Materials Science and Engineering C 32 (2012) 1531–1535
Contents lists available at SciVerse ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Preparation and characterization of a titanium bonding porcelain Litong Guo a,⁎, Yao Shi a, Lizhi Guo b, Qian Zhang c, Junlong Tian a, Yabo Zhu a, Tianwen Guo c a b c
School of Materials Science and Engineering, China University of Mining and Technology, Jiangsu Xuzhou 221116, P.R. China Xi'an Shiyou University, Xi'an 710065, P.R. China The Fourth Military Medical University, Xi'an 710032, P.R. China
a r t i c l e
i n f o
Article history: Received 30 March 2011 Received in revised form 2 February 2012 Accepted 20 April 2012 Available online 28 April 2012 Keywords: Dental material Titanium Bonding porcelain Cytotoxicity
a b s t r a c t The titanium bonding porcelain was synthesized through normal melting-derived route using borate–silicate system. The porcelain was characterized by thermal expansion, X-ray diffraction, scanning electron microscope and cytotoxicity tests. The results of X-ray diffraction showed that the main phase of the bonding porcelain was SnO2. The SnO2 microcrystals precipitated from the glass matrix when the SnO2 content was increased. The thermal expansion coefficient of bonding porcelains decreased with the increasing concentration of SiO2. The thermal expansion coefficient of bonding porcelains first decreased slightly with the increasing of B2O3 concentration (from 0 wt% to 10 wt%) and then increased to about 9.4 × 10− 6/°C(from 10 wt% to 12 wt%). As an intermediate, B2O3 can act as both network formers and modifiers, depending on the relationship between the concentration of basic oxides and intermediates. The Vickers hardness of bonding porcelains increased with the increase of SnO2 concentration. When SnO2 concentration was 6 wt%, only Si and Sn elements attended the reaction between titanium and porcelain and mainly adhesive fracture was found at Ti-porcelain interface. When SnO2 concentration was 12 wt%, failure of the titanium–porcelain predominantly occurred in the bonding porcelain and mainly cohesive fracture was found at Ti-porcelain interface. The methyl thiazolyl tetrazolium assay results demonstrated that the cytotoxicity of the titanium porcelain was ranked as 0. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Titanium is an attractive dental restorative material for its excellent characteristics such as excellent biocompatibility, corrosion resistance, light weight, and high strength and low cost [1,2]. However, inferior bonding in titanium–porcelain systems compared to the conventional metal-porcelain systems is still a major problem for its application [3]. The bonding of porcelain to an alloy is attributed to van-derWaal's forces, to a mechanical interlocking between both materials, and to chemical bonds between the porcelain and an oxide layer, which is built during the firing process on the surface of the alloy by oxidation of the base metals [4]. The main factors that affect the titanium-porcelain bond are (1) growth of an oxide layer on titanium at elevated temperatures, (2) adherence of the self-formed oxide to the Ti substrate, and (3) bonding of the self-formed oxide with the porcelain [4]. It has been proposed that the poor bonding strength between porcelain and titanium was partly because of continual oxidation of titanium during the porcelain fusing and formation of a nonadherent oxide layer [5]. ⁎ Corresponding author at: School of Materials Science & Engineering, China University of Mining and Technology, Xuzhou 221116, P.R. China. Tel.: +86 516 83591979; fax: +86 516 83591870. E-mail address: [email protected] (L. Guo). 0928-4931/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.msec.2012.04.038
Ti has a great affinity for oxygen, and relatively thick and nonadherent layers of titanium oxide tend to form at above 800 °C; therefore, porcelain should be fired below this temperature [4,5]. Special low fusing dental porcelains should be developed for titanium– porcelain bonding. For high temperature fusion, the coefficient of thermal expansion of the porcelains is important during the cooling process. The coefficient of thermal expansion of the porcelains developed should match with that of titanium (Δα b l × 10 − 6/°C). Therefore, in this work, a new Ti bonding porcelain was developed and the effect of composition on thermal expansion coefficient was also investigated. Because dental porcelains may be in contact with oral soft tissues for periods of time, assessment of dental material biocompatibility is gaining increasing importance for both patients and dentists. Clinical trials of dental materials are now clearly regulated and may only be initiated after successful completion of nonclinical evaluations and biocompatibility studies according to European regulations and ISO guidelines [6–8]. The aim of present investigation is to develop a new Ti bonding porcelain and characterize its properties such as phase structure, thermal expansion coefficient, mechanical properties and cytotoxicity [8–11]. At the same time, understanding the reactions between porcelain and titanium is important to elucidate the bonding mechanism. Therefore, the failure modes at the interface were investigated to gain insight into the possible mechanisms of Ti-porcelain bonding.
1532
L. Guo et al. / Materials Science and Engineering C 32 (2012) 1531–1535
modulus of titanium (ETi = 105.4 GPa). The coefficient K [8] can be expressed as:
2. Experimental 2.1. Preparation
2
K ¼ 54:78 dTi −73:15 dTi þ 27:65 ¼ 4:78 The titanium bonding porcelain was prepared by melting the corresponding chemicals (such as SiO2, H3BO3, Al2O3, Na2CO3 and K2CO3) for 2 h at 1450 °C. The cast glass was crushed and milled in an agate planetary mill, using a speed of rotation of 1500 rpm for 6 h. The titanium–porcelain test specimens were prepared for a three point bending test as specified in ISO 9693 [8]. Then GG Titanium Porcelain (self-made, Xi'an Jiaotong University, China) was fused in a Multimat 99 furnace (Dentsply, American) according to the manufacturers' instructions. A thin layer of bonding porcelain, opaque porcelain and dentin porcelain was fired at the central areaof the specimen, respectively. The porcelain built up was about 8 × 3 × 1 mm 3. The thickness of the bonding porcelain and opaque porcelain together was 0.4 ± 0.02 mm. The total thickness of the fired porcelain was 1 mm. After the dentin porcelain firing, specimens were ultrasonically cleaned in deionized water for 5 min and dried in air. 2.2. Characterization 2.2.1. Thermal expansion analysis The plate-shaped Ti porcelain specimens of the dimensions 25 mm × 4 mm × 4 mm were prepared. The thermal expansion properties were recorded using a computerized thermal dilatometer (NETZSCH-DIL 402 C, Germany) with a heating rate of 5 °C/min. Data were obtained from room temperature to 600 °C. For each specimen, three measurements were made and reproducible results were obtained (±5%). The theoretical thermal expansion coefficient of the bonding porcelain was calculated according to the following formula: α ¼ α 1 P1 þ α 2 P2 þ α 3 P3 þ ……
ð1Þ
where P1, P2 and P3 are the oxide concentration of porcelain. The α1, α2 and α3 are the calculating coefficient of the oxide, which was shown in Table 1 [9]. 2.2.2. Three-point bending test The three-point bending test was performed by a universal mechanical testing machine (DSS–25 T, Shimadzu, Japan) with a span distance of 20 mm. The specimens were placed in the testing machine with the porcelain positioned symmetrically on the side opposite to the applied load. The force was applied at a crosshead speed of 1.5 mm/min with the porcelain side opposite the center support until fracture. The three-point bonding strength was calculated according to the formula in ISO 9693 [8]. The bonding strength, τb, can be expressed as: τb ¼ K Ffail
ð2Þ
where Ffail is the applied load at failure, and K is a function of the thickness of the titanium (dTi = 0.5 mm), and the value of Young's
ð3Þ
For each group, a set of six samples was chosen by random sampling. The results were analyzed by one‐way ANOVA and Student– Newman–Kuels test at α= 0.05 using statistical software (SPSS 10.0 for Windows). 2.2.3. Vickers hardness test Vickers hardness testing was performed on an HV-5 hardness tester with an applied load of 5 kg for 15 s. For every sample, both diagonal lengths of each indentation were optically measured in the hardness apparatus. Vickers hardness was calculated by the following formula: 2
HV ¼ 1:854 P=d
ð4Þ
where HV is the Vickers hardness, P is the applied load, and d is the mean length of the two diagonal lines of the indentation. For each group, a set of six indentations was used. The results were analyzed by one‐way ANOVA and Student–Newman–Kuels test at α = 0.05 using statistical software (SPSS 10.0 for Windows). 2.2.4. XRD and SEM analysis XRD was performed at room temperature using Cu Kα radiation (λ = 1.5418 Å) on an X-ray diffractometer (D/max-r, Rigaku Corp., Japan). The acceleration voltage was 60 kV with an 80 mA current flux. Data were collected for 2θ in the range of 10–90° employing a step size of 0.02°, with a counting rate of 10.0°/min. The titanium surface and Ti–porcelain bonding interface were examined on a scanning electron microscope (JSM–6460, JEOL, Japan) coupled EDS apparatus (INCAX–sight, Oxford, England). 2.3.5. Cytotoxicity test The cytotoxicity properties of the samples were evaluated by methyl thiazolyl tetrazolium (MTT) assay in association with L929 mouse fibroblasts [11–15]. The extract solutions were made by immersing the samples (Φ 5 mm × 1 mm) in culture medium of Roswell Park Memorial Institute (RPMI) media 1640 at approximately 0.55 mL/cm 2, 37 °C for 72 h. Then the extract solution replaced the medium used in the test group. The L929 mouse fibroblast cells were collected and resuspended at 5 × 10 4 cell/mL. Cells were incubated into 96-well culture clusters, 100 μL in each well, and kept at 37 °C in a fully humidified atmosphere at 5% CO2 in air for 24 h. The medium in negative and positive controls was replaced by fresh culture medium and diluted hydroxybenzene solution (10 wt%), respectively. Cell proliferation was determined after being treated for 1, 3, 5 and 7 days. The medium was removed and 20 μL of MTT solution (2 mg/mL) was added to each well. After incubation at 37 °C for 3 h in a fully humidified atmosphere at 5% CO2 in air, the untransformed MTT was removed and washed twice with phosphate buffered saline(PBS). Then 150 μL of dimethyl sulphoxide (DMSO) was added and vibrated for 10 min. The spectrophotometric absorbance at a wavelength of 570 nm was measured using EL800 enzyme-linked
Table 1 Theoretical coefficient of oxide used in calculation of thermal expansion. Composition
Coefficient (×10− 7)
B2O3 ⁎
SiO2 34 ≤ P ≤ 67
67 ≤ P ≤ 100
ψ≤4
4≤ψ
35 + 0.5(67 − P)
35 − 1.0(P − 67)
12.4(4 − ψ) − 50
− 50
⁎Symbol note: ψ = ∑K RmOn/B2O3.
Al2O3
Na2O
K2O
SnO2
− 40
400
480
− 25
L. Guo et al. / Materials Science and Engineering C 32 (2012) 1531–1535 Table 2 Effect of SiO2 concentration on the thermal expansion coefficient of bonding porcelains. Sample
SiO2 1# 2# 3# 4# 5# 6#
α (×10− 6)
Composition (wt %)
50.0 52.0 54.0 56.0 58.0 60.0
Al2O3 10.0 10.0 10.0 10.0 10.0 10.0
B2O3 10.0 10.0 10.0 10.0 10.0 10.0
Na2O 9.0 9.0 9.0 9.0 9.0 9.0
K2O 9.0 9.0 9.0 9.0 9.0 9.0
SnO2 12.0 10.0 8.0 6.0 4.0 2.0
Calculated 9.37 9.32 9.27 9.22 9.17 9.12
Bonding strength (MPa)
Table 3 Effect of B2O3 concentration on the thermal expansion coefficient of bonding porcelains. Sample
7# 8# 9# 10# 11# 12# 1# 13#
α (×10− 6)
Composition (wt %) SiO2
Al2O3
B2O3
Na2O
K2O
SnO2
Calculated
Measured
60.0 59.0 58.0 56.0 54.0 52.0 50.0 48
10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0
0.0 1.0 2.0 4.0 6.0 8.0 10.0 12.0
9.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0
9.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0
12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0
9.58 9.47 9.44 9.44 9.44 9.43 9.37 9.40
9.70 9.56 9.57 9.58 9.56 9.55 9.46 9.51
immunosorbent assays reader. A mean value was obtained from the measurement of eight samples. The relative cell proliferation percent(Rcp) was calculated according to the following formula: Rcp ¼
Table 5 Bonding strength contrast between self-made and other titanium bonding porcelain systems.
Measured 9.46 9.40 9.35 9.29 9.24 9.21
Dt 100% Dn
ð5Þ
where Dt and Dn, are the absorbances of the test group and the negative control group [12,13]. Cytotoxicity was rated based on relative cell proliferation percent to controls as [10,14]; non-cytotoxic > 90% cell proliferation percent; slightly cytotoxic = 60–90% cell proliferation percent; moderately cytotoxic = 30–59% cell proliferation percent; severely cytotoxic ≤ 30% cell proliferation percent. The test was repeated using the same extracts. 3. Results and discussion Table 2 shows the effect of SiO2 concentration on thermal expansion coefficient (α) of bonding porcelains. Table 2 shows that the calculated thermal expansion coefficient of bonding porcelains linearly decreased with the increase of SiO2 concentration (or with the decrease of SnO2 concentration). This was because SiO2 was network former and SnO2 was network modifier. The network former reduced α but the network modifier increased α. The measured thermal expansion coefficients also decreased with the increase of SiO2 concentration as the calculated ones. The measured thermal expansion coefficients were all greater than the calculated ones. The deviations are below 5%, which can be attributed
1533
Self-made
Super porcelain Ti-22
Ti-bond
35 ± 2.28
36 ± 2.21
34 ± 3.28
to the apparatus deviations. The thermal expansion coefficients of this six bonding porcelains(both calculated and measured) were all slightly lower than that of titanium (9.5 × 10 − 6/°C). At the same time, the deviation of the coefficient of thermal expansion was lower than 1 × 10 − 6/°C (Δα b l × 10 − 6/°C). Table 3 shows the effect of B2O3 concentration on α of bonding porcelains. Table 3 shows that the α of bonding porcelains (both of the calculated and measured) first decreased slightly with the increasing of B2O3 concentration (from 0 wt% to 10 wt%) and then increased to about 9.4 × 10 − 6(from 10 wt% to 12 wt%). The maximal α value was obtained when the B2O3 concentration was 0 wt%. The minimum α value was obtained when the B2O3 concentration was 10 wt%. The one-way ANOVA test indicated that there was no significant difference within thermal expansion coefficients of 8 #, 9 #, 10 #, 11 # and 12 # bonding porcelains (p > 0.05). When the B2O3 concentration was greater than 10 wt%, the basic oxides concentration (including Na2O and K2O) was greater than that of Al2O3 and B2O3. The Al2O3 and B2O3 reacted as network formers. But for 1 # and 13 # bonding porcelains, the concentration of basic oxides was lower than that of the Al2O3 and B2O3 and the Al2O3 and B2O3 reacted as network modifiers. The different role of B2O3 resulted in the varying trend of the thermal expansion coefficients. The measured thermal expansion coefficients also varied as the calculated ones. The measured thermal expansion coefficients were all greater than the calculated ones and the errors are also below 5%. Table 4 shows the effect of SnO2 concentration on mechanical performance of bonding porcelains. The Vickers hardness of bonding porcelains slightly increased with SnO2 concentration. The one-way ANOVA test indicated that there was a significant difference (p b 0.05) between the Vickers hardness for 2 wt% SnO2 (681 ± 43.0 kgf/mm 2) and 12 wt% SnO2 (832± 30.5 kgf/mm2). However, the hardness for 2 wt% vs. 4 wt% (681 ± 43.0 vs. 701 ± 31.9 kgf/mm2), 6 wt% SnO2 vs. 8 wt% (736 ± 35 vs. 775 ± 38 kgf/mm2) and 10 wt % vs. 12 wt% is not significantly different (p > 0.05). The one-way ANOVA test also indicated that there was no significant difference among the flexure strength of this six bonding porcelains (p > 0.05). It indicated that SnO2 concentration did not affect the flexure strength of bonding porcelains in this research. In the standard ISO 9693, a lower limit of 50 MPa for flexure strength is fixed [8]. In this investigation, the measured average flexure strengths were higher than 70 MPa. Table 5 shows the Ti/porcelain bonding strength contrast between self-made and other commercial titanium bonding porcelain systems. The one-way ANOVA test indicated that there was no significant difference (p > 0.05) among the bonding strength of self-made, Noritake Super porcelain Ti-22 and Ti-bond titanium bonding porcelain systems. In the standard ISO 9693, a lower limit of 25 MPa for bonding strength is fixed [8]. In this investigation, all measured bonding strengths were higher than 25 MPa.
Table 4 Effect of SnO2 concentration on the mechanical performance of bonding porcelains. SnO2 concentration(wt%)
2
4
6
8
10
12
Flexure strength (MPa) Vickers hardness (kgf/mm2)
72 ± 5.6 681 ± 43.0
71 ± 4.8 701 ± 31.9
73 ± 4.1 736 ± 35.4
73 ± 5.4 775 ± 38.4
72 ± 4.9 818 ± 35.4
74 ± 5.2 832 ± 30.5
Note:ISO 9693: 1999 standard (flexure strength ≥ 50 MPa)[8].
1534
L. Guo et al. / Materials Science and Engineering C 32 (2012) 1531–1535
Fig. 1. The XRD patterns of the titanium bonding porcelain.
Fig. 1 shows the XRD patterns of the titanium bonding porcelain. The XRD pattern indicated that the main phase of the bonding porcelain(when the SnO2 content was 12 wt%) was SnO2 (tinstone, JCPDS 41–1445). The XRD pattern of the bonding porcelain(when the SnO2 content was 0 wt%) did not contain visible diffraction peaks, which indicated that this bonding porcelain was homogeneous vitreous and without crystalloids. Main ingredients of this bonding glass were silicon dioxide, boron oxide, alkali oxides (including sodium monoxide and potassium monoxide) and SnO2. The SnO2 gradually precipitated from the glass network with increasing of SnO2 content. Fig. 2 shows the SEM patterns of the titanium bonding porcelain (a) SnO2 content = 2 wt% and (b) SnO2 content = 12 wt %. When the SnO2 content was 2 wt%, as shown in Fig. 2(a), the bonding
Fig. 2. SEM patterns of the titanium bonding porcelain (a) SnO2 content = 0% and (b) SnO2 content = 12%.
Fig. 3. SEM micrograph of titanium surface debonded from (a) bonding porcelain (SnO2 = 6 wt%) and (b) bonding porcelain(SnO2 = 12 wt%).
porcelain was homogeneous vitreous, porosity, and did not contain any crystalloids. When the SnO2 content was increased to 12 wt%, as shown in Fig. 2(b), the SnO2 microcrystals gradually precipitated from the glass matrix. These results accorded with the XRD results. The SnO2 microcrystals reduced the thermal expansion coefficient. Fig. 3(a) shows the SEM micrograph of titanium surface debonded from porcelain. The 4 # bonding porcelain depicted adhesive fracture at Ti-porcelain interface. The EDS results showed that the titanium surface was composed of 36.0 wt% O, 54.8 wt% Ti, 8.1 wt% Si and 0.5 wt% Sn. Si and Sn came from the bonding porcelain and the other elements (Na, K, Al and B ) were not found at the interface, which indicated that only Si and Sn elements attended the reaction between titanium and porcelain. This also indicated that mainly adhesive fracture was found at Ti-porcelain interface and failure of the titanium–porcelain predominantly occurred at the titanium– oxide interface. Fig. 3(b) shows the SEM micrograph of titanium surface debonded from porcelain. The 1 # bonding porcelain depicted cohesive fracture at Ti-porcelain interface with presence of open pores. The EDS results showed that the titanium surface was composed of 48.4 wt% O, 27.4 wt% Ti, 1.0 wt% Al, 2.5 wt% Na, 1.4 wt% K, 10.5 wt% Si and 8.8 wt% Sn. Si, Sn, Na, K, Al and B came from the bonding porcelain. It indicated failure of the titanium–porcelain predominantly occurred in the bonding porcelain. The failure modes at the interface changed from adhesive fracture to cohesive fracture with the increase of SnO2 concentration. Table 6 shows the MTT experiment results for the extract solution, comparing with positive and negative controls. This table displays that all the OD values of the test group were greater than those of the negative control. The extract solution had no effect on L929 cell proliferation. The cytotoxic scale of the bonding porcelain was measured as zero, which corresponds to non-cytotoxicity.
L. Guo et al. / Materials Science and Engineering C 32 (2012) 1531–1535
1535
Table 6 Optical density value and cell proliferation percent for various test groups. Group
Negative control Extract solution Positive control
Cultured for 1 day
Cultured for 3 days
Cultured for 5 days
Cultured for 7 days
OD570nm
Rcp(%)
OD570nm
Rcp(%)
OD570nm
Rcp(%)
OD570nm
Rcp(%)
0.64 ± 0.01 0.63 ± 0.01 0.28 ± 0.01
100 98.4 43.8
2.30 ± 0.02 2.25 ± 0.02 0.23 ± 0.01
100 97.8 10.0
3.25 ± 0.03 3.20 ± 0.02 0.24 ± 0.01
100 98.5 7.4
4.00 ± 0.03 3.98 ± 0.03 0.25 ± 0.01
100 99.5 6.3
4. Conclusions The thermal expansion coefficient of bonding porcelains decreased with the increasing concentration of SiO2. As an intermediate, B2O3 acted as both network formers and modifiers, depending on the relationship between the concentration of basic oxides and intermediates. The main phase of the bonding porcelain was SnO2. The SnO2 microcrystals gradually precipitated from the glass matrix when the SnO2 concentration was increased. The Vickers hardness of bonding porcelains increased with the increase of SnO2 concentration. The failure modes at the interface changed from adhesive fracture to cohesive fracture with the increase of SnO2 concentration. The cytotoxicity of titanium porcelain was ranked as 0. Acknowledgements The authors gratefully acknowledge the Fourth Military Medical University for providing support for porcelain fusion and in vitro bioactivity tests. This work was supported by the National Natural Science Foundation of China (No. 81100789), China Postdoctoral Science Foundation (No. 20100481173), Doctoral Fund for New Youth Scholars of Ministry of Education of China (No. 20090095120017) and the Fundamental Research Funds for the Central Universities (No. JX111744 and No. 2012QNA04).
References [1] J. Yamazoe, M. Nakagawa, Y. Matono, A. Takeuchi, K. Ishikawa, Dent. Mater. J. 26 (2007) 260–267. [2] M.G. Troia Jr., G.E. Henriques, M.F. Mesquita, W. Fragoso, Dent. Mater. 24 (2008) 28–33. [3] L.T. Guo, H.T. Wu, X.C. Liu, Y.B. Zhu, J.Q. Gao, T.W. Guo, Mater. Lett. 63 (2009) 2486–2488. [4] S.S. Wang, X. Yang, L.B. Zhang, H.B. Guang, Shen, M.F.M. Zhang, Surf. Coat. Technol. 205 (2010) 1886–1891. [5] H. Zhang, T.W. Guo, Z.X. Song, X.J. Wang, K.W. Xu, Surf. Coat. Technol. 201 (2007) 5637–5640. [6] L.T. Guo, L.Z. Guo, Y.B. Zhu, C. Xu, H. Zhu, T.W. Guo, Mater. Manuf. Processes 25 (2010) 696–699. [7] G. Meriç, J.E. Dahlb, I.E. Ruyter, Dent. Mater. 24 (2008) 1201–1206. [8] ISO 9693. Metal-ceramic dental restorative systems, (1999). [9] Shaanxi University of Science & Technology, Glass Technology, first ed. China Light Industry Press, Beijing, 2006. [10] S. Bandyopadhyay-Ghosh, I.M. Reaney, I.M. Brook, K. Hurrell-Gillingham, A. Johnson, P.V. Hatton, J. Biomed. Mater. Res. A 80A (2007) 175–183. [11] K. Lee, H.C. Choe, B.H. Kim, Y.M. Ko, Surf. Coat. Technol. 205 (2010) S267–S270. [12] M. Schmid-Schwap, A. Franz, F. König, M. Bristela, T. Lucas, E. Piehslinger, et al., Dent. Mater. 25 (2009) 360–368. [13] Xiaobo Chen, Yuncang Li, Peter D. Hodgson, Cuie Wen, Mater. Sci. Eng. C 31 (2011) 1545–1552. [14] A. Saboori, M. Rabiee, F. Moztarzadeh, M. Sheikhi, M. Tahriri, M. Karimi, Mater. Sci. Eng. C 29 (2009) 335–340. [15] G. Ciofani, S. Danti, D.D. Alessandro, S. Moscato, A. Menciassi, Biochem. Biophys. Res. Commun. 394 (2010) 405–411.
Contents lists available at SciVerse ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Preparation and characterization of a titanium bonding porcelain Litong Guo a,⁎, Yao Shi a, Lizhi Guo b, Qian Zhang c, Junlong Tian a, Yabo Zhu a, Tianwen Guo c a b c
School of Materials Science and Engineering, China University of Mining and Technology, Jiangsu Xuzhou 221116, P.R. China Xi'an Shiyou University, Xi'an 710065, P.R. China The Fourth Military Medical University, Xi'an 710032, P.R. China
a r t i c l e
i n f o
Article history: Received 30 March 2011 Received in revised form 2 February 2012 Accepted 20 April 2012 Available online 28 April 2012 Keywords: Dental material Titanium Bonding porcelain Cytotoxicity
a b s t r a c t The titanium bonding porcelain was synthesized through normal melting-derived route using borate–silicate system. The porcelain was characterized by thermal expansion, X-ray diffraction, scanning electron microscope and cytotoxicity tests. The results of X-ray diffraction showed that the main phase of the bonding porcelain was SnO2. The SnO2 microcrystals precipitated from the glass matrix when the SnO2 content was increased. The thermal expansion coefficient of bonding porcelains decreased with the increasing concentration of SiO2. The thermal expansion coefficient of bonding porcelains first decreased slightly with the increasing of B2O3 concentration (from 0 wt% to 10 wt%) and then increased to about 9.4 × 10− 6/°C(from 10 wt% to 12 wt%). As an intermediate, B2O3 can act as both network formers and modifiers, depending on the relationship between the concentration of basic oxides and intermediates. The Vickers hardness of bonding porcelains increased with the increase of SnO2 concentration. When SnO2 concentration was 6 wt%, only Si and Sn elements attended the reaction between titanium and porcelain and mainly adhesive fracture was found at Ti-porcelain interface. When SnO2 concentration was 12 wt%, failure of the titanium–porcelain predominantly occurred in the bonding porcelain and mainly cohesive fracture was found at Ti-porcelain interface. The methyl thiazolyl tetrazolium assay results demonstrated that the cytotoxicity of the titanium porcelain was ranked as 0. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Titanium is an attractive dental restorative material for its excellent characteristics such as excellent biocompatibility, corrosion resistance, light weight, and high strength and low cost [1,2]. However, inferior bonding in titanium–porcelain systems compared to the conventional metal-porcelain systems is still a major problem for its application [3]. The bonding of porcelain to an alloy is attributed to van-derWaal's forces, to a mechanical interlocking between both materials, and to chemical bonds between the porcelain and an oxide layer, which is built during the firing process on the surface of the alloy by oxidation of the base metals [4]. The main factors that affect the titanium-porcelain bond are (1) growth of an oxide layer on titanium at elevated temperatures, (2) adherence of the self-formed oxide to the Ti substrate, and (3) bonding of the self-formed oxide with the porcelain [4]. It has been proposed that the poor bonding strength between porcelain and titanium was partly because of continual oxidation of titanium during the porcelain fusing and formation of a nonadherent oxide layer [5]. ⁎ Corresponding author at: School of Materials Science & Engineering, China University of Mining and Technology, Xuzhou 221116, P.R. China. Tel.: +86 516 83591979; fax: +86 516 83591870. E-mail address: [email protected] (L. Guo). 0928-4931/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.msec.2012.04.038
Ti has a great affinity for oxygen, and relatively thick and nonadherent layers of titanium oxide tend to form at above 800 °C; therefore, porcelain should be fired below this temperature [4,5]. Special low fusing dental porcelains should be developed for titanium– porcelain bonding. For high temperature fusion, the coefficient of thermal expansion of the porcelains is important during the cooling process. The coefficient of thermal expansion of the porcelains developed should match with that of titanium (Δα b l × 10 − 6/°C). Therefore, in this work, a new Ti bonding porcelain was developed and the effect of composition on thermal expansion coefficient was also investigated. Because dental porcelains may be in contact with oral soft tissues for periods of time, assessment of dental material biocompatibility is gaining increasing importance for both patients and dentists. Clinical trials of dental materials are now clearly regulated and may only be initiated after successful completion of nonclinical evaluations and biocompatibility studies according to European regulations and ISO guidelines [6–8]. The aim of present investigation is to develop a new Ti bonding porcelain and characterize its properties such as phase structure, thermal expansion coefficient, mechanical properties and cytotoxicity [8–11]. At the same time, understanding the reactions between porcelain and titanium is important to elucidate the bonding mechanism. Therefore, the failure modes at the interface were investigated to gain insight into the possible mechanisms of Ti-porcelain bonding.
1532
L. Guo et al. / Materials Science and Engineering C 32 (2012) 1531–1535
modulus of titanium (ETi = 105.4 GPa). The coefficient K [8] can be expressed as:
2. Experimental 2.1. Preparation
2
K ¼ 54:78 dTi −73:15 dTi þ 27:65 ¼ 4:78 The titanium bonding porcelain was prepared by melting the corresponding chemicals (such as SiO2, H3BO3, Al2O3, Na2CO3 and K2CO3) for 2 h at 1450 °C. The cast glass was crushed and milled in an agate planetary mill, using a speed of rotation of 1500 rpm for 6 h. The titanium–porcelain test specimens were prepared for a three point bending test as specified in ISO 9693 [8]. Then GG Titanium Porcelain (self-made, Xi'an Jiaotong University, China) was fused in a Multimat 99 furnace (Dentsply, American) according to the manufacturers' instructions. A thin layer of bonding porcelain, opaque porcelain and dentin porcelain was fired at the central areaof the specimen, respectively. The porcelain built up was about 8 × 3 × 1 mm 3. The thickness of the bonding porcelain and opaque porcelain together was 0.4 ± 0.02 mm. The total thickness of the fired porcelain was 1 mm. After the dentin porcelain firing, specimens were ultrasonically cleaned in deionized water for 5 min and dried in air. 2.2. Characterization 2.2.1. Thermal expansion analysis The plate-shaped Ti porcelain specimens of the dimensions 25 mm × 4 mm × 4 mm were prepared. The thermal expansion properties were recorded using a computerized thermal dilatometer (NETZSCH-DIL 402 C, Germany) with a heating rate of 5 °C/min. Data were obtained from room temperature to 600 °C. For each specimen, three measurements were made and reproducible results were obtained (±5%). The theoretical thermal expansion coefficient of the bonding porcelain was calculated according to the following formula: α ¼ α 1 P1 þ α 2 P2 þ α 3 P3 þ ……
ð1Þ
where P1, P2 and P3 are the oxide concentration of porcelain. The α1, α2 and α3 are the calculating coefficient of the oxide, which was shown in Table 1 [9]. 2.2.2. Three-point bending test The three-point bending test was performed by a universal mechanical testing machine (DSS–25 T, Shimadzu, Japan) with a span distance of 20 mm. The specimens were placed in the testing machine with the porcelain positioned symmetrically on the side opposite to the applied load. The force was applied at a crosshead speed of 1.5 mm/min with the porcelain side opposite the center support until fracture. The three-point bonding strength was calculated according to the formula in ISO 9693 [8]. The bonding strength, τb, can be expressed as: τb ¼ K Ffail
ð2Þ
where Ffail is the applied load at failure, and K is a function of the thickness of the titanium (dTi = 0.5 mm), and the value of Young's
ð3Þ
For each group, a set of six samples was chosen by random sampling. The results were analyzed by one‐way ANOVA and Student– Newman–Kuels test at α= 0.05 using statistical software (SPSS 10.0 for Windows). 2.2.3. Vickers hardness test Vickers hardness testing was performed on an HV-5 hardness tester with an applied load of 5 kg for 15 s. For every sample, both diagonal lengths of each indentation were optically measured in the hardness apparatus. Vickers hardness was calculated by the following formula: 2
HV ¼ 1:854 P=d
ð4Þ
where HV is the Vickers hardness, P is the applied load, and d is the mean length of the two diagonal lines of the indentation. For each group, a set of six indentations was used. The results were analyzed by one‐way ANOVA and Student–Newman–Kuels test at α = 0.05 using statistical software (SPSS 10.0 for Windows). 2.2.4. XRD and SEM analysis XRD was performed at room temperature using Cu Kα radiation (λ = 1.5418 Å) on an X-ray diffractometer (D/max-r, Rigaku Corp., Japan). The acceleration voltage was 60 kV with an 80 mA current flux. Data were collected for 2θ in the range of 10–90° employing a step size of 0.02°, with a counting rate of 10.0°/min. The titanium surface and Ti–porcelain bonding interface were examined on a scanning electron microscope (JSM–6460, JEOL, Japan) coupled EDS apparatus (INCAX–sight, Oxford, England). 2.3.5. Cytotoxicity test The cytotoxicity properties of the samples were evaluated by methyl thiazolyl tetrazolium (MTT) assay in association with L929 mouse fibroblasts [11–15]. The extract solutions were made by immersing the samples (Φ 5 mm × 1 mm) in culture medium of Roswell Park Memorial Institute (RPMI) media 1640 at approximately 0.55 mL/cm 2, 37 °C for 72 h. Then the extract solution replaced the medium used in the test group. The L929 mouse fibroblast cells were collected and resuspended at 5 × 10 4 cell/mL. Cells were incubated into 96-well culture clusters, 100 μL in each well, and kept at 37 °C in a fully humidified atmosphere at 5% CO2 in air for 24 h. The medium in negative and positive controls was replaced by fresh culture medium and diluted hydroxybenzene solution (10 wt%), respectively. Cell proliferation was determined after being treated for 1, 3, 5 and 7 days. The medium was removed and 20 μL of MTT solution (2 mg/mL) was added to each well. After incubation at 37 °C for 3 h in a fully humidified atmosphere at 5% CO2 in air, the untransformed MTT was removed and washed twice with phosphate buffered saline(PBS). Then 150 μL of dimethyl sulphoxide (DMSO) was added and vibrated for 10 min. The spectrophotometric absorbance at a wavelength of 570 nm was measured using EL800 enzyme-linked
Table 1 Theoretical coefficient of oxide used in calculation of thermal expansion. Composition
Coefficient (×10− 7)
B2O3 ⁎
SiO2 34 ≤ P ≤ 67
67 ≤ P ≤ 100
ψ≤4
4≤ψ
35 + 0.5(67 − P)
35 − 1.0(P − 67)
12.4(4 − ψ) − 50
− 50
⁎Symbol note: ψ = ∑K RmOn/B2O3.
Al2O3
Na2O
K2O
SnO2
− 40
400
480
− 25
L. Guo et al. / Materials Science and Engineering C 32 (2012) 1531–1535 Table 2 Effect of SiO2 concentration on the thermal expansion coefficient of bonding porcelains. Sample
SiO2 1# 2# 3# 4# 5# 6#
α (×10− 6)
Composition (wt %)
50.0 52.0 54.0 56.0 58.0 60.0
Al2O3 10.0 10.0 10.0 10.0 10.0 10.0
B2O3 10.0 10.0 10.0 10.0 10.0 10.0
Na2O 9.0 9.0 9.0 9.0 9.0 9.0
K2O 9.0 9.0 9.0 9.0 9.0 9.0
SnO2 12.0 10.0 8.0 6.0 4.0 2.0
Calculated 9.37 9.32 9.27 9.22 9.17 9.12
Bonding strength (MPa)
Table 3 Effect of B2O3 concentration on the thermal expansion coefficient of bonding porcelains. Sample
7# 8# 9# 10# 11# 12# 1# 13#
α (×10− 6)
Composition (wt %) SiO2
Al2O3
B2O3
Na2O
K2O
SnO2
Calculated
Measured
60.0 59.0 58.0 56.0 54.0 52.0 50.0 48
10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0
0.0 1.0 2.0 4.0 6.0 8.0 10.0 12.0
9.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0
9.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0
12.0 12.0 12.0 12.0 12.0 12.0 12.0 12.0
9.58 9.47 9.44 9.44 9.44 9.43 9.37 9.40
9.70 9.56 9.57 9.58 9.56 9.55 9.46 9.51
immunosorbent assays reader. A mean value was obtained from the measurement of eight samples. The relative cell proliferation percent(Rcp) was calculated according to the following formula: Rcp ¼
Table 5 Bonding strength contrast between self-made and other titanium bonding porcelain systems.
Measured 9.46 9.40 9.35 9.29 9.24 9.21
Dt 100% Dn
ð5Þ
where Dt and Dn, are the absorbances of the test group and the negative control group [12,13]. Cytotoxicity was rated based on relative cell proliferation percent to controls as [10,14]; non-cytotoxic > 90% cell proliferation percent; slightly cytotoxic = 60–90% cell proliferation percent; moderately cytotoxic = 30–59% cell proliferation percent; severely cytotoxic ≤ 30% cell proliferation percent. The test was repeated using the same extracts. 3. Results and discussion Table 2 shows the effect of SiO2 concentration on thermal expansion coefficient (α) of bonding porcelains. Table 2 shows that the calculated thermal expansion coefficient of bonding porcelains linearly decreased with the increase of SiO2 concentration (or with the decrease of SnO2 concentration). This was because SiO2 was network former and SnO2 was network modifier. The network former reduced α but the network modifier increased α. The measured thermal expansion coefficients also decreased with the increase of SiO2 concentration as the calculated ones. The measured thermal expansion coefficients were all greater than the calculated ones. The deviations are below 5%, which can be attributed
1533
Self-made
Super porcelain Ti-22
Ti-bond
35 ± 2.28
36 ± 2.21
34 ± 3.28
to the apparatus deviations. The thermal expansion coefficients of this six bonding porcelains(both calculated and measured) were all slightly lower than that of titanium (9.5 × 10 − 6/°C). At the same time, the deviation of the coefficient of thermal expansion was lower than 1 × 10 − 6/°C (Δα b l × 10 − 6/°C). Table 3 shows the effect of B2O3 concentration on α of bonding porcelains. Table 3 shows that the α of bonding porcelains (both of the calculated and measured) first decreased slightly with the increasing of B2O3 concentration (from 0 wt% to 10 wt%) and then increased to about 9.4 × 10 − 6(from 10 wt% to 12 wt%). The maximal α value was obtained when the B2O3 concentration was 0 wt%. The minimum α value was obtained when the B2O3 concentration was 10 wt%. The one-way ANOVA test indicated that there was no significant difference within thermal expansion coefficients of 8 #, 9 #, 10 #, 11 # and 12 # bonding porcelains (p > 0.05). When the B2O3 concentration was greater than 10 wt%, the basic oxides concentration (including Na2O and K2O) was greater than that of Al2O3 and B2O3. The Al2O3 and B2O3 reacted as network formers. But for 1 # and 13 # bonding porcelains, the concentration of basic oxides was lower than that of the Al2O3 and B2O3 and the Al2O3 and B2O3 reacted as network modifiers. The different role of B2O3 resulted in the varying trend of the thermal expansion coefficients. The measured thermal expansion coefficients also varied as the calculated ones. The measured thermal expansion coefficients were all greater than the calculated ones and the errors are also below 5%. Table 4 shows the effect of SnO2 concentration on mechanical performance of bonding porcelains. The Vickers hardness of bonding porcelains slightly increased with SnO2 concentration. The one-way ANOVA test indicated that there was a significant difference (p b 0.05) between the Vickers hardness for 2 wt% SnO2 (681 ± 43.0 kgf/mm 2) and 12 wt% SnO2 (832± 30.5 kgf/mm2). However, the hardness for 2 wt% vs. 4 wt% (681 ± 43.0 vs. 701 ± 31.9 kgf/mm2), 6 wt% SnO2 vs. 8 wt% (736 ± 35 vs. 775 ± 38 kgf/mm2) and 10 wt % vs. 12 wt% is not significantly different (p > 0.05). The one-way ANOVA test also indicated that there was no significant difference among the flexure strength of this six bonding porcelains (p > 0.05). It indicated that SnO2 concentration did not affect the flexure strength of bonding porcelains in this research. In the standard ISO 9693, a lower limit of 50 MPa for flexure strength is fixed [8]. In this investigation, the measured average flexure strengths were higher than 70 MPa. Table 5 shows the Ti/porcelain bonding strength contrast between self-made and other commercial titanium bonding porcelain systems. The one-way ANOVA test indicated that there was no significant difference (p > 0.05) among the bonding strength of self-made, Noritake Super porcelain Ti-22 and Ti-bond titanium bonding porcelain systems. In the standard ISO 9693, a lower limit of 25 MPa for bonding strength is fixed [8]. In this investigation, all measured bonding strengths were higher than 25 MPa.
Table 4 Effect of SnO2 concentration on the mechanical performance of bonding porcelains. SnO2 concentration(wt%)
2
4
6
8
10
12
Flexure strength (MPa) Vickers hardness (kgf/mm2)
72 ± 5.6 681 ± 43.0
71 ± 4.8 701 ± 31.9
73 ± 4.1 736 ± 35.4
73 ± 5.4 775 ± 38.4
72 ± 4.9 818 ± 35.4
74 ± 5.2 832 ± 30.5
Note:ISO 9693: 1999 standard (flexure strength ≥ 50 MPa)[8].
1534
L. Guo et al. / Materials Science and Engineering C 32 (2012) 1531–1535
Fig. 1. The XRD patterns of the titanium bonding porcelain.
Fig. 1 shows the XRD patterns of the titanium bonding porcelain. The XRD pattern indicated that the main phase of the bonding porcelain(when the SnO2 content was 12 wt%) was SnO2 (tinstone, JCPDS 41–1445). The XRD pattern of the bonding porcelain(when the SnO2 content was 0 wt%) did not contain visible diffraction peaks, which indicated that this bonding porcelain was homogeneous vitreous and without crystalloids. Main ingredients of this bonding glass were silicon dioxide, boron oxide, alkali oxides (including sodium monoxide and potassium monoxide) and SnO2. The SnO2 gradually precipitated from the glass network with increasing of SnO2 content. Fig. 2 shows the SEM patterns of the titanium bonding porcelain (a) SnO2 content = 2 wt% and (b) SnO2 content = 12 wt %. When the SnO2 content was 2 wt%, as shown in Fig. 2(a), the bonding
Fig. 2. SEM patterns of the titanium bonding porcelain (a) SnO2 content = 0% and (b) SnO2 content = 12%.
Fig. 3. SEM micrograph of titanium surface debonded from (a) bonding porcelain (SnO2 = 6 wt%) and (b) bonding porcelain(SnO2 = 12 wt%).
porcelain was homogeneous vitreous, porosity, and did not contain any crystalloids. When the SnO2 content was increased to 12 wt%, as shown in Fig. 2(b), the SnO2 microcrystals gradually precipitated from the glass matrix. These results accorded with the XRD results. The SnO2 microcrystals reduced the thermal expansion coefficient. Fig. 3(a) shows the SEM micrograph of titanium surface debonded from porcelain. The 4 # bonding porcelain depicted adhesive fracture at Ti-porcelain interface. The EDS results showed that the titanium surface was composed of 36.0 wt% O, 54.8 wt% Ti, 8.1 wt% Si and 0.5 wt% Sn. Si and Sn came from the bonding porcelain and the other elements (Na, K, Al and B ) were not found at the interface, which indicated that only Si and Sn elements attended the reaction between titanium and porcelain. This also indicated that mainly adhesive fracture was found at Ti-porcelain interface and failure of the titanium–porcelain predominantly occurred at the titanium– oxide interface. Fig. 3(b) shows the SEM micrograph of titanium surface debonded from porcelain. The 1 # bonding porcelain depicted cohesive fracture at Ti-porcelain interface with presence of open pores. The EDS results showed that the titanium surface was composed of 48.4 wt% O, 27.4 wt% Ti, 1.0 wt% Al, 2.5 wt% Na, 1.4 wt% K, 10.5 wt% Si and 8.8 wt% Sn. Si, Sn, Na, K, Al and B came from the bonding porcelain. It indicated failure of the titanium–porcelain predominantly occurred in the bonding porcelain. The failure modes at the interface changed from adhesive fracture to cohesive fracture with the increase of SnO2 concentration. Table 6 shows the MTT experiment results for the extract solution, comparing with positive and negative controls. This table displays that all the OD values of the test group were greater than those of the negative control. The extract solution had no effect on L929 cell proliferation. The cytotoxic scale of the bonding porcelain was measured as zero, which corresponds to non-cytotoxicity.
L. Guo et al. / Materials Science and Engineering C 32 (2012) 1531–1535
1535
Table 6 Optical density value and cell proliferation percent for various test groups. Group
Negative control Extract solution Positive control
Cultured for 1 day
Cultured for 3 days
Cultured for 5 days
Cultured for 7 days
OD570nm
Rcp(%)
OD570nm
Rcp(%)
OD570nm
Rcp(%)
OD570nm
Rcp(%)
0.64 ± 0.01 0.63 ± 0.01 0.28 ± 0.01
100 98.4 43.8
2.30 ± 0.02 2.25 ± 0.02 0.23 ± 0.01
100 97.8 10.0
3.25 ± 0.03 3.20 ± 0.02 0.24 ± 0.01
100 98.5 7.4
4.00 ± 0.03 3.98 ± 0.03 0.25 ± 0.01
100 99.5 6.3
4. Conclusions The thermal expansion coefficient of bonding porcelains decreased with the increasing concentration of SiO2. As an intermediate, B2O3 acted as both network formers and modifiers, depending on the relationship between the concentration of basic oxides and intermediates. The main phase of the bonding porcelain was SnO2. The SnO2 microcrystals gradually precipitated from the glass matrix when the SnO2 concentration was increased. The Vickers hardness of bonding porcelains increased with the increase of SnO2 concentration. The failure modes at the interface changed from adhesive fracture to cohesive fracture with the increase of SnO2 concentration. The cytotoxicity of titanium porcelain was ranked as 0. Acknowledgements The authors gratefully acknowledge the Fourth Military Medical University for providing support for porcelain fusion and in vitro bioactivity tests. This work was supported by the National Natural Science Foundation of China (No. 81100789), China Postdoctoral Science Foundation (No. 20100481173), Doctoral Fund for New Youth Scholars of Ministry of Education of China (No. 20090095120017) and the Fundamental Research Funds for the Central Universities (No. JX111744 and No. 2012QNA04).
References [1] J. Yamazoe, M. Nakagawa, Y. Matono, A. Takeuchi, K. Ishikawa, Dent. Mater. J. 26 (2007) 260–267. [2] M.G. Troia Jr., G.E. Henriques, M.F. Mesquita, W. Fragoso, Dent. Mater. 24 (2008) 28–33. [3] L.T. Guo, H.T. Wu, X.C. Liu, Y.B. Zhu, J.Q. Gao, T.W. Guo, Mater. Lett. 63 (2009) 2486–2488. [4] S.S. Wang, X. Yang, L.B. Zhang, H.B. Guang, Shen, M.F.M. Zhang, Surf. Coat. Technol. 205 (2010) 1886–1891. [5] H. Zhang, T.W. Guo, Z.X. Song, X.J. Wang, K.W. Xu, Surf. Coat. Technol. 201 (2007) 5637–5640. [6] L.T. Guo, L.Z. Guo, Y.B. Zhu, C. Xu, H. Zhu, T.W. Guo, Mater. Manuf. Processes 25 (2010) 696–699. [7] G. Meriç, J.E. Dahlb, I.E. Ruyter, Dent. Mater. 24 (2008) 1201–1206. [8] ISO 9693. Metal-ceramic dental restorative systems, (1999). [9] Shaanxi University of Science & Technology, Glass Technology, first ed. China Light Industry Press, Beijing, 2006. [10] S. Bandyopadhyay-Ghosh, I.M. Reaney, I.M. Brook, K. Hurrell-Gillingham, A. Johnson, P.V. Hatton, J. Biomed. Mater. Res. A 80A (2007) 175–183. [11] K. Lee, H.C. Choe, B.H. Kim, Y.M. Ko, Surf. Coat. Technol. 205 (2010) S267–S270. [12] M. Schmid-Schwap, A. Franz, F. König, M. Bristela, T. Lucas, E. Piehslinger, et al., Dent. Mater. 25 (2009) 360–368. [13] Xiaobo Chen, Yuncang Li, Peter D. Hodgson, Cuie Wen, Mater. Sci. Eng. C 31 (2011) 1545–1552. [14] A. Saboori, M. Rabiee, F. Moztarzadeh, M. Sheikhi, M. Tahriri, M. Karimi, Mater. Sci. Eng. C 29 (2009) 335–340. [15] G. Ciofani, S. Danti, D.D. Alessandro, S. Moscato, A. Menciassi, Biochem. Biophys. Res. Commun. 394 (2010) 405–411.
