Materials Science and Engineering C 53 (2015) 252–261

Contents lists available at ScienceDirect

Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec

Influence of ZnO/MgO substitution on sintering, crystallisation, and bio-activity of alkali-free glass-ceramics Saurabh Kapoor a, Ashutosh Goel b, Ana Filipa Correia a, Maria J. Pascual c, Hye-Young Lee d,e, Hae-Won Kim d,e, José M.F. Ferreira a,⁎ a

Department of Materials and Ceramics Engineering, University of Aveiro, CICECO, 3810-193 Aveiro, Portugal Department of Materials Science and Engineering, Rutgers, The State University of New Jersey, Piscataway, NJ 08854-8065, United States Instituto de Cerámica y Vidrio (CSIC), Kelsen 5, Campus de Cantoblanco, 28049 Madrid, Spain d Institute of Tissue Regeneration Engineering (ITREN) & College of Dentistry, Dankook University, Cheonan 330714, South Korea e Department of Nanobiomedical Science & BK21 PLUS NBM Global Research Centre for Regenerative Medicine, Dankook University, Cheonan 330714, South Korea b c

a r t i c l e

i n f o

Article history: Received 23 April 2014 Received in revised form 30 July 2014 Accepted 21 April 2015 Available online 22 April 2015 Keywords: Glass-ceramics Thermal properties Sintering Crystallisation Bioactivity

a b s t r a c t The present study reports on the influence of partial replacement of MgO by ZnO on the structure, crystallisation behaviour and bioactivity of alkali-free bioactive glass-ceramics (GCs). A series of glass compositions (mol%): 36.07 CaO–(19.24 − x) MgO–x ZnO–5.61 P2O5–38.49 SiO2–0.59 CaF2 (x = 2–10) have been synthesised by melt–quench technique. The structural changes were investigated by solid-state magic angle spinning nuclear magnetic resonance (MAS-NMR), X-ray diffraction and differential thermal analysis. The sintering and crystallisation behaviours of glass powders were studied by hot-stage microscopy and differential thermal analysis, respectively. All the glass compositions exhibited good densification ability resulting in well sintered and mechanically strong GCs. The crystallisation and mechanical behaviour were studied under non-isothermal heating conditions at 850 °C for 1 h. Diopside was the primary crystalline phase in all the GCs followed by fluorapatite and rankinite as secondary phases. Another phase named petedunnite was identified in GCs with ZnO content N4 mol. The proliferation of mesenchymal stem cells (MSCs) and their alkaline phosphatase activity (ALP) on GCs was revealed to be Zn-dose dependent with the highest performance being observed for 4 mol% ZnO. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Since the discovery of 45S5 Bioglass® in 1971 [1] there has been heightened interest in the science and biomedical applications of bioactive glasses over the last two decades. Many artificial biomaterials based on, or inspired by Hench's glasses have been developed and successfully employed in clinical applications for repairing and replacing parts of the human body. In general, bioactive glasses are used as powders to fabricate porous scaffolds for bone regeneration and tissue engineering [2], or as bioactive glass coatings applied onto metallic implants by plasma spraying [3]. The successful development of such biomedical devices based on bioactive glasses requires high temperature treatments. Thus a good understanding of the factors affecting the sintering and crystallisation behaviours of GCs is essential to draw the expected benefits from the developed materials. It has been demonstrated that Zn, an essential trace element, manifests stimulatory effects on bone formation in vitro and in vivo. In fact, the slow release of Zn incorporated into an implant material promotes ⁎ Corresponding author. E-mail address: [email protected] (J.M.F. Ferreira).

http://dx.doi.org/10.1016/j.msec.2015.04.023 0928-4931/© 2015 Elsevier B.V. All rights reserved.

bone formation around the implant and accelerates patient recovery [4,5]. Moreover, it is well-known that the addition of Zn to silicate glasses can promote the formation of crystalline phases that increase the bending strength of glass ceramics. Considering the beneficial aspects of Zn in bone tissue, the incorporation of Zinc in bioactive glasses and GCs intended for various orthopaedic applications has received special interest in recent years [6–18]. However, despite these very interesting features, a literature survey reveals that most of the zinc-containing bioactive glasses or GC compositions studied until now are either based on – or inspired by – 45S5 Bioglass® [19] and A − W (apatite–wollastonite) GCs [13,14,20–23] and sheds minimal light over the influence of zinc over the structure, sintering behaviour and crystallisation of bioactive GCs. Even though A − W GCs possess much better mechanical properties in comparison to those derived from 45S5 Bioglass® type compositions [24], still there is a dearth of effective bioactive GCs that can exhibit a better balance between good mechanical function along with good bioactive properties. The aim of the present paper is to elucidate about the effect of partial substitution of MgO by ZnO on the structure and bioactivity of glasses and ceramics obtained upon heat treatments. The sintering and crystallisation behaviours were studied by hot-stage microscopy and

S. Kapoor et al. / Materials Science and Engineering C 53 (2015) 252–261

differential thermal analysis, respectively. The aim was to gather insights into the thermal behaviour of glass powders and to discriminate between the densification and crystallisation events. This knowledge is essential for understanding the sintering behaviour and selecting the most suitable thermal schedules for successful fabrication of porous scaffolds or coatings on metallic implants from bioactive glasses [25]. The structural evolutions were assessed through MAS-NMR and X-ray powder diffraction as complementary techniques. 2. Experimental procedure

253

points were obtained from the photographs taken during the HSM experiment. 2.4. Mechanical properties of sintered GCs The mechanical properties were evaluated by measurements of three-point bending strength of rectified parallelepiped bars (3 mm × 4 mm × 50 mm) of sintered GCs (Shimadzu Autograph AG 25 TA, 0.5 mm/min displacement; the mean values and their standard deviation presented have been obtained from measurements from 12 bars).

2.1. Designing of the glass compositions 2.5. Structural transformation in glasses during heat treatment Alkali-free glasses with compositions (mol%): 36.07 CaO–(19.24 − x) MgO–x ZnO–5.61 P2O5–38.49 SiO2–0.59 CaF2 (x = 0–10) have been synthesised by melt–quench technique. The parent glass composition has been designed in the glass forming region of diopside (Di; CaMgSi2O6) – fluorapatite [FA; Ca5(PO4)3F] – tricalcium phosphate (TCP; 3CaO∙P2O5) ternary system [26]. The glass compositions have been labelled as Zn-0 (parent composition), Zn-2, Zn-4, Zn-6, Zn-8 and Zn-10 in accordance with their respective ZnO contents (for example: Zn-2 corresponds to 2 mol% ZnO). 2.2. Materials and methods High purity powders of SiO2 (BDH Chemicals Ltd., UK, purity N99.5%), CaCO3 (BDH Chemicals Ltd., UK, purity N99.5%), MgCO3 (BDH Chemicals Ltd., UK, purity N99.0%), ZnO (Sigma Aldrich, Germany, N99.0%), NH4H2PO4 (Sigma Aldrich, Germany, N99.0%) and CaF2 (Sigma Aldrich, Germany, 325 mesh, N99.9%) were used. Homogeneous mixtures of batches (~100 g) obtained by ball milling, were preheated at 900 °C for 1 h for decarbonisation and then melted in Pt crucibles at 1550 °C for 1 h, depending on the composition of glass. The glasses were obtained in frit form by quenching of glass melts in cold water. The frits were dried and then milled in a high-speed agate mill resulting in fine glass powders with mean particle sizes of ~ 10–20 μm (determined by light scattering technique; Coulter LS 230, Beckman Coulter Fullerton CA; Fraunhofer optical model). The amorphous/crystalline nature of frits was confirmed by X-ray diffraction (XRD) analysis (Rigaku Geigerflex D/Max, Tokyo, Japan; C Series; CuKα radiation; 2θ angle range 10–60°; step 0.02° s−1).

The qualitative analysis of crystalline phases in the GCs (crushed to particle size b25 μm) was made by XRD analysis using a conventional Bragg–Brentano diffractometer (Philips PW 3710, Eindhoven, The Netherlands) with Ni-filtered Cu-Kα radiation. Infrared spectra of the sintered glass powder compacts were obtained using an Infrared Fourier spectrometer (FTIR, model Mattson Galaxy S-7000, USA). For this purpose GC pellets crushed into fine powder were mixed with KBr in the proportion of 1/150 (by weight) and pressed into a pellet using a hand press. 64 scans for background and 64 scans per sample were made with signal gain 1. The resolution was 4 cm−1. 2.6. Apatite forming ability Apatite forming ability of the GCs was investigated by immersion of sintered GC circular discs (green diameter: 10 mm) in 50 ml simulated body fluid (SBF) solution at 37 °C. SBF had an ionic concentration (Na+ 2− 142.0, K+ 5.0, Ca2 + 2.5, Mg2 + 1.5, Cl− 148.8, HPO− 4.2, 4 1.0, HCO3 −1 SO2− 0.5 mmol l ) nearly equivalent to human plasma, as discussed 4 by Tas [27]. The GC–SBF mixtures were immediately sealed into sterilised plastic flasks and were placed in an oven at 37 °C (± 0.5 K) which was agitating in a circular motion at 120 rpm. The sampling took place at different times varying between 1 day and 14 days. The experiments were performed in triplicate in order to ensure the accuracy of results. After each experiment, the GC discs were separated from the liquids and their apatite forming ability on glass powders was followed by SEM and XRD analysis. 2.7. Cell proliferation and alkaline phosphatase activity

2.3. Thermal analysis of glasses The differential thermal analysis (DTA, SetaramLabsys, Setaram Instrumentation, Caluire, France) of glass powder weighing 50 mg was carried out in air from room temperature to 1000 °C at 5 K min−1 and 20 K min−1 in an alumina crucible using α-alumina powder as reference material. The standard deviation values reported for the glass transition temperature (Tg), crystallisation onset temperature (Tc) and peak temperature of crystallisation (Tp) as obtained from DTA are within the range of ±2 °C. The sintering behaviour of the glass powders was investigated by hot-stage microscopy (HSM). A side-view HSM EM 201 equipped with an image analysis system and electrical furnace (1750/15 Leica) was used. The image analyser takes into account the thermal expansion of the alumina substrate while measuring the height of the sample during firing, with the base as a reference. The HSM software calculates the percentage of decrease in height, width and area of the sample images. The measurements were conducted in air at a heating rate of 5 K min− 1. The cylindrical samples with height and diameter of ~ 3 mm were prepared by cold pressing the glass powders. The cylindrical samples were placed on a 10 × 15 × 1 mm3 alumina (N 99.5 wt.% Al2O3) support. The temperature was measured with a Pt/Rh (6/30) thermocouple contacted under the alumina support. The temperatures corresponding to the characteristic viscosity

2.7.1. Mesenchymal stem cell proliferation The biological performance of the as developed GCs was addressed by the in vitro cellular responses, including cell proliferation and osteoblastic differentiation. Zn − 0, Zn − 4 and Zn − 8 have been used as representative samples, and results were compared with the tissue culture plastic used as a control. For the cellular study, mesenchymal stem cells (MSCs) derived from rat bone marrow were used. MSCs have been a potential source for the regenerative therapy of tissues including bone, because of their multipotent and self-renewal capacity without the concern of ethical issues. The experimental procedures were based on our previous work [28] and followed by the guidelines approved by the Animal Ethics Committee of Dankook University. MSCs gathered from the bone marrow of rats were maintained in a normal culture medium containing α-minimal essential medium (MEM) supplemented with 10% foetal bovine serum (FBS), 100 U ml− 1 penicillin and 100 mg ml− 1 streptomycin in a humidified atmosphere of 5% CO2 in air at 37 °C. MSCs maintained up to 3–4 passages were used for cellular study. The GCs were sterilised with 70% ethanol for 1 h prior to seeding cells. For the cell growth study, MSCs were seeded on each sample (15 mm × 2 mm disc type) and then cultured in the normal culture medium condition as described above. After culture for 3, 7 and 14 days, the cell growth level was analysed by the cell counting kit (CCK) assay. At each time of culturing (3 and 7 days), the CCK-8 reagent was

254

S. Kapoor et al. / Materials Science and Engineering C 53 (2015) 252–261

added to each sample and incubated for 3 h at 37 °C. The absorbance was measured at a wavelength of 450 nm using a microplate reader (Molecular Devices, USA). The cell growth morphology was observed by SEM (Hitachi) after fixation the cells with 2.5% glutaraldehyde, dehydration with a graded series of ethanol (50, 70, 90 and 100%) and coating with gold. 2.7.2. Alkaline phosphatase activity The osteoblastic differentiation of the MSCs on the GCs was determined by measuring the alkaline phosphatase (ALP) activity. Cells were seeded on each sample (20 mm × 2 mm disc type), and cultured in the osteogenic medium containing 50 μg ml− 1 sodium ascorbate, 10 mM β-glycerol phosphate, and 10 nM dexamethasone. After culture for 7 and 14 days, cellular samples were gathered and added to the ALP reaction media to allow an enzymatic reaction according to the manufacturer's instructions (Sigma). The quantity of samples used for the reaction was determined based on the total protein content which was determined by using a commercial DC protein assay kit (BioRad). As a result of the reaction, the product p-nitrophenol appears as the colour change which was then measured at an absorbance of 405 nm using a spectrophotometer. The CCK and ALP assays were performed on three replicate samples (n = 3). Data were represented as means ± standard deviations. Statistical analysis was carried out by analysis of variance (ANOVA) and significance level was considered at p b 0.05. 3. Results and discussion 3.1. Glass-forming ability For all the investigated glass compositions (x = 0–10 mol%) melting at 1550 °C for 1 h was sufficient to obtain bubble-free, transparent and XRD-amorphous glasses. 3.2. Thermal analysis The DTA thermographs of glass powders with β = 20 K min−1 are shown in Fig. 1. All the glass compositions feature a single endothermic dip before the onset of crystallisation (Tc) corresponding to Tg, which is followed by the exothermic crystallisation effect. The characteristic parameters obtained from DTA thermographs are summarised in Table 1. In general, addition of ZnO resulted in gradual decrease in Tg from 752 °C to 718 °C. This decline in Tg can be attributed to the low bond energy of Zn–O bond in comparison to the Mg–O bonds [29] due to the weaker cation field strength of Zn in comparison to Mg [30], which in

β = 20 K min

β = 20 K min-1

-1

Tc

Tg

Tp

Exo

Zn-10 Zn-8 Zn-6

ΔT (μV)

ΔT (μV)

Zn-10

Zn-8 Zn-6

Zn-4

Zn-4

Zn-2

Zn-2

Tg

Zn-0

Zn-0

650

750

850

950

Temperature (ºC)

650

700

750

800

Temperature (ºC)

Fig. 1. DTA thermographs of glasses under a heating rate of 20 K min−1.

850

Table 1 Thermal parameters measured from DTA and HSM at 5 K min−1.

TFS (±5 °C) TMS1 (±5 °C) TMS2 (±5 °C) Tga (±2 °C) Tc (±2 °C) Tp (±2 °C) Sc (=Tc − TMS1) A/A0 at TMS1 a

Zn-0

Zn-2

Zn-4

Zn-6

Zn-8

Zn-10

757 820 – 752 850 895 30 0.66

753 809 883 737 843 881 48 0.67

748 797 888 733 852 878 55 0.66

766 817 – 725 843 870 26 0.63

747 797 – 722 837 868 41 0.65

750 795 850 718 822 861 27 0.67

β = 20 K min−1.

turn reduces the rigidity of the glass network, even though both possess similarities in size and charge. This in turn translates into lowering the viscosity of the glasses [31] thus, decreasing their Tc values. Such changes could be observed in glass powder compacts sintered at 800 °C (not shown here) which remained amorphous for the lower zinc contents, and visible XRD peaks appeared in the glasses only with higher zinc contents. Hence, the results obtained in the present work for zinccontaining GCs are in good agreement with other literature reports [13,30,31]. 3.3. Sintering and crystallisation behaviours A comparison between DTA and HSM results using same heating conditions can be useful in investigating the effect of glass composition on sintering and devitrification phenomena. Fig. 2 presents variation in the relative area and heat flow with respect to temperature as obtained from HSM and DTA, respectively, for all the investigated glasses at a similar heating rate of 5 K min−1. Table 1 summarises the values of the temperature of first shrinkage (TFS; log η = 9.1 ± 0.1, η is viscosity in dPa s), temperature for maximum shrinkage (TMS; log η = 7.8 ± 0.1) and ratio of the final area/initial area of the glass powder compact (A/A0) at TMS1, as obtained from the HSM along with temperature for onset of crystallisation (Tc) and peak temperature of crystallisation (Tp), as obtained from the DTA of the glasses. (i) In general, TFS values show a decreasing tendency with increasing content of ZnO. (ii) In agreement with the parent glass (here referred to as Zn-0; [2]), two stage sintering was observed in glasses Zn-2 and Zn-4 while a gradual shift to single stage shrinkage was observed for glasses thereafter (Fig. 2). The conclusion of the first sintering stage is characterised at TMS1 while end of second sintering stage is characterised by TMS2. Interestingly, in all the glass compositions TMS1 was observed at temperatures lower than Tc (i.e., TMS1 b Tc). Thus sintering precedes crystallisation in all the glasses therefore one can expect well sintered and mechanically strong glass powder compacts. (iii) The value for TMS2 was higher than Tc in the glasses exhibiting two stage sintering behaviour, thus depicting that shrinkage continued in the glass powders even after onset of crystallisation, probably due to the existence of residual glassy phase in the GCs. (iv) Table 1 lists the values of sinterability parameter (Sc), [32] where Sc = Tc − TMS. The parameter Sc is the measure of ability of sintering vs. crystallisation: the greater this difference, the more independent are the kinetics of both processes. However, in the present study no particular trend could be observed for the variation of Sc. Nevertheless, all the investigated glasses show larger values of Sc (greater than 25 °C) (26 °C–48 °C) indicating good sintering behaviour. (v) The values of Tp varied between 895 °C–861 °C. As evident from Table 1, the highest Tp value was observed for Zn-0. Adding increasing amounts of ZnO to the parent glass at the expenses of MgO resulted in gradual decreases in Tp.

S. Kapoor et al. / Materials Science and Engineering C 53 (2015) 252–261

Tp

Exo.

0.85 .

0.75

1.05

HSM

TFS

Tp

DTA

DTA

0.85

Tc .

0.75

Tc

0.65

TMS

0.65

TMS1

β = 5 K min-1

β = 5 K min-1

0.55

0.55 650

700

750

800

850

900

650

950

700

750

(d) HSM

TFS Exo.

0.85 0.75 0.65

.

DTA

Tc TMS1

β = 5 K min-1

850

900

950

Zn-8

Tp

TFS Exo.

0.85 DTA

.

0.75

Tc

0.65

0.55 650

HSM

0.95

A/A0

A/A0

0.95

1.05

Zn-6

Tp

ΔT(μV)

1.05

800

Temperature (ºC)

Temperature (ºC)

(c)

Zn-4

Exo.

0.95

ΔT(μV)

0.95

A/A0

(b)

Zn-2

TFS

ΔT(μV)

HSM

ΔT(μV)

1.05

A/A0

(a)

255

TMS1

β = 5 K min-1

0.55 700

750

800

850

900

950

650

700

750

Temperature (ºC) 1.05

HSM

A/A0

900

950

Zn-10

Tp

TFS Exo.

0.95 0.85

850

β = 5 K min-1

0.75

.

DTA TMS1

0.65

TMS2 Tc

0.55 650

ΔT(μV)

(e)

800

Temperature (ºC)

700

750

800

850

900

950

Temperature (ºC) Fig. 2. Comparison of DTA and HSM curves on the same temperature scale for compositions: (a) Zn-2, (b) Zn-4, (c) Zn-6, (d) Zn-8 and (e) Zn-10.

(vi) Values of A/A0 ranged from 0.63 to 0.67 (Table 1) for all the glasses resulting in good densification levels of 95–98% [33].

Scale: 20000 cps

850 oC Zn-10

3.4.1. Crystalline phase evolution by XRD The exact nature of the phases produced during the heat treatment is of great interest in bioactive glasses as they are intended to be often exposed to high temperature during sintering for different biomedical applications [34]. Based on DTA thermographs, Tc was established for all the compositions and used as a criterion for selecting the heat treatment temperature to induce crystallisation. In agreement with HSM and DTA, well sintered and dense GCs were obtained after heat treating the glass-powder compacts at 850 °C for 1 h. Fig. 3 presents the qualitative crystalline phase analysis of the sintered GCs as depicted by XRD data. Thus heat treating at 850 °C for 1 h resulted in well-defined XRD phase reflections in all the compositions indicating their crystalline nature (Fig. 3). Diopside [CaMgSi2O6; ICSD: 10223] and fluorapatite [Ca5(PO4)3F; ICSD: 163790] were the two the primary crystalline phases formed in all the GCs, followed by rankinite [Ca3Si2O7; ICSD: 2282] as secondary phase. Another phase named petedunnite (CaZnSi2O6; ICSD: 158143) crystallised in GCs with ZnO contents N4 mol%. Consequently, with the addition of ZnO for MgO, Zn2 + replaced Mg2 + in

Intensity (a.u.)

3.4. Effect of heat treatment on the structure

Zn-8 Zn-6 Petedunite Zn-4 Zn-2 Zn-0 Rn FA Di

20

25

30

35

40

Fig. 3. X-ray diffractograms of GCs heat treated at 850 °C for 1 h.

256

S. Kapoor et al. / Materials Science and Engineering C 53 (2015) 252–261

the diopside, thus forming the major crystalline phase with higher intensity than fluorapatite. The presence of fluorapatite phase is highly desirable as fluorapatite is more stable than hydroxyapatite. Therefore, GCs containing fluorapatite phase would be recommended when coatings having low dissolution rate are required [35]. Hence, with the addition of ZnO to the parent composition the intensity of XRD peaks corresponding to diopside increased whereas the intensity of fluorapatite peaks remained almost constant throughout the series.

3.4.2. MAS-NMR of glass-ceramics The XRD data was further supported by MAS-NMR studies (Fig. 4). The full width half maximum has reduced considerably in comparison to the original glasses [36]. Further, as shown in Fig. 4a, all the spectra have a sharp peak centred ~− 84.4 ppm and a small shoulder at − 74.5 ppm, which becomes more pronounced with the increase of zinc content in glasses, where the former is close to the pure diopside [37] while the latter corresponds to the presence of minor phase rankinite [37]. In our study of XRD a total of two silicate phases pertaining to chain silicates were observed in the GCs. However, discrimination between these phases is difficult due to the similarity in their silica connectivity. Further, the presence of a residual glassy phase in the GCs gives rise to a certain amount of line broadening. Therefore the resonance around −84 ppm can be due to the contribution from diopside and petedunnite (ionosilicates). Fig. 4b shows 31P NMR spectra of glass ceramics heat treated at 850 °C for 1 h. All the spectra have a sharp peak at 3.3–3.5 ppm corresponding to an orthophosphate environment and a small shoulder at around − 5.3 and −14.6 ppm indicating the presence of pyrophosphate units. The band at 3.5 ppm is close to that reported for fluorapatite [38]. Further, the formation of pyrophosphate phase around −5.3 and −14.6 ppm may be due to the reduction in the charge balancing cations, which migrated towards their respective crystalline phases upon heat treatment. Further, diopside being a mechanically strong phase renders high mechanical strength to the resultant GCs. It is worth noting that the presence of petedunnite as a minor phase indicates that zinc predominantly remains in the residual glassy phase. The residual glassy phase being Zn enriched with respect to the GCs will enhance the release of Zn2+ which may be important for osteoblast differentiation [39], increasing osteoblast DNA content [8] and inhibiting osteoclastic bone resorption in vitro [40].

(a)

3.5. FTIR of glasses and glass-ceramics The room temperature FTIR transmission spectra are shown in Fig. 5. The spectra of all the investigated glasses Fig. 5a show three broad bands in the region ranging from 350–1300 cm− 1. The broadness clearly shows the disorder in the glass structure with uneven distribution of Qn units. The band ranging from 800–1300 cm − 1 corresponds to the presence of two optical modes of Si–O–Si groups; Si–O bending mode identified around 800 cm− 1 and the asymmetric stretching mode Si − O(s) in 1000–1300 cm− 1 region. The transmission band for different Qn units (n = 1, 2, 3, 4) are centred around 1200, 1000, 950, 900 and 850 cm− 1 respectively. Further in this region (800–1300 cm− 1) the band centred at 1041 cm− 1 (Zn-0) indicating the distribution of Qn around Q3 units [41–44]. Further, this band corresponding to Si–O stretching mode slightly shifts towards lower wave number pointing towards some rearrangement of the glass network compared with parent glass, as substitution of ZnO for MgO leads to decrease in viscosity thus making the structure less rigid [31,45]. It should be noted that the high frequency band at 1035 cm− 1 may also be attributed to asymmetric stretching of PO4 units which may be due to the presence of P2O5 in the investigated glasses [46]. The ~500 cm−1 band can be attributed to Si–O–Si bending modes [47], while the weak shoulder around 740 cm−1 may be due to Si–O–Si symmetric stretching with simultaneous Si cation motions [48]. Further, the band centred around 559 cm−1 corresponds to P–O bending modes [49]. The FTIR spectra recorded for all the glass powder compacts sintered at 850 °C displayed in Fig. 5b are in clear agreement with the XRD data. The broad peaks in the glasses decompose into narrow regions thus indicating the crystalline nature of the samples which increases with increase in temperature. According to Omori et al. [47], the absorption bands at 1070 cm− 1 and 965 cm− 1 correspond to the stretching vibrations of Si–O (B) (B refers to bridging) and bands at 920 and 865 correspond to the Si–O (NB) (NB refers to nonbridging) therefore the bands around ~ 1068, 964–968, 916–919, and 862–867 cm − 1 may be assigned to the bands as described by Omori [47]. The bands appearing around ~560 and 600 cm−1 may be attributed to the PO3− 4 groups present in the structure, the bands at around 600 corresponding to the P–O bend vibrations in a crystal, and the band at around 560 cm−1 being attributed to the P–O bending vibrations in glass as described earlier. The ~507 cm−1 band can be assigned to the Si–O–Si bending modes [47].

(b)

-84.4

3.5 -74.4

Zn-8

-14.6 -5.6 Zn-8

Zn-4

Zn-4

Zn-0

Zn-0

-200

-150 29

-100

-50

0

Si Chemical Shift (ppm)

50

-40

-20

0

20

31

P Chemical Shift (ppm)

Fig. 4. 29Si MAS-NMR (a) and 31P MAS-NMR (b) spectra of GCs heat treated at 850 °C for 1 h.

40

S. Kapoor et al. / Materials Science and Engineering C 53 (2015) 252–261

(a)

(b) 740

Zn-8

1035 507 509 397

1031

Zn-6

1027

Zn-4

501

Zn-2

1029 399

499

1033

493

350

673 507 634 703 385 605 468 563 709

Zn-10

Transmittance (%)

559 505

Transmittance (%)

257

950

Zn-8

Zn-4 513 470

Zn-0

1250

964 1068 867

Zn-6

916 864 966

Zn-2

919

Zn-0

862 968

1041

650

850 ºC Zn-10

350

Wavenumber (cm-1)

650

950

Wavenumber

1250 (cm-1)

Fig. 5. FTIR spectra of all the investigated glasses (a), and corresponding GCs obtained upon sintering glass powder compacts at 850 °C for 1 h (b).

3.6. Density and molar volume The experimental results showed that substitution of MgO by ZnO caused an increase in the density (Table 2) of the glasses due to the higher density of ZnO (5.6 g cm−3) with respect to MgO (3.6 g cm−3). Further, the molar volume Vm was calculated using the apparent density data for the bulk glasses using following relations: M : ρ

3.8. Apatite forming ability of GCs

Here M is the molar mass of the glass and ρ is the apparent density of the bulk glasses. The molar volume of the glasses increased with increase in Zn content in glasses. The increase in molar volume can be attributed to the small bond strength of Zn–O bond in comparison to Mg– O bond which results in higher bond lengths thus causing an expansion in the network. 3.7. Mechanical behaviour Evaluating the mechanical strength of a biomaterial is very important to judge its suitability for a given intended application, such as in the case of scaffolds for bone regeneration. The mechanical performance of a glass-ceramic material is a complex function of several concurrent factors such as composition, sintering temperature, and the resulting phase assemblage. It is likely that mechanical strength of a given glass-ceramic will be dependent on the mechanical properties of the constituting crystalline phases, their volume fractions, the size/shape of the crystals formed, and on the interfacial bonding between the amorphous and crystalline phases. The flexural strength was calculated for powder compacts sintered at 850 °C. High flexural strength values were obtained for glass powder compacts, which can be attributed the predominance of diopside in all GCs (Fig. 6). In general, an increase in Table 2 Density and molar volume of the investigated glasses. Glasses

Density (g cm−1)

Molar volume

Zn-0 Zn-2 Zn-4 Zn-6 Zn-8 Zn-10

2.92 2.95 2.98 3.02 3.05 3.08

20.40 20.47 20.51 20.55 20.59 20.66

The apatite formation at the surface of samples immersed in simulated body fluid (SBF) is a relatively simple and common method to evaluate the biomineralisation activity in vitro. However, the preparation method is tricky and there are some reservations regarding the reliability of results obtained from SBF tests in the literature [50]. Despite the lack of consensus, a great number of correlations have been attempted between the bio-mineralisation ability of a given bioactive material assessed by this method and its performance in vivo [51,52]. The SEM-EDX results of the GC samples before and after immersion in SBF solution for different time periods are shown in Fig. 7. The surface morphology of Zn-0, Zn-4 and Zn-8 samples before immersion in SBF

230 210

Flexural stength (MPa)

Vm ¼

mechanical strength with increasing Zn contents in glasses was observed, with the highest mechanical strength being measured for the GC Zn-8. As deduced from HSM and DTA analysis, this enhancement can be attributed to two complementary reasons: (i) an improved sintering ability due to a decrease in the viscosity of glasses translated by decreasing T g values; and (ii) a more extensive subsequent crystallisation promoted by the enhanced diffusivity of ionic species in the glassy matrix.

190 170 150 130 110 90 70 50 0

2

4

6

8

10

ZnO (mol. %) Fig. 6. Flexural strength of the GCs heat treated at 850 °C for 1 h.

258

(a)

(b)

7 days

14 Days

(a) Zn-0

Zn-0 5µm

10µm

(b)

Zn-4

5 µm

Zn-4 10µm

Zn-8

10µm

(c) 5 µm

Zn-8 10µm

10µm

Fig. 7. SEM/EDS micrographs images of unpolished and non-etched GCs samples Zn-0, Zn-4, Zn-8. (a) Before immersion and (b) after immersion in SBF for 7 and 14 days.

S. Kapoor et al. / Materials Science and Engineering C 53 (2015) 252–261

10 µm

S. Kapoor et al. / Materials Science and Engineering C 53 (2015) 252–261

259

obvious. Various reasons can be cited for this delayed biomineralisation process in ZnO-containing GCs: 1. Presence of ZnO: The delayed formation of HA can be due to intertetrahedral non-covalent linkages in the residual glassy phase. Similar results have been reported in our previous work [36]. Further, the concomitant presence of Mg2 + and Zn2 + inhibits HA crystal growth. The literature reports suggest a Langmuir-type adsorption of these ionic species at active growth sites on the HA crystals [53]. 2. Low residual glassy phase: The enhanced crystallisation of ZnOcontaining GCs (lower Tc values) and the consequent lower amount of residual glassy phase is also expected to decrease the apatite forming ability [54]. 3.9. Mesenchymal cell activity and alkaline phosphatase activity

Fig. 8. X-ray diffractograms of GC Zn-0 heat before immersion and after immersion in SBF for 7 and 14 days (HA = hydroxyapatite; FA fluorapatite). .

(Fig. 7a) is similar, a situation that was altered for the samples immersed in SBF (Fig. 7b). The EDX data was gathered over the entire areas shown in the SEM micrographs to assure reliable averaging of the signals and the scale of all EDX spectra was always kept constant to enable a direct comparison. Although all the spectra reveal the presence of all glass-constituting elements, their relative intensities changed significantly after immersing the samples in SBF. It can be seen that the EDX spectrum of the Zn-0 sample (Fig. 7a) significantly differs from those obtained after immersing in SBF (Fig. 7b). Namely, the intensities of Ca and P became more and more predominant with increasing immersion time. This constitutes an obvious evidence of apatite formation at the surface of the samples. Fig. 8 compares the XRD patterns of Zn-0 GC before and after immersion in SBF for 7 and 14 days. A clear evolution of the XRD spectra towards that of HA can be observed as immersion time increases. The close proximity of the lattice parameters and of the XRD lines of fluorapatite (FA), already formed in the glass-ceramics, and the newly formed HA makes it difficult to distinguish both crystalline phases. The formation of HA at the surface of ZnO-containing GCs is less

Fig. 9 shows the cell viability on the GCs and control during culture for up to 7 days, as assessed by a CCK method. Apparently, the growth kinetics of GCs with 4 mol% ZnO was significantly higher with respect to that of tissue culture plastic control as well as to that of other investigated GCs. Interestingly, all other zinc containing GCs showed lower cell viability in comparison to that for tissue culture plastic. The representative cell growth images at day 3 and day 7 on the GC samples are presented in Fig. 10. The cells were not readily noticed on the Zn-0 glass-ceramic after 3 days while the cell proliferation was comparatively better on GC Zn-8. On the other hand, when grown on Zn-4, the cells were readily observed covering the surface almost completely with active cytoskeletal processes and elongated filopodia indicating that GC Zn-4 provides favourable conditions for adherence, spreading and proliferation of MSCs. After 7 days, although the cell proliferation on GCs other than Zn-4 was better in comparison to that observed after day 3 but still the GC Zn-4 exhibited the highest rate of cell proliferation. The lower cell viability in the absence of zinc can be explained by the fact that after iron, zinc is the most abundant trace metal in human organisms and is essential for various metabolic activities as has been discussed above. However, the decreasing cell viability with increasing zinc concentration in glasses may be attributed to its dose-dependent effect. Zn2 + has been reported to play a dual role in effecting cell death: although zinc seems to be an inhibitor of many forms of apoptosis [55], exposure to excessive concentrations may contribute to neuronal cell death in acute neurological disorders and to apoptosis of human peripheral blood lymphocytes [56]. The mechanism by which Zn2+ induces cell death and oxidative stress is still unresolved. According to

Fig. 9. Influence of ZnO content in GCs on cell viability during culture for up to 7 days, as assessed by CCK method. GCs with 4 mol% ZnO showed highest growth with respect to tissue culture plastic control as well as to that of other investigated GCs.

260

S. Kapoor et al. / Materials Science and Engineering C 53 (2015) 252–261

Zn-0 (3d)

Zn-4 (3d)

100 μm

Zn-0 (7d)

Zn-8 (3d)

100 μm

Zn-4 (7d)

100 μm

100 μm

Zn-8 (7d)

100 μm

100 μm

Fig. 10. SEM images of the MSCs grown on the sintered glass powder compacts (Zn-0, Zn-4 and Zn-8) during culture for 3 and 7 days. After 7 days, cell proliferation on GCs was better in comparison to that observed after 3 days. GC Zn-4 exhibited the highest rate of cell proliferation.

the available literature, an increase in Zn2+ concentration in cellular medium higher than ~ 2 ppm can lead to cytotoxicity and cell death [57]. Similar results with respect to proliferation of endothelial cells and osteoblasts have also been reported by Aina et al. [12,58] where it has been shown that 45S5 Bioglass® doped with 5 wt.% ZnO enhanced cell proliferation on its surface while increasing zinc concentration led to significant release of lactate dehydrogenase (index of cytotoxicity) thus, resulting in cell death. With respect to the bioactivity of GC Zn-4, the cell growth kinetics in this sample is significantly higher in comparison to that reported for apatite–wollastonite [59] as well as diopsidefluorapatite [60] GCs, respectively. The osteogenic differentiation of the MSCs cultured on the GC samples was investigated in terms of ALP activity (Fig. 11). Under the osteogenic medium used herein, rat bone marrow MSCs have been shown to switch to the lineage of osteoblastic cells [28]. When cultured on the control dish, MSCs showed an increased ALP level from days 7 to

14. The ALP levels for all the investigated GCs were higher in comparison to control irrespective of their zinc content. However, among all the investigated GCs, Zn-4 exhibited the highest ALP activity of cells during 7 and 14 days of culture. It should be noted that ALP is a homodimeric metalloenzyme that catalyses the hydrolysis of almost any phosphor monoester with the release of phosphate and alcohol. Its properties have been associated with the mineralisation process and phosphate homeostasis in bone tissue [61]. ALP acts as a nucleation centre for hydroxyapatite as it is able to release phosphate and capture calcium ions. In this way, the presence of ALP and a solution rich in Ca2+ ions induces the deposition of hydroxyapatite (bone forming mineral) crystals and amorphous calcium phosphate. Therefore, the significant improvement in ALP level of MSCs indicates that the Zn-GCs are proper materials to stimulate osteogenic differentiation and to possibly enhance synthesis of bone extracellular matrices, a hypothesis that remains as further in-depth study.

Fig. 11. Alkaline phosphatase activity of the MSCs during culture for 7 and 14 days on the glass compacts (Zn-0 to Zn-8) and on tissue culture plastic control. For all the investigated GCs, irrespective of their ZnO content, ALP levels were always higher in comparison to that of control (p b 0.05, ANOVA, n = 3).

S. Kapoor et al. / Materials Science and Engineering C 53 (2015) 252–261

4. Conclusions A detailed investigation about the influence of adding increasing amounts of ZnO to partially replace MgO in a parent bioactive glass composition was shown to have consequences in terms of sintering and crystallisation behaviours and on the relevant properties of the resulting GCs, including mechanical and biological responses. The glass forming ability of the glasses was unaffected by the level of MgO substitution. All the glass compositions investigated exhibited good sintering ability upon heat treating at 850 °C for 1 h thus, resulting in well sintered and mechanically strong GCs with diopside and fluorapatite as the primary crystalline phases. A minor crystalline phase (petedunnite) was also formed for ZnO N 4 mol%. The ZnO content was also revealed to play an essential role on the in vitro bioactivity. The GC composition with ZnO = 4 mol% exhibited the highest levels of mesenchymal cell proliferation and alkaline phosphatase activity, while further increasing ZnO contents led to a significant decrease in the in vitro performance of the investigated GCs. Acknowledgements The support received from CICECO, University of Aveiro, and from the Foundation for Science and Technology (FCT) of Portugal in the frame of the Project Reference PTDC/CTM/99489/2008 is acknowledged. This study was also partially supported by a research grant from Priority Research Centers Program (2009-0093829), National Research Foundation, South Korea. References [1] L.L. Hench, R.J. Splinter, W.C. Allen, T.K. Greenlee, J. Biomed. Mater. Res. 5 (1971) 117–141. [2] Q. Fu, E. Saiz, M.N. Rahaman, A.P. Tomsia, Mater. Sci. Eng. C 31 (2011) 1245–1256. [3] L. Pawlowski, The Science and Engineering of Thermal Spray Coatings, John Wiley & Sons Ltd, England, 2008. [4] P. Zorrilla, L.A. Gomez, J.A. Salido, A. Silva, A. Lopez-Alonso, Wound Repair Regen. 14 (2006) 119–122. [5] M.S. Agren, Acta. Derm. Venereol. Suppl. (Stockh.) 154 (1990) 1–36. [6] G. Lusvardi, D. Zaffe, L. Menabue, C. Bertoldi, G. Malavasi, U. Consolo, Acta Biomater. 5 (2009) 419–428. [7] G. Lusvardi, G. Malavasi, L. Menabue, M.C. Menziani, A. Pedone, U. Segre, V. Aina, A. Perardi, C. Morterra, F. Boccafoschi, S. Gatti, M. Bosetti, M. Cannas, J. Biomater. Appl. 22 (2008) 505–526. [8] S. Murphy, D. Boyd, S. Moane, M. Bennett, J. Mater. Sci. Mater. Med. 20 (2009) 2207–2214. [9] S. Murphy, A.W. Wren, M.R. Towler, D. Boyd, J. Mater. Sci. Mater. Med. 21 (2010) 2827–2834. [10] R.L. Du, J. Chang, J. Inorg. Mater. 19 (2004) 1353–1358. [11] R.L. Du, J. Chang, Bio-Med. Mater. Eng. 16 (2006) 229–236. [12] V. Aina, A. Perardi, L. Bergandi, G. Malavasi, L. Menabue, C. Morterra, D. Ghigo, Chem. Biol. Interact. 167 (2007) 207–218. [13] M. Kamitakahara, C. Ohtsuki, H. Inada, M. Tanihara, T. Miyazaki, Acta Biomater. 2 (2006) 467–471. [14] R.L. Du, J. Chang, S.Y. Ni, W.Y. Zhai, J.Y. Wang, J. Biomater. Appl. 20 (2006) 341–360. [15] V. Aina, F. Bonino, C. Morterra, M. Miola, C.L. Bianchi, G. Malavasi, M. Marchetti, V. Bolis, J. Phys. Chem. C 115 (2011) 2196–2210.

261

[16] H. Oudadesse, E. Dietrich, Y.L. Gal, P. Pellen, B. Bureau, A.A. Mostafa, G. Cathelineau, Biomed. Mater. 6 (2011). [17] A.J. Salinas, S. Shruti, G. Malavasi, L. Menabue, M. Vallet-Regi, Acta Biomater. 7 (2011) 3452–3458. [18] A. Karamanov, M. Pelino, J. Eur. Ceram. Soc. 28 (2008) 555–562. [19] G. Malavasi, G. Lusvardi, A. Pedone, M.C. Menziani, M. Dappiaggi, A. Gualtieri, L. Menabue, J. Phys. Chem. A 111 (2007) 8401–8408. [20] G. Da Li, D.L. Zhou, T.H. Pan, G.S. Chen, Y. Lin, M. Mao, G. Yan, Biomed. Mater. 4 (2009). [21] G. Li, D. Zhou, M. Xue, W. Yang, Q. Long, B. Cao, D. Feng, Appl. Surf. Sci. 255 (2008) 559–561. [22] G.D. Li, D.L. Zhou, D.G. Feng, M. Mao, Y. Guo, S.M. Li, J. Inorg. Mater. 23 (2008) 621–625. [23] M. Mao, Y. Lin, D.L. Zhou, G.D. Li, Y. Guo, G.S. Chen, T.H. Pan, J. Clin. Rehab. Tissue Eng. Res. 12 (2008) 8863–8866. [24] S. Yang, K.F. Leong, Z. Du, C.K. Chua, Tissue Eng. 7 (2001) 679–689. [25] A.R. Boccaccini, Q. Chen, L. Lefebvre, L. Gremillard, J. Chevalier, Faraday Discuss 136 (2007) 27–44. [26] A. Goel, S. Kapoor, R.R. Rajagopal, M.J. Pascual, H.W. Kim, J.M.F. Ferreira, Acta Biomater. 8 (2012) 361–372. [27] A.C. Tas, Biomaterials 21 (2000) 1429–1438. [28] S.A. O., S.H. K., J.E. W., J.J. K., U.S. S., H.W. K., J. Tissue Eng. 2010 (2010) 475206. [29] Y.-R. Luo, Comprehensive Handbook of Chemical Bond Energies, CRC Press, 2010. [30] S. Toyoda, S. Fujino, K. Morinaga, J. Non-Cryst. Solids 321 (2003) 169–174. [31] R.H. Doremus, John Wiley & Sons, Inc., New York, 1994, pp. 105. [32] C. Lara, M.J. Pascual, A. Durán, J. Non-Cryst. Solids 348 (2004) 149–155. [33] M.J. Pascual, A. Duran, M.O. Prado, Phys. Chem. Glasses 46 (2005) 512–520. [34] D. Bellucci, V. Cannillo, A. Sola, Sci. Sinter. 42 (2010) 307–320. [35] C. Jana, M. Braun, Biomaterials 17 (1996) 2065–2069. [36] A. Goel, S. Kapoor, A. Tilocca, R.R. Rajagopal, J.M.F. Ferreira, J. Mater. Chem. B 1 (2013) 3073–3082. [37] Chapter 4 29Si NMR, in: J.D.M. Kenneth, E.S. Mark (Eds.), Pergamon Materials Series, Pergamon 2002, pp. 201–268. [38] H.A. Abo-Mosallam, R.G. Hill, N. Karpukhina, R.V. Law, J. Mater. Chem. 20 (2010) 790–797. [39] J.R. Popp, B.J. Love, A.S. Goldstein, J. Biomed. Mater. Res. 81A (2007) 766–769. [40] A. Ito, H. Kawamura, M. Otsuka, M. Ikeuchi, H. Ohgushi, K. Ishikawa, K. Onuma, N. Kanzaki, Y. Sogo, N. Ichinose, Mater. Sci. Eng. C 22 (2002) 21–25. [41] S. Agathopoulos, D.U. Tulyaganov, J.M.G. Ventura, S. Kannan, A. Saranti, M.A. Karakassides, J.M.F. Ferreira, J. Non-Cryst. Solids 352 (2006) 322–328. [42] Y. Tsunawaki, N. Iwamoto, T. Hattori, A. Mitsuishi, J. Non-Cryst. Solids 44 (1981) 369–378. [43] W.L. Konijnendijk, J.M. Stevels, J. Non-Cryst. Solids 21 (1976) 447–453. [44] L. Stoch, M. Środa, J. Mol. Struct. 511–512 (1999) 77–84. [45] S.I. Lara-Cabrera, D.M. Spooner, Plant Syst. Evol. 248 (2004) 129–142. [46] Q. Williams, E. Knittle, J. Phys. Chem. Solids 57 (1996) 417–422. [47] K. Omori, Am. Mineral. 56 (1971) 1607–1616. [48] F.L. Galeener, P.N. Sen, Phys. Rev. B 17 (1978) 1928–1933. [49] O. Peitl, E.D. Zanotto, L.L. Hench, J. Non-Cryst. Solids 292 (2001) 115–126. [50] M. Bohner, J. Lemaitre, Biomaterials 30 (2009) 2175–2179. [51] T. Kokubo, H. Takadama, Biomaterials 27 (2006) 2907–2915. [52] S. Fujibayashi, M. Neo, H.M. Kim, T. Kokubo, T. Nakamura, Biomaterials 24 (2003) 1349–1356. [53] L. Wang, G.H. Nancollas, Chem. Rev. 108 (2008) 4628–4669. [54] P. Li, Q. Yang, F. Zhang, T. Kokubo, J. Mater. Sci. Mater. Med. 3 (1992) 452–456. [55] F.G. Chai, A.Q. Truong-Tran, L.H. Ho, P.D. Zalewski, Immunol. Cell Biol. 77 (1999) 272–278. [56] D.W. Choi, J.Y. Koh, Annu. Rev. Neurosci. 21 (1998) 347–375. [57] A. Hoppe, N.S. Guldal, A.R. Boccaccini, Biomaterials 32 (2011) 2757–2774. [58] V. Aina, G. Malavasi, A.F. Pla, L. Munaron, C. Morterra, Acta Biomater. 5 (2009) 1211–1222. [59] H. Zhang, X.J. Ye, J.S. Li, Biomed. Mater. 4 (2009). [60] I. Kansal, A. Goel, D.U. Tulyaganov, M.J. Pascual, H.Y. Lee, H.W. Kim, J.M.F. Ferreira, J. Mater. Chem. 21 (2011) 16247–16256. [61] E. Verne, S. Ferraris, C. Vitale-Brovarone, S. Spriano, C.L. Bianchi, A. Naldoni, M. Morra, C. Cassinelli, Acta Biomater. 6 (2010) 229–240.