BIOACTIVITY EVALUATION OF 60%SiO2-36%CaO-4%P2O5GLASSES MICRO AND NANO SIZED Agda Aline Rocha de Oliveira, Sandhra Maria de Carvalho, and Marivalda de Magalhães Pereira
Departamento de Engenharia Metalúrgica e de Materiais – Universidade Federal de Minas Gerais, Brasil
Abstract. Bioactive glasses (BG) bond to bone by forming a hydroxiapatite (HA) layer in vivo. In solution, the surface of BG undergoes a time-dependent modification. The formation of a HA layer in vitro on a material surface is believed to indicate its bioactive potential in vivo. Parameters such as surface charge, composition, structure, and morphology will be important in the formation of the Ca-P layer, as well as in the interaction between the material surface and the surrounding medium, proteins and cells. The BG surface zeta potential variations in an electrolyte solution correspond to, and may directly influence, Ca-P layer formation. In this study the bioactivity of BG particles micro and nano sized was investigated by the time-dependent variations in zeta potential of BG immersed in SBF solution. FTIR, XRD and SEM analysis were used to confirm the formation of HA layer. The cell viability by MTT assay was used to evaluate the behavior of particles in direct contact with osteoblast cells. The zeta potential variations occurred faster and had higher variations for nanoparticles (NP). This result suggests that the kinetic of HA formation on BG particles is influenced by the particle size. The NP presented significant increase in cell viability when compared to microparticles (MP). These results support the hypothesis that BGNP are more bioactive than BGMP. Keywords: Bioactive glass nanoparticles, Bioactive glass microparticles, Hydroxyapatite, Cell viability
Bioactive glasses (BGs) bond to bones by forming an in vivo hydroxiapatite (HA) layer. In solutions, the surface of BG undergoes a time-dependent modification. The formation of HA layers in vitro on a material surface is believed to indicate its bioactive potential in vivo. Clearly, a bioactive behavior is an interface-driven phenomenon. Parameters such as surface charge, composition, structure, and morphology will be important in the formation of the Ca/P layer as well as in the interaction between the material surface and the surrounding medium, proteins, and cells [1, 2].
The HA layer is spontaneously formed on the surface of glasses in the CaO–SiO2–P2O5 system after exposure to simulated body fluid (SBF) [1, 3]. Apatite is formed preferentially on a glass surface mainly composed of CaO and SiO2 due to the fact that the Ca2+ ions released from the glass increases the degree of supersaturation with respect to the apatite of the surrounding fluid. In addition, the Si–OH groups of the hydrated silica gel formed on the surface induce a heterogeneous nucleation of apatite. These crystals grow by consuming calcium and phosphate ions from the body fluid and those that migrate from the bulk to the surface of the glass .
Compared to microparticles, bioactive glass nanoparticles (BGNPs) have advantages in bone repair and regeneration, with the decrease in grain size promoting an increase in cellular adhesion, enhanced osteoblast proliferation and differentiation, and an increase in the biomineralization process . The use of nanosized particles may provide a means for a more rapid release of Ca and P  and low negative zeta potential in biological medium, which has important effects in vivo  and promotes cell attachment and proliferation .
The present work aims to study the effect of BG particle size on their in vitro bioactivity, investigated by the immersion of SiO2-CaO-P2O5 glasses, both micro (38-150 m) and nano (83-91 nm) sized, in SBF. The time-dependent variations in BG zeta potentials, as well as FTIR, XRD, SEM, and EDS analyses, were used to confirm the formation of HA layers. Resazurine in vitro assays were used to determine osteoblast cell behavior in direct contact with micro and nano-sized particles of BGs.
MATERIALS AND METHODS
2.1Sample preparation The reagents used in the synthesis of BG were 98% Tetraethyl Orthosilicate (TEOS) and 99% Triethyl Phosphate (TEP) (Sigma-Aldrich, USA), Methanol and Calcium Nitrate (Ca(NO3)2.4H2O) (Synth), Nitric Acid (HNO3) and 33% Ammonium Hydroxide solution (NH4OH) (Merck). Bioactive glass microparticles (BGMPs), with the nominal composition wt % of 60% SiO2, 36% CaO, and 4% P2O5 were prepared as described in previous works  by the acid hydrolysis of the alcoxides. The solids were ground and separated by sieving in the range of 38-150 m.
The BGNPs were also prepared as described previously  by the mixing of methanol, ammonium hydroxide and water, followed by the addition (dropwise) of TEOS and TEP. The sol was mechanically stirred for 48 h and then placed in an oven at 50°C until the ammonium had been completely evaporated. After, the sol was filtered in a 0.22 m Milipore, and Ca(NO3)2.4H2O was dissolved in the sol and mixed for 24 h. The nanoparticles formed were separated by subsequent filtrations in a 0.22 and 0.11 m Milipore and submitted to freeze drying. The powders obtained were thermally treated at 200ºC for 40 min, at a heating rate of 1ºC/min. At the end of the process, well-dispersed bioactive glass nanoparticles, with an average diameter of (87 ± 5) nm, were obtained without grinding or sieving.
2.2 In vitro bioactivity evaluation The formation of a surface HA layer on the particles was evaluated by means of immersion in SBF solution at time periods of 24, 48, and 72 hours. Surface structural observation was carried out after the immersed samples were removed from the SBF, washed three times with de-ionized water, and dried in an air circulation drying oven. The chemical structure of the BG particles was analyzed by Perkin Elmer 100 Spectrum Fourier transform infrared (FTIR) spectra in the mid-infrared range from 550 to 4000 cm-1 in ATR mode using a Zinc Selenide prism. Samples for FTIR analysis were diluted and ground in KBr with a sample to KBr dilution ratio of 1:100. X-ray Diffraction (XRD) spectra were collected on a Philips PW1700 series automated powder diffractometer using Cu K radiation at 40 KV/40 mA. Data was collected between 4.05º and 89.95º with a step of 0.06° and a time of 1.5 seconds to identify any crystallization of the particles.
The morphology and atomic composition of the BG particles were observed by a Tecnai G220 FEI Scanning Electron Microscope (SEM) equipped with energy-dispersive X-ray (EDS).
Zeta potential data were measured using the Zetasizer 3000 HS Data Type 1256. BG particles were immersed in SBF at a weight-to-solution volume ratio of 0.1 mg/ml. Each sample was incubated at 37°C for 0, 6, and 24 hours, and 3 and 7 days. Each data point for time-dependent changes in BG zeta potential in SBF represents the mean values calculated based on the average of nine histograms (n = 9).
2.3 Biological response The biological tests were performed according to ISO 10993 (Part 5 - tests for in vitro cytotoxicity). Osteoblast viability was evaluated by Resazurine assay, based on a commercially available resazurin solution (Sigma Aldrich USA). A culture medium with BG particles was added to the osteoblast cells (103 cell/ mm3) cultured in 96-well plates in DMEM containing 10% FBS, penicillin G sodium (10 units/ml), and streptomycin sulfate (10 mg/ ml). After incubation times, the medium was removed and 180 l of new culture medium supplemented with 10% of FBS was added to each well, and 20 l of Resazurine (0.01 mg/ml) was added to each well and incubated for 24 hours at 37°C under 5% CO2. After, the absorbance was measured using an ADAP 1.6 spectrophotometer at 570nm and 595 dual filter. Phosphate-buffered saline (PBS) was used as a positive control, while polyethylene was used as a negative control.
RESULTS AND DISCUSSION
3.1 Zeta potential measurements Zeta potential changes in BG immersed in SBF are shown in Fig. 1. The data are indicative of a dynamic surface, as two sign reversals in the surface zeta potential were measured at different time periods for micro and nano-sized BG particles . The initial BG surface was negative in charge, with an average zeta potential of approximately -13 mV. The surface potential increased with an increase in soaking time up to a maximum positive value. For the NP, the surface became positive after 6 hours, reaching an average zeta potential of +21.0 mV. For the MP, the positive potential of +4.5 mV was only observed after 3 days in SBF. After, these potentials decreased with an increase in soaking time and once again attained negative values, remaining negative after 7 days of immersion. The zeta potential values reached at this time were -6.9 and -2.1 mV for NP and MP, respectively.
Figure 1: Zeta Potential of BG micro and nano sized particles immersed in SBF solution in different periods of time. Each data point represents the mean of values calculated based on the average of nine histograms (n = 9). The error bar represents the standard deviation of the mean.
The zeta potential variations occurred faster and with higher variations in BGNP, suggesting that the kinetics of HA formation on particles is in fact influenced by particle size, most likely due to the significant difference in surface area, when micro (80 cm3/g) and nano (530 cm3/g)-sized particles are compared, as suggested in previous works [10, 11]. It is consistent to assume that the particles with greater surface areas contain a larger number of -OH groups, which subsequently improves the catalytic effect of the Si-OH groups for HA nucleation. With the progress of chemical reactions, the surface becomes completely covered by the formed apatite and no longer contains the Si-OH available to generate new HA nodule precipitation. Thus, the reaction kinetics tend to stabilize. These results indicate that NP increases the kinetics of HA nucleation at the early stage of SBF soaking, but this effect decreases over time and the differences in growth of HA on NP and MP surface, respectively, tend to be minimized after prolonged soaking times.
3.2 Structural characterization After soaking in SBF for seven days, the BG particles were filtered and the obtained powders were analyzed by SEM. Figures 2 and 3 show the HA layer formed on BG micro and nano-sized particles, respectively. In both particles, the surface was covered by HA layer nodules. In the MP this nodules merged to form larger particles that populated the glass surface and can be easily identified in the SEM image (Figs 2 (c)). In the NP, it is quite difficult to conclude whether or not these HA nodules were formed or if the HA layer only covered the surface of the glass, significantly increasing particle sizes (Fig. 3 (c)).
Figure 2: SEM images of BGMP after 7 days of immersion in SBF. Images (b) and (c) are magnifications of image (a).
Figure 3: SEM images of BGNP after 7 days of immersion in SBF. Images (b) and (c) are magnifications of image (a).
FTIR was used to confirm the formation of the Ca/P layer on BG particles by detecting characteristic vibration modes of the P-O and P=O bonds after 7 days of immersion in SBF (Fig. 4). The changes that occurred can be monitored by the appearance of several absorption bands in the spectra of samples immersed in SBF. The absorption band at 960 cm-1 is assigned to the P–O bond. The C-O stretching vibration appeared between 890 and 800 cm−1 suggesting the presence of carbonated Ca/P on the BG particles after SBF immersion. The divided P-O bending vibration peak between 600 and 500 cm−1, an indication of the development of a crystalline Ca/P layer, can clearly be seen in both BG spectra. The band at 470 cm-1 is assigned to the asymmetric stretching of PO43- groups [1, 3, 10, 11].
Figure 4: FTIR spectra of BG micro and nanoparticles after soaking in SBF for 7 days.
Fig. 5 shows the XRD patterns obtained from the surfaces of the particles after soaking in SBF. Crystalline peaks appear in the XRD patterns of BG after 7 days in SBF. The peaks at 2 value of 25.8 and 31.7 are assigned to (0 0 2) and (2 1 1) reflections of crystalline HA [1, 3, 10, 11].Apatite formation on the surface of the glass particles in the SBF is governed by the chemical reaction of the surface of the matrix with the fluid. The intensity of two major reflections, (0 0 2) and (2 1 1), increase with a rise in the concentration of Ca2+ and PO43- ions on the surface of the glass immersed in SBF for a number of days. The peaks assigned to these reflections are slightly better defined in nano-sized particles, which suggests that NPs present a higher crystalline quality in HA crystals than do MPs.
Figure 5: XRD patterns of BG micro and nano sized after soaking in SBF for 7 days.
3.3 Biological tests The Resazurine assay (Fig. 6) showed an osteoblast proliferation of 27% and 16% higher in the presence of NPs, when compared to MPs, at periods of 6 and 24 hours, respectively. At 72 hours no significant difference could be found between the particles. The results indicated that both MP and NP are biocompatible. The higher cell viability results for NPs at the early stage of incubation can be attributed to the more rapid release of Ca and P, which have a positive effect on the bioactivity of these particles. These results support the hypothesis that BGNP are more bioactive and biocompatible than BGMP. It is consistent to assume that the increased HA nucleation on the nanoparticle surfaces is responsible for the higher stimulating effect on osteoblast at the early stages of incubation. This effect decreased over time, and the differences in osteoblast proliferation when in contact with nano and microparticles tends to be minimized after prolonged soaking .
Figure 6: Cell viability of BG micro and nano particles after 6, 24 and 72 hours of culture measured by resazurine assay. * Represents significant difference compared to control and º represents significant difference between MP and NP at a significance level of 0.05 %.
The present study examined the relationship between time-dependent variations in surface charge and the corresponding surface composition, structure, and morphology changes in BG micro and nano-sized particles. BG zeta potential variations proved to be directly related to surface Ca/P layer formation. The zeta potential variations occurred faster and presented higher variations for BGNP, suggesting that the kinetics of HA formation in particles is in fact influenced by particle size, most likely due to the significant difference in surface area, when nano and microparticles are compared. The increased HA nucleation on the nanoparticle surfaces may well be responsible for the higher cell viabilit in the osteoblast cultures at the early stages of incubation. These results support the hypothesis that BGNP are more bioactive and biocompatible than BGMP.
ACKNOWLEDGMENTS The authors gratefully acknowledge financial support from CNPq, CAPES, and FAPEMIG/Brazil and the technical support from the Center of Microscopy- UFMG/Brazil and the technical support from the Center of Microscopy- UFMG/Brazil.
REFERENCES 1. Kokubo T, Kushitani H, Ohtsuki C, Sakka S. Chemical reaction of bioactive glass and glass–ceramics with a simulated body fluid. J Mater Sci: Mater Med (1992) 3:79–83
2. Ueshima M, Nakamura S, Yamashita K. Huge, Millicoulomb charge storage in ceramic hydroxyapatite by biomedical electric polarization. Adv Mater (2002)14:591–4.
3. Marlene B Coelho and Marivalda M Pereira. Sol-gel synthesis of bioactive glass scaffolds for tissue engineering: effect of surfactant type and concentration. Journal of Biomedical Materials Research Part B: Applied Biomaterials (2005) 75B (2) 451-456.
4. Kokubo T, Kushitani H, Ohtsuki C, Sakka S. Chemical reaction of bioactive glass and glass–ceramics with a simulated body fluid. J Mater Sci: Mater Med (1992) 3:79–83.
5. Liu H, Webster TJ. Nanomedicine for implants: a review of studies and necessary experimental tools. Biomaterials (2007) 28:354–69.
6. Superb K. Misra, Dirk Mohn, Tobias J. Brunner, Wendelin J. Stark, Sheryl E. Philip, Ipsita Roy, Vehid Salih, Jonathan C. Knowles, Aldo R. Boccaccini. Comparison of nanoscale and microscale bioactive glass on the properties of P(3HB)/Bioglass composites. Biomaterials 29 (2008) 1750-1761.
7. R. Smeets, A. Kolk, M. Gerressen, O. Driemel, O. Maciejewski, B. Hermanns-Sachweh, D. Riediger, J.M. Stein, A new biphasic osteoinductive calcium composite material with a negative zeta potential for bone augmentation, Head Face Med. 5 (2009) 13.
8. J. Cooper, J.A. Hunt, The significance of zeta potential in osteogenesis, in: Transactions of the 31st Annual Meeting for Biomaterials, Society for Biomaterials, (2006), p. 592.
9. Agda Aline Rocha de Oliveira, Sandhra Maria de Carvalho, Maria de Fátima Leita, Rodrigo Lambert Oréfice, Marivalda de Magalhães Pereira. Development of biodegradable polyurethane and bioactive glass nanoparticles scaffolds for bone tissue engineering applications (2012). DOI: 10.1002/jbm.b.32710.
10. Helen H. Lu, Solomon R. Pollack, Paul Ducheyne. Temporal zeta potential variations of 45S5 bioactive glass immersed in an electrolyte solution. J Biomed Mater Res. (2000) 51 (1):80-7.
11. Ali Doostmohammadi, Ahmad Monshi, Rasoul Salehi, Mohammad Hossein Fathi, Zahra Golniya, Alma. U. Daniels. Bioactive glass nanoparticles with negative zeta potential. Ceramics International 37 (2011) 2311–2316.