- Email: [email protected]
Gallium-containing phosphosilicate glasses: Functionalization and in-vitro bioactivity
Materials Science and Engineering C 33 (2013) 3190–3196
Contents lists available at SciVerse ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Gallium-containing phosphosilicate glasses: Functionalization and in-vitro bioactivity Gigliola Lusvardi, Gianluca Malavasi, Ledi Menabue ⁎, Shruti Shruti Department of Chemical and Geological Sciences, University of Modena and Reggio Emilia, Via G. Campi 183, 41125 Modena, Italy
a r t i c l e
i n f o
Article history: Received 8 October 2012 Received in revised form 4 March 2013 Accepted 28 March 2013 Available online 6 April 2013 Keywords: Glasses functionalization Phospho-silicate glasses Gallium oxide Bioactivity test
a b s t r a c t A gallium containing glass 45.7SiO2·24.1Na2O·26.6CaO·2.6P2O5·1.0Ga2O3 (referred to as “Ga1.0”) and a parent Ga-free glass 46.2SiO2·24.3Na2O·26.9CaO·2.6P2O5 (hereinafter represented as “H”), corresponding to Bioglass® 45S5, were functionalized with Tetraethoxysilane (TEOS) and (3-Aminopropyl)triethoxysilane (APTS) in order to improve their ability to bond with biomolecules, such as drugs, proteins, and peptides. Functionalization with TEOS and APTS promoted the increment in OH groups and formation of NH2 groups on the glass surface, respectively. The presence of OH or NH2 groups was investigated by means of IR spectroscopy and elemental analysis. Moreover, in vitro study of these functionalized glasses was performed in simulated body fluid (SBF) so as to investigate the effect of functionalization on the bioactive behavior of H and Ga1.0. The results showed that the functionalization was obtained along with maintaining their bioactivity. The surfaces of both functionalized glasses were covered by a layer of apatite within 30 days of SBF immersion. In addition, CaCO3 was also identified on the surface of APTS functionalized glasses. However, no gallium release was detected during SBF soaking. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Introduction of biomolecules (peptides, proteins, drugs, etc.) on the surface of bioactive glasses is one of the emerging topics in the field of bioceramics. The pioneering study done by Weetall et al. lead to the discovery of silanization of the ceramic surface with a sol–gel precursor which is now a common approach for the bonding of biomolecules on glass surface [1,2] Recently Vernè et al. reported different cleaning and silanization methods to covalently bond bone morphogenetic proteins (BMP-2) and alkaline phosphatase on the bioactive glass surface [3,4]. Their work brought forth the importance of functional groups (such as \OH and \NH2) in grafting of biomolecules. Tetrahaethoxyosilane (TEOS) and (3-Aminopropyl)triethoxysilane (APTS) are widely used substances for silanization [5,6]. TEOS increases the number of \SiOH groups on the glass surface due to hydrolysis of ethoxy group. On the other hand, APTS enriches the surface with aminoalkyl chains. Bioactive glasses containing gallium(III) ions are worth investigating keeping in view the antimicrobial properties of gallium [7–9]. Recently, gallium containing phosphate glasses showed antimicrobial properties against Pseudomonas aeruginosa [10]. In addition, highly porous Ga-releasing scaffolds, based on Bioglass 45S5, have shown antibacterial activity against Staphylococcus aureus [11]. In the last two years, phosphosilicate glasses (based on Bioglass® 45S5 [12]) and ⁎ Corresponding author. Tel.: +39 0592055042; fax: +39 059373543. E-mail address: [email protected] (L. Menabue). 0928-4931/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2013.03.046
mesoporous sol–gel glasses (based on the composition 80% SiO2–15% CaO–5% P2O5) were modified and prepared in our laboratory by adding different amounts of Ga2O3 [13,14]. In addition to this, in vitro studies were carried out in simulated body fluid (SBF) to check their ability to form a hydroxyapatite layer and the release of gallium ions. In all the cases a layer of hydroxyapatite or hydroxycarbonate apatite was formed and the release of Ga 3+ ions was maintained well below the toxicity level. In this paper we report functionalization with TEOS and APTS and in-vitro bioactivity of a glass based on Bioglass® 45S5, whose composition has been doped with 1 mol% Ga2O3. On the other hand, the plain Bioglass® 45S5, referred to as H, is used as a reference. The solid state investigation before and after soaking of glasses in SBF will enable us to evaluate: i) the effect of gallium on functionalization, ii) the effect of functionalization on the ability of the glasses to form an apatite layer, and iii) the classification of glasses in terms of apatite rate formation on the surface and Ga 3+ release. 2. Experimental 2.1. Materials synthesis The Ga-free glass 46.2SiO2·24.3Na2O·26.9CaO·2.6P2O5, referred to as H, and Ga-containing glass 45.7SiO2·24.1Na2O·26.6CaO·2.6P2O5· 1.0Ga2O3, referred to as Ga1.0, were prepared following the protocol mentioned in the previous publication [12]. Their composition was determined with EDS analysis on ball milled powders because during
G. Lusvardi et al. / Materials Science and Engineering C 33 (2013) 3190–3196
the fusion process P-compounds could be partly volatilized. The data obtained were in the range ± 0.3% with respect to the theoretical ones. Before functionalization the glasses were milled in agate mortar to obtain grains of size below 32 μm, with the aim at maximizing the surface to be functionalized. The functionalization was performed according to the method reported in the previous publication [6]. In the first attempt, 0.5 g of the selected glass was added to a sealed polyethylene bottle containing a solution comprising of 20 mL of ethanol and 2 mL of TEOS or APTS corresponding to ≈9 × 10−3 mol. The pH of the solution was adjusted to 8 by using 1 M HCl. In the second attempt, 5 mL of TEOS or APTS was used, corresponding to ≈2.2 × 10−2 mol. The mixtures were then maintained under continuous stirring at room temperature for a time interval between 1 and 7 days. After reaction the glasses were rinsed with water and ethanol followed by drying at 120 °C for 24 h. During the functionalization step with TEOS, the pH of the solution did not change. However, in case of APTS it slightly increased to 8.2. The best conditions for TEOS functionalization were estimated by the percentage of Si% determined by EDS analysis. The highest value of Si% was obtained after 7 days of reaction with 5 mL of TEOS (Ga1.0TEOS). The same conditions were applied also for the functionalization of H glass. Tables 1 and 2 report the results of elemental analysis and atomic ratios. The best conditions for APTS functionalization of Ga1.0 glass were estimated by the percentage of C% and N%, using elemental analysis. The highest values of C% and N% were obtained after 1 day of reaction with 5 mL of APTS (Ga1.0APTS) and the same conditions were applied to H glass. 2.2. In vitro bioactivity tests In vitro bioactivity test was performed by soaking 250 mg of the powder sample in 50 mL of simulated body fluid (SBF) at 37 °C as proposed by Kokubo et al. [15]. Each sample was soaked for four different time intervals (1, 4, 15 and 30 days) under magnetic stirring in order to maintain homogeneous composition of the system during the test. The ion concentrations of SBF (mM) were: Na + = 142.0, K + = 5.0, Ca 2+ = 2.5, Mg 2+ = 1.5, Cl − = 147.8, HCO3− = 4.2, HPO42− = 1.0 and pH 7.4. After soaking, the solution was filtered with a Whatman membrane filter having a pore size of 0.45 μm and a diameter of 50 mm. The SBF solutions were tested by ICP (Inductively Coupled Plasma, Perkin Elmer Optima 4200 DV, USA) to determine the concentration (ppm) of Na, Ca, Ga, Si and P for each time period (Table 3). In order to evaluate the glass bioactivity the concentration (ppm) of Na, Ca and P was determined also for SBF solution and the result is reported in Table 3 as SBF t = 0. The concentrations reported in Table 3 represent the mean value of four different determinations
Table 2 Mean atomic % determined by EDS analysis of the glasses before (H and Ga1.0) and after (HTEOS, Ga1.0TEOS, HAPTS and Ga1.0APTS) functionalization stepa. Atom Hb O Si Ca Na P Ga C N
H a.s. Ga1.0 a.s. HTEOS 7d Ga1.0TEOS 1d Ga1.0TEOS 2d Ga1.0TEOS 3d Ga1.0TEOS 7d HAPTS 1d Ga1.0APTS 1d Ga1.0APTS 2d Ga1.0APTS 3d Ga1.0APTS 7d
EDS C/N [atomic]
H/C H/N [atomic] [atomic]
%Si [atomic%]
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.26 1.09 0.64 0.29 0.33
/ / / / / / / 2.87 3.32 3.57 3.82 4.32
6.94 3.30 32.8 178 80.7 51.2 12.6 4.27 2.72 2.79 3.00 1.56
15.7 15.2 32.8 19.6 22.4 23.7 24.8 / / / / /
0.24 0.18 0.53 0.20 0.23 0.34 0.83 0.88 3.28 2.14 1.13 1.40
0.14 0.05 0.94 0.35 0.54 0.71 0.74 0.37 0.76 0.51 0.29 0.21
/ / / / / / / 12.3 9.04 9.97 11.5 6.70
The italics identify the best samples with the maximum of functionalization.
± ± ± ± ± ± ± ±
HTEOS
Ga1.0TEOS HAPTS
Ga1.0APTS
3.5 55.7 ± 2.6 61.1 ± 5.0 70.1 ± 6.7 62.5 ± 4.2 50.7 ± 1.5 15.2 ± 2.1 32.8 ± 5.7 24.8 ± 3.1 12.5 ± 1.0 7.9 ± 1.2 9.6 ± 1.0 0.9 ± 0.1 0.9 ± 0.2 8.9 ± 1.8 4.4 ± 0.4 16.9 ± 0.4 1.3 ± 0.1 1.5 ± 0.1 10.2 ± 1.0 8.0 ± 0.2 1.8 ± 0.4 0.0 ± 0.0 0.0 ± 0.0 1.1 ± 0.1 0.9 ± 0.0 0.8 ± 0.1 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.3 ± 0.1 0.2 ± 0.1 3.7 ± 0.6 2.5 ± 0.6 3.6 ± 0.7 21.4 ± 0.0 0.0 ± 0.0 0.3 ± 0.1 0.2 ± 0.1 1.2 ± 0.3 6.5 ±
6.2 1.1 0.8 0.8 0.1 0.1 2.1 1.3
on the same sample solution. The concentration (Si, Ca, Na and Ga) difference between four replicated samples was b5%. Table 3 does not contain gallium concentration because in all the samples the value was lower than the lowest standard used (0.2 ppm). 2.3. Infrared spectroscopy The conventional KBr pellet method was followed in order to track the changes in the vibrational modes of the glasses induced by the functionalization with TEOS or APTS. Powder samples were mixed with KBr (glass: KBr = 2: 200 weight ratio) and then IR spectra of KBr pellets were recorded in the 4000–400 cm −1 spectral range by means of a Jasco FT/IR-4200. 2.4. X-ray powder diffraction XRPD (X-ray powders diffraction) patterns of milled glasses (dimension b 32 μm) were recorded on a PANalytical X'Pert Pro Bragg–Brentano diffractometer, using Ni-filtered Cu Kα radiation (λ = 1.54060 Å) with X'Celerator detector. The patterns were taken over the diffraction angle 2θ range = 10–50° and with a time step of 50 s and a step size of 0.03°. 2.5. Environmental scanning electron microscopy The surface morphology of the specimens was observed by means of ESEM (Environmental Scanning Electron Microscopy, FEI Quanta Table 3 Concentrationa (ppm) of Si, Na, Ca and P (±std. dev.) in the plain SBF solution and sample solutions separated after different soaking times in SBF. Time [Days]
Si ± 2 ppm Na ± 50 ppm Ca ± 5 ppm P ± 2 ppm [ppm] [ppm] [ppm] [ppm]
SBF (t = 0) 0 HTEOS
%N %C %H [wt.%] [wt.%] [wt.%]
55.2 15.7 9.7 17.5 1.9 0.0 0.2 0.0
Ga1.0b
±std. dev. obtained by two independent samples each analyzed onto two different areas. a Functionalization step: 7 days for TEOS, 1 day for HAPTS. b Data reported in ref. [12] for O, Si, Ca, Na, P and Ga.
Table 1 %N, %C and %H [wt.%] determined by elemental analysis (EA) and Si [atom%] by EDS analysis of samples before and after different times of functionalization with TEOS and APTS. EA
3191
1 4 15 30 1 4 Ga1.0TEOS 15 30 1 4 HAPTS 15 30 1 4 Ga1.0APTS 15 30
49 51 60 65 50 53 53 53 65 73 55 77 56 67 71 71
3480
76
25
3530 3690 3670 3700 3630 3630 3830 3950 3810 3810 4260 5180 4080 4380 4360 5170
146 178 128 55 155 137 110 146 251 188 88 228 235 215 223 230
5 0 0 0 5 0 0 0 1 0 0 0 1 0 0 0
a Reported data are an average of four different determinations on the same sample solution.
3192
G. Lusvardi et al. / Materials Science and Engineering C 33 (2013) 3190–3196
200, Fei Company, the Netherlands). Images were acquired with an electrostatic voltage of 25 kV and a working distance of 9–10 mm. The samples were analyzed without metallization with C. Microanalysis, obtained with an Energy Dispersive Spectrometer (EDS, INCA 350, Oxford Instruments, UK) coupled with ESEM was also performed to identify and label the elements on the surface. EDS measurements were performed in duplicate for each examined particle and areas of the surface and the results are presented as means ± std. dev. (Table 2). 3. Results 3.1. Functionalization The functionalized glasses HTEOS, HAPTS, Ga1.0TEOS, and Ga1.0APTS and corresponding non functionalized H and Ga1.0 glasses were analyzed by elemental analysis to find out C%, H% and N%, EDS analysis for checking Si%. Moreover, the IR spectroscopic analysis of the same glasses was also performed. These characterization techniques helped to verify and determine the amount of functionalization as a function of reaction time. Furthermore, we also calculated the atomic ratios C/N, H/C, and H/N (Table 1) as an expression of degree of functionalization,
hydrolysis of ethoxy groups and cross-linking of APTS or TEOS residues, respectively. Before calculating the atomic ratios, elemental analysis results of plain H and Ga1.0 glasses were subtracted from the C% and H% of HAPTS, Ga1.0APTS, HTEOS and Ga1.0TEOS. The result obtained by elemental analysis of HTEOS and Ga1.0TEOS after 1 day treatment with TEOS showed higher C% and H% with respect to H and Ga1.0 glasses. In addition, EDS analysis revealed more Si% than that determined for H and Ga1.0 glasses (Table 2). Moreover, the Si% and C% increased with the time of reaction (Table 1). On the other hand, the experimental C/N ratio for Ga1.0APTS increased from 3.32 to 4.32 after 7 days (Table 1). However, with the increase in the reaction time a decrease was observed in experimental H/C and H/N ratios from 2.72 to 1.56 and 9.04 to 6.70, respectively. The maximum difference between experimental and theoretical values was found after 7 days of reaction. Fig. 1 collects the FTIR spectra of Ga1.0 and H before and after functionalization. The IR spectrum (Fig. 1, section b) of Ga1.0 showed a weak broad band in the 3500–3300 cm −1 spectral region due to the hydroxylation of unsaturated Si\O groups of the glass surface. In addition, two intense bands in the range 1100–1020 and 930–920 cm −1 are assigned to bridging (BO) and not bridging (NBO) Si\O bonds, respectively. It is also interesting to note very weak bands in the range
Fig. 1. FTIR spectra, in the range 4000–500 cm−1, of the bioactive glasses H (section a) and Ga1.0 (section b) as synthesized (a.s.) and after functionalization with APTS and TEOS.
G. Lusvardi et al. / Materials Science and Engineering C 33 (2013) 3190–3196
3193
2960–2850 cm−1 attributed to the stretching vibrations of alkyl groups due to atmospheric hydrocarbons adsorbed on the glass surface [16]. After 1 day of reaction with APTS, Ga1.0APTS IR spectrum (Fig. 1, section b) still presented the Si\O and aminopropyl residue bands. In addition, an intense broad band is observed centered at 3400 cm−1 consistent both with the stretching of NH2 and OH groups on the surface. Medium intensity bands at ~1595 cm−1 consistent with the NH2 bending vibration and 2960–2850 and 1480–1350 cm−1 in the region of stretching and bending vibrations of alkyl groups, respectively, are also observed. The IR spectrum of Ga1.0TEOS was more or less similar to Ga1.0 spectrum. The main differences can be summarized as: i) higher ratio between the band intensity at 1100–1020 cm−1 and 920 cm−1 suggesting the increment of Si-BO units with respect to Si\NBO; ii) presence of two bands at 1220 and 800 cm−1 due to the formation of a high-area silica gel [16]. In Fig. 1, section a the spectra of H, HAPTS and HTEOS samples before and after functionalization are reported and showed similar bands as found in case of Ga1.0, Ga1.0APTS and Ga1.0TEOS. Finally, in the spectra of all the samples a band is observed at ~1620 cm−1 due to the scissor deformation mode of surface Lewis-coordinated water molecule [16]. 3.2. Solid state ESEM-EDS Fig. 2 represents the micrographs of Ga1.0 sample before (section a) and after functionalization with TEOS (section b) and APTS (section c). Before functionalization, the morphology of the sample was characteristic of powders with dimensions less than 32 μm. After TEOS functionalization, the sample morphology changed significantly showing the formation of agglomerates of bigger dimensions with respect to the powder glass probably due to linkage among several particles covered by a silica-gel layer. This was confirmed by the increment of Si% on the sample surface determined by EDS analysis (Table 2). At the same time it is interesting to note that the concentrations (atomic %) of Ca, Na and Ga were low. The APTS functionalization did not modify significantly the powder morphology. However, the EDS analysis (Table 2) of Ga1.0APTS sample showed high percentage of C and N. The ratio between them fit very well with the theoretical value (C/N/Si 3/1/1) provided that a complete polymerization of APTS on the glass surface has occurred. The HAPTS glass was characterized by lower C% and N% with respect to Ga1.0APTS suggesting less amount of APTS functionalized on the glass surface and the C/N ratio corresponded to the theoretical value of 3. 3.3. In vitro bioactivity tests 3.3.1. SBF solution Table 3 reports the concentrations (in ppm) of Na, Ca, Si and P in SBF before and during in vitro bioactivity. Si concentration increased slightly with increase of soaking time along with Na concentration with the highest value determined for APTS functionalized glasses. After 1 day of soaking in SBF, Ca concentration was found higher for HAPTS and Ga1.0APTS as compared to TEOS functionalized glasses. With soaking time Ca concentration for HTEOS and HAPTS significantly decreased, while for Ga1.0TEOS and Ga1.0APTS the change was not significant. However, for longer time period (>15 days) the Ca concentration increased except for HTEOS. The decrease of P concentration, after 1 day of SBF soaking, was very sharp for HAPTS and Ga1.0APTS. The Ga concentration for both Ga1.0TEOS and Ga1.0APTS was under the lowest standard used (0.2 ppm) during measurements and was independent of the time of SBF soaking. The non-monotone variation of Ca corresponds to the behavior of species involved in dissolution and precipitation processes [12]. 3.3.2. Solid state XRD The XRD powder spectra of H and Ga1.0 glasses before and after functionalization are reported in Fig. 3. Before functionalization,
Fig. 2. ESEM images of the bioactive Ga1.0 glass as synthesized (section a) and after functionalization with TEOS (section b) and APTS (section c).
a comparison between XRD spectra of powder samples having dimensions less than 32 μm and 250–500 μm [12] range showed no difference. They all exhibited a broad hump in the range 30–33°
3194
G. Lusvardi et al. / Materials Science and Engineering C 33 (2013) 3190–3196
a) HTEOS HA002
b) Ga1.0TEOS HA002
C HA211
C HA211 30d
30d
15d
15d 4d
4d 1d
1d Ga1.0TEOS
Ga1.0 a.s.
HTEOS
H a.s. 10,00
15,00
20,00
25,00
30,00
35,00
40,00
45,00
50,00
10,00
15,00
20,00
25,00
2θ [°]
35,00
40,00
45,00
50,00
2θ [°]
c) HAPTS
d) Ga1.0APTS
C HA002
30,00
C HA211
HA211 30d
30d
15d
15d
4d
4d 1d HAPTS
1d Ga1.0APTS
Ga1.0 a.s.
H a.s. 10,00
15,00
20,00
25,00
30,00
35,00
40,00
45,00
50,00
2θ [°]
10,00
15,00
20,00
25,00
30,00
35,00
40,00
45,00
50,00
2θ [°]
Fig. 3. XRD patterns of the bioactive functionalized glasses HTEOS (section a), Ga1.0TEOS (section b), HAPTS (section c) and GA1.0APTS (section d) as synthesized (a.s.) and after 1, 4, 15 and 30 days of reaction in SBF. (HA: hydroxyapatite, C: calcite CaCO3).
(2θ) typical of sodium–calcium–phosphosilicate glasses [12]. This hump existed even after functionalization with TEOS, (Fig. 3, HTEOS, Ga1.0TEOS) although slightly masked by an intense and broad amorphous silica peak at 21.36° (2θ) [17]. This indicated the formation of an amorphous silica layer on the glass surface due to the hydrolysis and polymerization of TEOS. During in vitro bioactivity measurements, the XRD patterns of HTEOS and Ga1.0TEOS were almost unchanged after 1 day except for a reduction of peak intensity. The peak of HA was observed in the XRD pattern of HTEOS after 4 days of SBF soaking. After 15 days, a peak centered at ~ 30° (2θ value) attributed to calcite was also identified. The crystalline and amorphous phases identified in the XRD patterns of Ga1.0TEOS were the same as found in HTEOS after a corresponding time of SBF soaking, but the relative intensity of HA and calcite peaks was low. The XRD spectra of as quenched samples treated with APTS (Fig. 3) presented a very weak broad band in the region of silica-gel band (around 21°, in 2θ value) in addition to the hump in the range 30–33° (2θ). These remain unchanged after 1 day of SBF soaking. The HAPTS glass presented an XRD pattern where the main peak of calcite was distinguished after 4 days. Silica gel and HA peaks were also clearly identified by 15 days whereas calcite remained the main crystalline phase even after 30 days. In case of Ga1.0APTS, the identification of calcite and HA peaks was possible after 4 and 15 days, respectively. No significant change in silica gel band was observed up to 30 days. In both HAPTS and Ga1.0APTS, the calcite phase was more abundant than in HTEOS and Ga1.0TEOS. 4. Discussion 4.1. Functionalization The reaction between glass and silanization agents was conducted in ethanol (a small contribution of H2O derived from the HCl 1 M solution) to reduce the polymerization of APTS without functionalization of the
glass surface which is favored by an aqueous medium. The parameters directly correlated to hydrolysis and cross-linking of APTS are the atomic ratios C/N, H/C and H/N and their values for free APTS are 9, 2.55 and 23, respectively. The atomic ratios C/N, H/C and H/N having values 3, 2.67 and 8, respectively, are expected after hydrolysis of ethoxy groups if APTS residues are completely cross-linked. In addition, C/N ratio can be used for checking cross-linking of APTS. The C/N ratio calculated after 1 day (Table 1) is consistent with a complete hydrolysis of alkoxy groups of Ga1.0APTS along with the formation of silanol groups followed by cross-linking between adjacent molecules. However, the H/N ratio was found greater than the theoretical value which could be due to both the presence of a significant number of Si\OH as a result of incomplete cross-linking and the solvent which remained trapped into the network of condensed APTS. The IR bands assigned to silanol, amino and alkyl groups observed in the spectra of HAPTS and Ga1.0APTS (Fig. 1) confirm silanization of glass surface. At longer time of reaction all the element percentage decrease most of H% and N% and the atomic ratio strongly deviates from both free APTS and completely hydrolyzed and condensed APTS. This can be explained by the partial degradation of glass surface during the functionalization step due to leaching of glass components in the hydro-alcoholic medium. The results indicate that leaching is faster than the functionalization process because of the steric hindrance of the aminoalkyl chain (Table 2). In fact the concentrations of Ca, Na and Si of APTS functionalized glasses were lower than the corresponding values of plain H and Ga1.0 glasses. On the basis of these results, the best conditions for functionalization with APTS are identified as 1 day of reaction and the same condition was also adopted for HAPTS. After 1 day, the experimental atomic ratios C/N, H/C and H/N for HAPTS were 2.87, 4.27 and 12.3, respectively, suggesting that the complete hydrolysis has occurred. On the basis of N% value, we can calculate the ratio between the number of APTS moles and the mass of glass, molAPTS/gglass. The ratio value for Ga1.0APTS is ≈8 × 10 −4, indicating a good yield for the functionalization reaction and the value calculated for
G. Lusvardi et al. / Materials Science and Engineering C 33 (2013) 3190–3196
HAPTS is ≈ 1.5 × 10 −4 molAPTS/gglass. On the basis of these values, we can say that Ga 3+ ions favor the grafting of APTS onto the glass surface. Sol–gel Ga-containing glasses based on the formula 77.3SiO2· 14.5CaO·4.8P2O5·3.4Ga2O3 [14] showed that low content gallium can act as network former giving rise to Brönsted acidic sites with protonation of bridging oxygens of the Si\O\Ga groups. The positive charge on the hydrogen atom attracts the amino tail of the alkyl chain of an APTS molecule. As a consequence, the partially negative oxygen atom reacts with the Si atom forming a covalent Si\O bond and a C2H5OH molecule is released with a pathway similar to that proposed by Kanan et al. [18]. The higher H/C and H/N ratios for HAPTS with respect to completely polymerized APTS suggest that the cross-linking among APTS moieties by means of Si\O\Si bridges is lower as compared to Ga1.0APTS. This is because the lower amount of APTS residues functionalized on the glass surface maintains them at a distance that does not enable an extensive condensation between silanol groups. As a result the functionalized Ga1.0APTS glass possesses surface enriched by NH2 groups and is not covered by a silica gel layer as demonstrated by the XRD pattern and EDS analysis. The C/N, H/C and H/N atomic ratios determined for HAPTS and Ga1.0APTS ensure that APTS is immobilized on the glass surface forming layered interfaces due to internal polymerization with significant cross-linking between polymer monomers but sparse cross-links between polymer and substrate [19]. The functionalization by means of TEOS can be followed by the Si% determined on the glass surface. Si% of Ga1.0TEOS increases with time along with C% indicating that the glass surface is enriched by silicon atoms. The decrease of Si/C ratio with the time of reaction indicates that at increased grafting of TEOS on the glass surface the hydrolysis of ethoxy groups is slowed down. This is possibly because of the formation of a thick silica gel layer, as showed by SEM and XRD analysis, that makes difficult the diffusion of C2H5OH formed after ethoxy group condensation. The presence of a thick silica gel layer is indicated also by IR spectra where a medium band at 800 cm−1, typical of silica gel, is observed along with a decrease of the band at 920 cm−1, assigned to Si\NBO bonds, as compared to the bands assigned to Si\BO bonds. The silica gel layer that covers the glass surface after 1 day preserves the glasses from fast leaching of glass components as indicated by the increment of Si% with respect to plain glasses. 4.2. Bioactivity A significant number of silanol groups deprotonate [20] when HTEOS and Ga1.0TEOS are soaked in SBF causing release of H + to the solution. On the other hand, increase of Na + and Ca 2+ concentration in solution (Table 3) is slowed down due to their interaction with negatively charged \Si\O − groups. So the cations are trapped by the silica gel layer without reaching the solution and in particular form \Si\O\Ca + groups that further interact with phosphate ions. After 4 days, the glasses exchanged sufficient Na + with H + of the solution to produce the OH − necessary for crystallization of HA leading to the formation of visible HA peaks. This behavior is different from the five step mechanism proposed by Hench [21] but the role of the silica gel layer for HA nucleation is confirmed. The XRD spectra of the parent H and Ga1.0 glasses showed signals of HA after 1 day [12]. XRD and SEM results indicate that the functionalization with APTS gives rise to a very thin silica gel layer mainly because the aminoalkyl side chain hinders the structural arrangement of the silica gel which is responsible for the band at 21.36°, 2θ. During the synthesis of HAPTS and Ga1.0APTS, an equilibrium can take place between amino and silanol groups on the glass surface [22] or formed upon hydrolysis of ethoxy groups: þ
−
`Si\ðCH2 Þ3 NH3 þ `Si\O ↔`Si\ðCH2 Þ3 NH2 þ `Si\OH
ð1Þ
that reduces the condensation degree of silanols and silica gel formation. When glasses are soaked in SBF the equilibrium (1) favors the
3195
observed silica release while the aminoalkyl chain disfavors a subsequent fast condensation and repolymerization of silanols. The low amount of silica gel reduces its efficacy to interact with released cations and to favor HA nucleation. The relatively high Ca 2+ concentration enables the separation of a simple binary compound as CaCO3 clearly identified after 4 days. On the other hand, the interaction of charged species of equilibrium (1) with the ions of the solution as well as with the ions released from the glasses is important especially with multiple charged ions such as Ca 2+ and phosphate ions. This can explain both the Na + concentration higher than measured for plain H and Ga1.0 glasses and the low concentration of phosphorus after 1 day. Then the immobilized phosphate and calcium ions are slowly converted into HA. After 15 days of SBF soaking, the crystallization of HA is evident along with complete consumption of phosphate. The relative intensity of CaCO3 signals undergoes only a slight increase. At this stage the Ca 2+ ions continuously released from the glasses are no longer consumed causing an increase in their concentration. The low amount of silica gel layer delays the formation of calcium phosphate and favors calcite precipitation [23–25]. The present results indicate that functionalization slowed down bioactivity of the glasses as compared to parent H and Ga1.0 glasses. Thus, the derived bioactivity order is as follows: H ≅ Ga1:0 > HTEOS ≅ Ga1:0TEOS > HAPTS ≅ Ga1:0APTS: 5. Conclusion H and Ga-containing glasses can be silanized with TEOS or APTS at room temperature. TEOS covalently bonded to the glass surface is largely polymerized and cross-linked forming a thick silica gel layer. In addition, the functionalized glass when in contact with SBF completes TEOS polymerization. The negatively charged silica gel layer strongly interacts with Ca 2+ and HA precipitation is slightly delayed with respect to plain glasses. No significant differences are observed between HTEOS and Ga1.0TEOS. The amount of APTS immobilized on the glass surface is lower as compared to TEOS and is only partially polymerized and cross-linked along with a thin silica gel layer formation. The presence of gallium enables a greater amount of APTS to be attached to the glass as compared to gallium free glass. The functionalization with APTS favors the separation of calcite after short SBF soaking time (4 days). HA is recognized after 15 days of SBF immersion. On the other hand, functionalization with TEOS favors the separation of HA after short SBF soaking time (4 days). Gallium ion concentration was always lower than 0.2 ppm, while for Ga1.0, a continuous gallium release was observed up to 0.6 ppm [12]. Acknowledgments The authors wish to thank the “Centro Interdipartimentale Grandi Strumenti” (CIGS) of the University of Modena e Reggio Emilia for instrument availability and assistance and FAR2008-UNIMORE for the financial support. References [1] H.H. Weetall, in: M. Klaus (Ed.), Methods Enzymol., vol. 44, Academic Press, London, 1976, p. 134. [2] H.H. Weetall, Trends Biotechnol. 3 (1985) 276. [3] E. Vernè, C. Vitale-Brovarone, E. Bui, C.L. Bianchi, A.R. Boccaccini, J. Biomed. Mater. Res. 90A (2009) 981. [4] E. Vernè, S. Ferrarsi, C. Vitale-Brovarone, S. Spriano, C.L. Bianchi, A. Naldoni, M. Morra, C. Cassinelli, Acta Biomater. 6 (2010) 229. [5] W. Wang, M.W. Vaughn, Scanning 30 (2008) 65. [6] Q.Z. Chen, I. Ahmed, J.C. Knowles, S.N. Nazhat, A.R. Boccaccini, K. Rezwan, J. Biomed. Mater. Res. 86A (2008) 987. [7] Y. Kaneko, M. Thoendel, O. Olakanmi, B.E. Britigan, P.K. Singh, J. Clin. Invest. 117 (2007) 877. [8] O. Olakanmi, B.E. Britigan, L.S. Schlesinger, Infect. Immun. 68 (2000) 5619.
3196
G. Lusvardi et al. / Materials Science and Engineering C 33 (2013) 3190–3196
[9] J.R. Harrington, R.J. Martens, N.D. Cohen, L.R. Bernstein, J. Vet. Pharmacol. Ther. 29 (2006) 121. [10] S.P. Valappil, D. Ready, E.A. Abou Neel, D.M. Pickup, L.A. O'Dell, W. Chrzanowski, J. Pratten, R.J. Newport, M.E. Smith, M. Wilson, J.C. Knowle, Acta Biomater. 5 (2009) 1198. [11] V. Mourino, P. Newby, A.R. Boccaccini, Adv. Eng. Mater. 12 (2010) B283. [12] M. Franchini, G. Lusvardi, G. Malavasi, L. Menabue, Mater. Sci. Eng. C 32 (2012) 1401–1406. [13] A.J. Salinas, S. Shruti, G. Malavasi, L. Menabue, M. Vallet-Regí, Acta Biomater. 7 (2011) 3452–3458. [14] V. Aina, C. Morterra, G. Lusvardi, G. Malavasi, L. Menabue, S. Shruti, C.L. Bianchi, V. Bolis, J. Phys. Chem. C 115 (2011) 22461–22474. [15] T. Kokubo, H. Kushitani, S. Sakka, T. Kitsugi, T. Yamamuro, J. Biomed. Mater. Res. 24 (1990) 721.
[16] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds, 3rd edn Wiley, New York, 1978. [17] W.D. Kingery, H.K. Bowen, D.R. Uhlmann, Introduction to Ceramics, 2nd ed. John Wiley & Sons, Canada, 1976. [18] S.M. Kanan, W.T.Y. Tze, C.P. Tripp, Langmuir 18 (2002) 6623. [19] B. Bhushan, K.J. Kwak, S. Gupta, S.C. Lee, J. R. Soc. Interface 6 (2009) 719. [20] J.M. Rosenholm, T. Czuryszkiewicz, F. Kleitz, J.B. Rosenholm, M. Lindén, Langmuir 23 (2007) 4315. [21] L.L. Hench, J. Am. Ceram. Soc. 81 (1998) 1705. [22] C.-H. Chiang, H. Ishida, J.L. Koenig, J. Colloid Interface Sci. 74 (1980) 396. [23] M.R. Figuerias, G.R. Torre, L.L. Hench, J. Biomed. Mater. Res. 23 (1993) 445. [24] P. Li, C. Ohtsuki, T. Kokubo, K. Nakanishi, N. Soga, T. Nakamura, et al., J. Am. Ceram. Soc. 75 (1992) 2094. [25] T. Kokubo, H.-M. Kim, M. Kawashita, Biomaterials 24 (2003) 2161.
Contents lists available at SciVerse ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Gallium-containing phosphosilicate glasses: Functionalization and in-vitro bioactivity Gigliola Lusvardi, Gianluca Malavasi, Ledi Menabue ⁎, Shruti Shruti Department of Chemical and Geological Sciences, University of Modena and Reggio Emilia, Via G. Campi 183, 41125 Modena, Italy
a r t i c l e
i n f o
Article history: Received 8 October 2012 Received in revised form 4 March 2013 Accepted 28 March 2013 Available online 6 April 2013 Keywords: Glasses functionalization Phospho-silicate glasses Gallium oxide Bioactivity test
a b s t r a c t A gallium containing glass 45.7SiO2·24.1Na2O·26.6CaO·2.6P2O5·1.0Ga2O3 (referred to as “Ga1.0”) and a parent Ga-free glass 46.2SiO2·24.3Na2O·26.9CaO·2.6P2O5 (hereinafter represented as “H”), corresponding to Bioglass® 45S5, were functionalized with Tetraethoxysilane (TEOS) and (3-Aminopropyl)triethoxysilane (APTS) in order to improve their ability to bond with biomolecules, such as drugs, proteins, and peptides. Functionalization with TEOS and APTS promoted the increment in OH groups and formation of NH2 groups on the glass surface, respectively. The presence of OH or NH2 groups was investigated by means of IR spectroscopy and elemental analysis. Moreover, in vitro study of these functionalized glasses was performed in simulated body fluid (SBF) so as to investigate the effect of functionalization on the bioactive behavior of H and Ga1.0. The results showed that the functionalization was obtained along with maintaining their bioactivity. The surfaces of both functionalized glasses were covered by a layer of apatite within 30 days of SBF immersion. In addition, CaCO3 was also identified on the surface of APTS functionalized glasses. However, no gallium release was detected during SBF soaking. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Introduction of biomolecules (peptides, proteins, drugs, etc.) on the surface of bioactive glasses is one of the emerging topics in the field of bioceramics. The pioneering study done by Weetall et al. lead to the discovery of silanization of the ceramic surface with a sol–gel precursor which is now a common approach for the bonding of biomolecules on glass surface [1,2] Recently Vernè et al. reported different cleaning and silanization methods to covalently bond bone morphogenetic proteins (BMP-2) and alkaline phosphatase on the bioactive glass surface [3,4]. Their work brought forth the importance of functional groups (such as \OH and \NH2) in grafting of biomolecules. Tetrahaethoxyosilane (TEOS) and (3-Aminopropyl)triethoxysilane (APTS) are widely used substances for silanization [5,6]. TEOS increases the number of \SiOH groups on the glass surface due to hydrolysis of ethoxy group. On the other hand, APTS enriches the surface with aminoalkyl chains. Bioactive glasses containing gallium(III) ions are worth investigating keeping in view the antimicrobial properties of gallium [7–9]. Recently, gallium containing phosphate glasses showed antimicrobial properties against Pseudomonas aeruginosa [10]. In addition, highly porous Ga-releasing scaffolds, based on Bioglass 45S5, have shown antibacterial activity against Staphylococcus aureus [11]. In the last two years, phosphosilicate glasses (based on Bioglass® 45S5 [12]) and ⁎ Corresponding author. Tel.: +39 0592055042; fax: +39 059373543. E-mail address: [email protected] (L. Menabue). 0928-4931/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2013.03.046
mesoporous sol–gel glasses (based on the composition 80% SiO2–15% CaO–5% P2O5) were modified and prepared in our laboratory by adding different amounts of Ga2O3 [13,14]. In addition to this, in vitro studies were carried out in simulated body fluid (SBF) to check their ability to form a hydroxyapatite layer and the release of gallium ions. In all the cases a layer of hydroxyapatite or hydroxycarbonate apatite was formed and the release of Ga 3+ ions was maintained well below the toxicity level. In this paper we report functionalization with TEOS and APTS and in-vitro bioactivity of a glass based on Bioglass® 45S5, whose composition has been doped with 1 mol% Ga2O3. On the other hand, the plain Bioglass® 45S5, referred to as H, is used as a reference. The solid state investigation before and after soaking of glasses in SBF will enable us to evaluate: i) the effect of gallium on functionalization, ii) the effect of functionalization on the ability of the glasses to form an apatite layer, and iii) the classification of glasses in terms of apatite rate formation on the surface and Ga 3+ release. 2. Experimental 2.1. Materials synthesis The Ga-free glass 46.2SiO2·24.3Na2O·26.9CaO·2.6P2O5, referred to as H, and Ga-containing glass 45.7SiO2·24.1Na2O·26.6CaO·2.6P2O5· 1.0Ga2O3, referred to as Ga1.0, were prepared following the protocol mentioned in the previous publication [12]. Their composition was determined with EDS analysis on ball milled powders because during
G. Lusvardi et al. / Materials Science and Engineering C 33 (2013) 3190–3196
the fusion process P-compounds could be partly volatilized. The data obtained were in the range ± 0.3% with respect to the theoretical ones. Before functionalization the glasses were milled in agate mortar to obtain grains of size below 32 μm, with the aim at maximizing the surface to be functionalized. The functionalization was performed according to the method reported in the previous publication [6]. In the first attempt, 0.5 g of the selected glass was added to a sealed polyethylene bottle containing a solution comprising of 20 mL of ethanol and 2 mL of TEOS or APTS corresponding to ≈9 × 10−3 mol. The pH of the solution was adjusted to 8 by using 1 M HCl. In the second attempt, 5 mL of TEOS or APTS was used, corresponding to ≈2.2 × 10−2 mol. The mixtures were then maintained under continuous stirring at room temperature for a time interval between 1 and 7 days. After reaction the glasses were rinsed with water and ethanol followed by drying at 120 °C for 24 h. During the functionalization step with TEOS, the pH of the solution did not change. However, in case of APTS it slightly increased to 8.2. The best conditions for TEOS functionalization were estimated by the percentage of Si% determined by EDS analysis. The highest value of Si% was obtained after 7 days of reaction with 5 mL of TEOS (Ga1.0TEOS). The same conditions were applied also for the functionalization of H glass. Tables 1 and 2 report the results of elemental analysis and atomic ratios. The best conditions for APTS functionalization of Ga1.0 glass were estimated by the percentage of C% and N%, using elemental analysis. The highest values of C% and N% were obtained after 1 day of reaction with 5 mL of APTS (Ga1.0APTS) and the same conditions were applied to H glass. 2.2. In vitro bioactivity tests In vitro bioactivity test was performed by soaking 250 mg of the powder sample in 50 mL of simulated body fluid (SBF) at 37 °C as proposed by Kokubo et al. [15]. Each sample was soaked for four different time intervals (1, 4, 15 and 30 days) under magnetic stirring in order to maintain homogeneous composition of the system during the test. The ion concentrations of SBF (mM) were: Na + = 142.0, K + = 5.0, Ca 2+ = 2.5, Mg 2+ = 1.5, Cl − = 147.8, HCO3− = 4.2, HPO42− = 1.0 and pH 7.4. After soaking, the solution was filtered with a Whatman membrane filter having a pore size of 0.45 μm and a diameter of 50 mm. The SBF solutions were tested by ICP (Inductively Coupled Plasma, Perkin Elmer Optima 4200 DV, USA) to determine the concentration (ppm) of Na, Ca, Ga, Si and P for each time period (Table 3). In order to evaluate the glass bioactivity the concentration (ppm) of Na, Ca and P was determined also for SBF solution and the result is reported in Table 3 as SBF t = 0. The concentrations reported in Table 3 represent the mean value of four different determinations
Table 2 Mean atomic % determined by EDS analysis of the glasses before (H and Ga1.0) and after (HTEOS, Ga1.0TEOS, HAPTS and Ga1.0APTS) functionalization stepa. Atom Hb O Si Ca Na P Ga C N
H a.s. Ga1.0 a.s. HTEOS 7d Ga1.0TEOS 1d Ga1.0TEOS 2d Ga1.0TEOS 3d Ga1.0TEOS 7d HAPTS 1d Ga1.0APTS 1d Ga1.0APTS 2d Ga1.0APTS 3d Ga1.0APTS 7d
EDS C/N [atomic]
H/C H/N [atomic] [atomic]
%Si [atomic%]
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.26 1.09 0.64 0.29 0.33
/ / / / / / / 2.87 3.32 3.57 3.82 4.32
6.94 3.30 32.8 178 80.7 51.2 12.6 4.27 2.72 2.79 3.00 1.56
15.7 15.2 32.8 19.6 22.4 23.7 24.8 / / / / /
0.24 0.18 0.53 0.20 0.23 0.34 0.83 0.88 3.28 2.14 1.13 1.40
0.14 0.05 0.94 0.35 0.54 0.71 0.74 0.37 0.76 0.51 0.29 0.21
/ / / / / / / 12.3 9.04 9.97 11.5 6.70
The italics identify the best samples with the maximum of functionalization.
± ± ± ± ± ± ± ±
HTEOS
Ga1.0TEOS HAPTS
Ga1.0APTS
3.5 55.7 ± 2.6 61.1 ± 5.0 70.1 ± 6.7 62.5 ± 4.2 50.7 ± 1.5 15.2 ± 2.1 32.8 ± 5.7 24.8 ± 3.1 12.5 ± 1.0 7.9 ± 1.2 9.6 ± 1.0 0.9 ± 0.1 0.9 ± 0.2 8.9 ± 1.8 4.4 ± 0.4 16.9 ± 0.4 1.3 ± 0.1 1.5 ± 0.1 10.2 ± 1.0 8.0 ± 0.2 1.8 ± 0.4 0.0 ± 0.0 0.0 ± 0.0 1.1 ± 0.1 0.9 ± 0.0 0.8 ± 0.1 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.3 ± 0.1 0.2 ± 0.1 3.7 ± 0.6 2.5 ± 0.6 3.6 ± 0.7 21.4 ± 0.0 0.0 ± 0.0 0.3 ± 0.1 0.2 ± 0.1 1.2 ± 0.3 6.5 ±
6.2 1.1 0.8 0.8 0.1 0.1 2.1 1.3
on the same sample solution. The concentration (Si, Ca, Na and Ga) difference between four replicated samples was b5%. Table 3 does not contain gallium concentration because in all the samples the value was lower than the lowest standard used (0.2 ppm). 2.3. Infrared spectroscopy The conventional KBr pellet method was followed in order to track the changes in the vibrational modes of the glasses induced by the functionalization with TEOS or APTS. Powder samples were mixed with KBr (glass: KBr = 2: 200 weight ratio) and then IR spectra of KBr pellets were recorded in the 4000–400 cm −1 spectral range by means of a Jasco FT/IR-4200. 2.4. X-ray powder diffraction XRPD (X-ray powders diffraction) patterns of milled glasses (dimension b 32 μm) were recorded on a PANalytical X'Pert Pro Bragg–Brentano diffractometer, using Ni-filtered Cu Kα radiation (λ = 1.54060 Å) with X'Celerator detector. The patterns were taken over the diffraction angle 2θ range = 10–50° and with a time step of 50 s and a step size of 0.03°. 2.5. Environmental scanning electron microscopy The surface morphology of the specimens was observed by means of ESEM (Environmental Scanning Electron Microscopy, FEI Quanta Table 3 Concentrationa (ppm) of Si, Na, Ca and P (±std. dev.) in the plain SBF solution and sample solutions separated after different soaking times in SBF. Time [Days]
Si ± 2 ppm Na ± 50 ppm Ca ± 5 ppm P ± 2 ppm [ppm] [ppm] [ppm] [ppm]
SBF (t = 0) 0 HTEOS
%N %C %H [wt.%] [wt.%] [wt.%]
55.2 15.7 9.7 17.5 1.9 0.0 0.2 0.0
Ga1.0b
±std. dev. obtained by two independent samples each analyzed onto two different areas. a Functionalization step: 7 days for TEOS, 1 day for HAPTS. b Data reported in ref. [12] for O, Si, Ca, Na, P and Ga.
Table 1 %N, %C and %H [wt.%] determined by elemental analysis (EA) and Si [atom%] by EDS analysis of samples before and after different times of functionalization with TEOS and APTS. EA
3191
1 4 15 30 1 4 Ga1.0TEOS 15 30 1 4 HAPTS 15 30 1 4 Ga1.0APTS 15 30
49 51 60 65 50 53 53 53 65 73 55 77 56 67 71 71
3480
76
25
3530 3690 3670 3700 3630 3630 3830 3950 3810 3810 4260 5180 4080 4380 4360 5170
146 178 128 55 155 137 110 146 251 188 88 228 235 215 223 230
5 0 0 0 5 0 0 0 1 0 0 0 1 0 0 0
a Reported data are an average of four different determinations on the same sample solution.
3192
G. Lusvardi et al. / Materials Science and Engineering C 33 (2013) 3190–3196
200, Fei Company, the Netherlands). Images were acquired with an electrostatic voltage of 25 kV and a working distance of 9–10 mm. The samples were analyzed without metallization with C. Microanalysis, obtained with an Energy Dispersive Spectrometer (EDS, INCA 350, Oxford Instruments, UK) coupled with ESEM was also performed to identify and label the elements on the surface. EDS measurements were performed in duplicate for each examined particle and areas of the surface and the results are presented as means ± std. dev. (Table 2). 3. Results 3.1. Functionalization The functionalized glasses HTEOS, HAPTS, Ga1.0TEOS, and Ga1.0APTS and corresponding non functionalized H and Ga1.0 glasses were analyzed by elemental analysis to find out C%, H% and N%, EDS analysis for checking Si%. Moreover, the IR spectroscopic analysis of the same glasses was also performed. These characterization techniques helped to verify and determine the amount of functionalization as a function of reaction time. Furthermore, we also calculated the atomic ratios C/N, H/C, and H/N (Table 1) as an expression of degree of functionalization,
hydrolysis of ethoxy groups and cross-linking of APTS or TEOS residues, respectively. Before calculating the atomic ratios, elemental analysis results of plain H and Ga1.0 glasses were subtracted from the C% and H% of HAPTS, Ga1.0APTS, HTEOS and Ga1.0TEOS. The result obtained by elemental analysis of HTEOS and Ga1.0TEOS after 1 day treatment with TEOS showed higher C% and H% with respect to H and Ga1.0 glasses. In addition, EDS analysis revealed more Si% than that determined for H and Ga1.0 glasses (Table 2). Moreover, the Si% and C% increased with the time of reaction (Table 1). On the other hand, the experimental C/N ratio for Ga1.0APTS increased from 3.32 to 4.32 after 7 days (Table 1). However, with the increase in the reaction time a decrease was observed in experimental H/C and H/N ratios from 2.72 to 1.56 and 9.04 to 6.70, respectively. The maximum difference between experimental and theoretical values was found after 7 days of reaction. Fig. 1 collects the FTIR spectra of Ga1.0 and H before and after functionalization. The IR spectrum (Fig. 1, section b) of Ga1.0 showed a weak broad band in the 3500–3300 cm −1 spectral region due to the hydroxylation of unsaturated Si\O groups of the glass surface. In addition, two intense bands in the range 1100–1020 and 930–920 cm −1 are assigned to bridging (BO) and not bridging (NBO) Si\O bonds, respectively. It is also interesting to note very weak bands in the range
Fig. 1. FTIR spectra, in the range 4000–500 cm−1, of the bioactive glasses H (section a) and Ga1.0 (section b) as synthesized (a.s.) and after functionalization with APTS and TEOS.
G. Lusvardi et al. / Materials Science and Engineering C 33 (2013) 3190–3196
3193
2960–2850 cm−1 attributed to the stretching vibrations of alkyl groups due to atmospheric hydrocarbons adsorbed on the glass surface [16]. After 1 day of reaction with APTS, Ga1.0APTS IR spectrum (Fig. 1, section b) still presented the Si\O and aminopropyl residue bands. In addition, an intense broad band is observed centered at 3400 cm−1 consistent both with the stretching of NH2 and OH groups on the surface. Medium intensity bands at ~1595 cm−1 consistent with the NH2 bending vibration and 2960–2850 and 1480–1350 cm−1 in the region of stretching and bending vibrations of alkyl groups, respectively, are also observed. The IR spectrum of Ga1.0TEOS was more or less similar to Ga1.0 spectrum. The main differences can be summarized as: i) higher ratio between the band intensity at 1100–1020 cm−1 and 920 cm−1 suggesting the increment of Si-BO units with respect to Si\NBO; ii) presence of two bands at 1220 and 800 cm−1 due to the formation of a high-area silica gel [16]. In Fig. 1, section a the spectra of H, HAPTS and HTEOS samples before and after functionalization are reported and showed similar bands as found in case of Ga1.0, Ga1.0APTS and Ga1.0TEOS. Finally, in the spectra of all the samples a band is observed at ~1620 cm−1 due to the scissor deformation mode of surface Lewis-coordinated water molecule [16]. 3.2. Solid state ESEM-EDS Fig. 2 represents the micrographs of Ga1.0 sample before (section a) and after functionalization with TEOS (section b) and APTS (section c). Before functionalization, the morphology of the sample was characteristic of powders with dimensions less than 32 μm. After TEOS functionalization, the sample morphology changed significantly showing the formation of agglomerates of bigger dimensions with respect to the powder glass probably due to linkage among several particles covered by a silica-gel layer. This was confirmed by the increment of Si% on the sample surface determined by EDS analysis (Table 2). At the same time it is interesting to note that the concentrations (atomic %) of Ca, Na and Ga were low. The APTS functionalization did not modify significantly the powder morphology. However, the EDS analysis (Table 2) of Ga1.0APTS sample showed high percentage of C and N. The ratio between them fit very well with the theoretical value (C/N/Si 3/1/1) provided that a complete polymerization of APTS on the glass surface has occurred. The HAPTS glass was characterized by lower C% and N% with respect to Ga1.0APTS suggesting less amount of APTS functionalized on the glass surface and the C/N ratio corresponded to the theoretical value of 3. 3.3. In vitro bioactivity tests 3.3.1. SBF solution Table 3 reports the concentrations (in ppm) of Na, Ca, Si and P in SBF before and during in vitro bioactivity. Si concentration increased slightly with increase of soaking time along with Na concentration with the highest value determined for APTS functionalized glasses. After 1 day of soaking in SBF, Ca concentration was found higher for HAPTS and Ga1.0APTS as compared to TEOS functionalized glasses. With soaking time Ca concentration for HTEOS and HAPTS significantly decreased, while for Ga1.0TEOS and Ga1.0APTS the change was not significant. However, for longer time period (>15 days) the Ca concentration increased except for HTEOS. The decrease of P concentration, after 1 day of SBF soaking, was very sharp for HAPTS and Ga1.0APTS. The Ga concentration for both Ga1.0TEOS and Ga1.0APTS was under the lowest standard used (0.2 ppm) during measurements and was independent of the time of SBF soaking. The non-monotone variation of Ca corresponds to the behavior of species involved in dissolution and precipitation processes [12]. 3.3.2. Solid state XRD The XRD powder spectra of H and Ga1.0 glasses before and after functionalization are reported in Fig. 3. Before functionalization,
Fig. 2. ESEM images of the bioactive Ga1.0 glass as synthesized (section a) and after functionalization with TEOS (section b) and APTS (section c).
a comparison between XRD spectra of powder samples having dimensions less than 32 μm and 250–500 μm [12] range showed no difference. They all exhibited a broad hump in the range 30–33°
3194
G. Lusvardi et al. / Materials Science and Engineering C 33 (2013) 3190–3196
a) HTEOS HA002
b) Ga1.0TEOS HA002
C HA211
C HA211 30d
30d
15d
15d 4d
4d 1d
1d Ga1.0TEOS
Ga1.0 a.s.
HTEOS
H a.s. 10,00
15,00
20,00
25,00
30,00
35,00
40,00
45,00
50,00
10,00
15,00
20,00
25,00
2θ [°]
35,00
40,00
45,00
50,00
2θ [°]
c) HAPTS
d) Ga1.0APTS
C HA002
30,00
C HA211
HA211 30d
30d
15d
15d
4d
4d 1d HAPTS
1d Ga1.0APTS
Ga1.0 a.s.
H a.s. 10,00
15,00
20,00
25,00
30,00
35,00
40,00
45,00
50,00
2θ [°]
10,00
15,00
20,00
25,00
30,00
35,00
40,00
45,00
50,00
2θ [°]
Fig. 3. XRD patterns of the bioactive functionalized glasses HTEOS (section a), Ga1.0TEOS (section b), HAPTS (section c) and GA1.0APTS (section d) as synthesized (a.s.) and after 1, 4, 15 and 30 days of reaction in SBF. (HA: hydroxyapatite, C: calcite CaCO3).
(2θ) typical of sodium–calcium–phosphosilicate glasses [12]. This hump existed even after functionalization with TEOS, (Fig. 3, HTEOS, Ga1.0TEOS) although slightly masked by an intense and broad amorphous silica peak at 21.36° (2θ) [17]. This indicated the formation of an amorphous silica layer on the glass surface due to the hydrolysis and polymerization of TEOS. During in vitro bioactivity measurements, the XRD patterns of HTEOS and Ga1.0TEOS were almost unchanged after 1 day except for a reduction of peak intensity. The peak of HA was observed in the XRD pattern of HTEOS after 4 days of SBF soaking. After 15 days, a peak centered at ~ 30° (2θ value) attributed to calcite was also identified. The crystalline and amorphous phases identified in the XRD patterns of Ga1.0TEOS were the same as found in HTEOS after a corresponding time of SBF soaking, but the relative intensity of HA and calcite peaks was low. The XRD spectra of as quenched samples treated with APTS (Fig. 3) presented a very weak broad band in the region of silica-gel band (around 21°, in 2θ value) in addition to the hump in the range 30–33° (2θ). These remain unchanged after 1 day of SBF soaking. The HAPTS glass presented an XRD pattern where the main peak of calcite was distinguished after 4 days. Silica gel and HA peaks were also clearly identified by 15 days whereas calcite remained the main crystalline phase even after 30 days. In case of Ga1.0APTS, the identification of calcite and HA peaks was possible after 4 and 15 days, respectively. No significant change in silica gel band was observed up to 30 days. In both HAPTS and Ga1.0APTS, the calcite phase was more abundant than in HTEOS and Ga1.0TEOS. 4. Discussion 4.1. Functionalization The reaction between glass and silanization agents was conducted in ethanol (a small contribution of H2O derived from the HCl 1 M solution) to reduce the polymerization of APTS without functionalization of the
glass surface which is favored by an aqueous medium. The parameters directly correlated to hydrolysis and cross-linking of APTS are the atomic ratios C/N, H/C and H/N and their values for free APTS are 9, 2.55 and 23, respectively. The atomic ratios C/N, H/C and H/N having values 3, 2.67 and 8, respectively, are expected after hydrolysis of ethoxy groups if APTS residues are completely cross-linked. In addition, C/N ratio can be used for checking cross-linking of APTS. The C/N ratio calculated after 1 day (Table 1) is consistent with a complete hydrolysis of alkoxy groups of Ga1.0APTS along with the formation of silanol groups followed by cross-linking between adjacent molecules. However, the H/N ratio was found greater than the theoretical value which could be due to both the presence of a significant number of Si\OH as a result of incomplete cross-linking and the solvent which remained trapped into the network of condensed APTS. The IR bands assigned to silanol, amino and alkyl groups observed in the spectra of HAPTS and Ga1.0APTS (Fig. 1) confirm silanization of glass surface. At longer time of reaction all the element percentage decrease most of H% and N% and the atomic ratio strongly deviates from both free APTS and completely hydrolyzed and condensed APTS. This can be explained by the partial degradation of glass surface during the functionalization step due to leaching of glass components in the hydro-alcoholic medium. The results indicate that leaching is faster than the functionalization process because of the steric hindrance of the aminoalkyl chain (Table 2). In fact the concentrations of Ca, Na and Si of APTS functionalized glasses were lower than the corresponding values of plain H and Ga1.0 glasses. On the basis of these results, the best conditions for functionalization with APTS are identified as 1 day of reaction and the same condition was also adopted for HAPTS. After 1 day, the experimental atomic ratios C/N, H/C and H/N for HAPTS were 2.87, 4.27 and 12.3, respectively, suggesting that the complete hydrolysis has occurred. On the basis of N% value, we can calculate the ratio between the number of APTS moles and the mass of glass, molAPTS/gglass. The ratio value for Ga1.0APTS is ≈8 × 10 −4, indicating a good yield for the functionalization reaction and the value calculated for
G. Lusvardi et al. / Materials Science and Engineering C 33 (2013) 3190–3196
HAPTS is ≈ 1.5 × 10 −4 molAPTS/gglass. On the basis of these values, we can say that Ga 3+ ions favor the grafting of APTS onto the glass surface. Sol–gel Ga-containing glasses based on the formula 77.3SiO2· 14.5CaO·4.8P2O5·3.4Ga2O3 [14] showed that low content gallium can act as network former giving rise to Brönsted acidic sites with protonation of bridging oxygens of the Si\O\Ga groups. The positive charge on the hydrogen atom attracts the amino tail of the alkyl chain of an APTS molecule. As a consequence, the partially negative oxygen atom reacts with the Si atom forming a covalent Si\O bond and a C2H5OH molecule is released with a pathway similar to that proposed by Kanan et al. [18]. The higher H/C and H/N ratios for HAPTS with respect to completely polymerized APTS suggest that the cross-linking among APTS moieties by means of Si\O\Si bridges is lower as compared to Ga1.0APTS. This is because the lower amount of APTS residues functionalized on the glass surface maintains them at a distance that does not enable an extensive condensation between silanol groups. As a result the functionalized Ga1.0APTS glass possesses surface enriched by NH2 groups and is not covered by a silica gel layer as demonstrated by the XRD pattern and EDS analysis. The C/N, H/C and H/N atomic ratios determined for HAPTS and Ga1.0APTS ensure that APTS is immobilized on the glass surface forming layered interfaces due to internal polymerization with significant cross-linking between polymer monomers but sparse cross-links between polymer and substrate [19]. The functionalization by means of TEOS can be followed by the Si% determined on the glass surface. Si% of Ga1.0TEOS increases with time along with C% indicating that the glass surface is enriched by silicon atoms. The decrease of Si/C ratio with the time of reaction indicates that at increased grafting of TEOS on the glass surface the hydrolysis of ethoxy groups is slowed down. This is possibly because of the formation of a thick silica gel layer, as showed by SEM and XRD analysis, that makes difficult the diffusion of C2H5OH formed after ethoxy group condensation. The presence of a thick silica gel layer is indicated also by IR spectra where a medium band at 800 cm−1, typical of silica gel, is observed along with a decrease of the band at 920 cm−1, assigned to Si\NBO bonds, as compared to the bands assigned to Si\BO bonds. The silica gel layer that covers the glass surface after 1 day preserves the glasses from fast leaching of glass components as indicated by the increment of Si% with respect to plain glasses. 4.2. Bioactivity A significant number of silanol groups deprotonate [20] when HTEOS and Ga1.0TEOS are soaked in SBF causing release of H + to the solution. On the other hand, increase of Na + and Ca 2+ concentration in solution (Table 3) is slowed down due to their interaction with negatively charged \Si\O − groups. So the cations are trapped by the silica gel layer without reaching the solution and in particular form \Si\O\Ca + groups that further interact with phosphate ions. After 4 days, the glasses exchanged sufficient Na + with H + of the solution to produce the OH − necessary for crystallization of HA leading to the formation of visible HA peaks. This behavior is different from the five step mechanism proposed by Hench [21] but the role of the silica gel layer for HA nucleation is confirmed. The XRD spectra of the parent H and Ga1.0 glasses showed signals of HA after 1 day [12]. XRD and SEM results indicate that the functionalization with APTS gives rise to a very thin silica gel layer mainly because the aminoalkyl side chain hinders the structural arrangement of the silica gel which is responsible for the band at 21.36°, 2θ. During the synthesis of HAPTS and Ga1.0APTS, an equilibrium can take place between amino and silanol groups on the glass surface [22] or formed upon hydrolysis of ethoxy groups: þ
−
`Si\ðCH2 Þ3 NH3 þ `Si\O ↔`Si\ðCH2 Þ3 NH2 þ `Si\OH
ð1Þ
that reduces the condensation degree of silanols and silica gel formation. When glasses are soaked in SBF the equilibrium (1) favors the
3195
observed silica release while the aminoalkyl chain disfavors a subsequent fast condensation and repolymerization of silanols. The low amount of silica gel reduces its efficacy to interact with released cations and to favor HA nucleation. The relatively high Ca 2+ concentration enables the separation of a simple binary compound as CaCO3 clearly identified after 4 days. On the other hand, the interaction of charged species of equilibrium (1) with the ions of the solution as well as with the ions released from the glasses is important especially with multiple charged ions such as Ca 2+ and phosphate ions. This can explain both the Na + concentration higher than measured for plain H and Ga1.0 glasses and the low concentration of phosphorus after 1 day. Then the immobilized phosphate and calcium ions are slowly converted into HA. After 15 days of SBF soaking, the crystallization of HA is evident along with complete consumption of phosphate. The relative intensity of CaCO3 signals undergoes only a slight increase. At this stage the Ca 2+ ions continuously released from the glasses are no longer consumed causing an increase in their concentration. The low amount of silica gel layer delays the formation of calcium phosphate and favors calcite precipitation [23–25]. The present results indicate that functionalization slowed down bioactivity of the glasses as compared to parent H and Ga1.0 glasses. Thus, the derived bioactivity order is as follows: H ≅ Ga1:0 > HTEOS ≅ Ga1:0TEOS > HAPTS ≅ Ga1:0APTS: 5. Conclusion H and Ga-containing glasses can be silanized with TEOS or APTS at room temperature. TEOS covalently bonded to the glass surface is largely polymerized and cross-linked forming a thick silica gel layer. In addition, the functionalized glass when in contact with SBF completes TEOS polymerization. The negatively charged silica gel layer strongly interacts with Ca 2+ and HA precipitation is slightly delayed with respect to plain glasses. No significant differences are observed between HTEOS and Ga1.0TEOS. The amount of APTS immobilized on the glass surface is lower as compared to TEOS and is only partially polymerized and cross-linked along with a thin silica gel layer formation. The presence of gallium enables a greater amount of APTS to be attached to the glass as compared to gallium free glass. The functionalization with APTS favors the separation of calcite after short SBF soaking time (4 days). HA is recognized after 15 days of SBF immersion. On the other hand, functionalization with TEOS favors the separation of HA after short SBF soaking time (4 days). Gallium ion concentration was always lower than 0.2 ppm, while for Ga1.0, a continuous gallium release was observed up to 0.6 ppm [12]. Acknowledgments The authors wish to thank the “Centro Interdipartimentale Grandi Strumenti” (CIGS) of the University of Modena e Reggio Emilia for instrument availability and assistance and FAR2008-UNIMORE for the financial support. References [1] H.H. Weetall, in: M. Klaus (Ed.), Methods Enzymol., vol. 44, Academic Press, London, 1976, p. 134. [2] H.H. Weetall, Trends Biotechnol. 3 (1985) 276. [3] E. Vernè, C. Vitale-Brovarone, E. Bui, C.L. Bianchi, A.R. Boccaccini, J. Biomed. Mater. Res. 90A (2009) 981. [4] E. Vernè, S. Ferrarsi, C. Vitale-Brovarone, S. Spriano, C.L. Bianchi, A. Naldoni, M. Morra, C. Cassinelli, Acta Biomater. 6 (2010) 229. [5] W. Wang, M.W. Vaughn, Scanning 30 (2008) 65. [6] Q.Z. Chen, I. Ahmed, J.C. Knowles, S.N. Nazhat, A.R. Boccaccini, K. Rezwan, J. Biomed. Mater. Res. 86A (2008) 987. [7] Y. Kaneko, M. Thoendel, O. Olakanmi, B.E. Britigan, P.K. Singh, J. Clin. Invest. 117 (2007) 877. [8] O. Olakanmi, B.E. Britigan, L.S. Schlesinger, Infect. Immun. 68 (2000) 5619.
3196
G. Lusvardi et al. / Materials Science and Engineering C 33 (2013) 3190–3196
[9] J.R. Harrington, R.J. Martens, N.D. Cohen, L.R. Bernstein, J. Vet. Pharmacol. Ther. 29 (2006) 121. [10] S.P. Valappil, D. Ready, E.A. Abou Neel, D.M. Pickup, L.A. O'Dell, W. Chrzanowski, J. Pratten, R.J. Newport, M.E. Smith, M. Wilson, J.C. Knowle, Acta Biomater. 5 (2009) 1198. [11] V. Mourino, P. Newby, A.R. Boccaccini, Adv. Eng. Mater. 12 (2010) B283. [12] M. Franchini, G. Lusvardi, G. Malavasi, L. Menabue, Mater. Sci. Eng. C 32 (2012) 1401–1406. [13] A.J. Salinas, S. Shruti, G. Malavasi, L. Menabue, M. Vallet-Regí, Acta Biomater. 7 (2011) 3452–3458. [14] V. Aina, C. Morterra, G. Lusvardi, G. Malavasi, L. Menabue, S. Shruti, C.L. Bianchi, V. Bolis, J. Phys. Chem. C 115 (2011) 22461–22474. [15] T. Kokubo, H. Kushitani, S. Sakka, T. Kitsugi, T. Yamamuro, J. Biomed. Mater. Res. 24 (1990) 721.
[16] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds, 3rd edn Wiley, New York, 1978. [17] W.D. Kingery, H.K. Bowen, D.R. Uhlmann, Introduction to Ceramics, 2nd ed. John Wiley & Sons, Canada, 1976. [18] S.M. Kanan, W.T.Y. Tze, C.P. Tripp, Langmuir 18 (2002) 6623. [19] B. Bhushan, K.J. Kwak, S. Gupta, S.C. Lee, J. R. Soc. Interface 6 (2009) 719. [20] J.M. Rosenholm, T. Czuryszkiewicz, F. Kleitz, J.B. Rosenholm, M. Lindén, Langmuir 23 (2007) 4315. [21] L.L. Hench, J. Am. Ceram. Soc. 81 (1998) 1705. [22] C.-H. Chiang, H. Ishida, J.L. Koenig, J. Colloid Interface Sci. 74 (1980) 396. [23] M.R. Figuerias, G.R. Torre, L.L. Hench, J. Biomed. Mater. Res. 23 (1993) 445. [24] P. Li, C. Ohtsuki, T. Kokubo, K. Nakanishi, N. Soga, T. Nakamura, et al., J. Am. Ceram. Soc. 75 (1992) 2094. [25] T. Kokubo, H.-M. Kim, M. Kawashita, Biomaterials 24 (2003) 2161.