast differentiation. Figure 10

) and assignments of Fourier transform infrared spectra of investigated various glass samples in the glassy system
SiO2-Al2O3-MgO-K2CO3-CaO-MgF2-CNT