- Email: [email protected]
Heat treatment mechanism and biodegradable characteristics of ZAX1330 Mg alloy
Materials Science and Engineering C 51 (2015) 300–308
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Heat treatment mechanism and biodegradable characteristics of ZAX1330 Mg alloy Da-Jun Lin a, Fei-Yi Hung a,⁎, Truan-Sheng Lui a, Ming-Long Yeh b a b
Department of Materials Science and Engineering, National Cheng Kung University, Tainan 701, Taiwan Department of Biomedical Engineering, National Cheng Kung University, Tainan 701, Taiwan
a r t i c l e
i n f o
Article history: Received 26 July 2014 Received in revised form 25 February 2015 Accepted 7 March 2015 Available online 10 March 2015 Keywords: Magnesium alloy Biodegradable Heat treatment Mechanical properties Corrosion Cytocompatibility
a b s t r a c t Heat treatments are key processes in the development of biodegradable magnesium implants. The aim of this study is to investigate the factors of microstructures and metallurgical segregation on the functionality of biodegradable magnesium alloy. The solid solution heat treatment and strain induced melting activation heat treatment were employed to alter the microstructures of ZAX1330 alloy in this study. Heat treatments caused a significant change on grain size and distribution of secondary phases. The fine-grained microstructure enhanced the mechanical strength, corrosion resistance and achieved the lowest degradation rate in simulated body fluid solution. In coarse-grained microstructure systems, grain growth followed liquid phase formation. The corrosion rate increased due to a larger cathodic region. The status of micro-alloyed calcium (in solid solution or segregated) influenced the microstructural evolution mechanisms, mechanical strength, and degradation properties. A cytotoxicity test and a live/dead assay showed that ZAX1330 had good cytocompatibility, which varied with heat treatment, and no cell toxicity. The results suggest that heat treatment should be controlled precisely in order to improve the cytocompatibility of magnesium alloys for application in orthopedic implants. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Magnesium alloys have recently received substantial scientific and clinical interests for use as new generation implant materials in biomedical application [1,2]. Metallic implants made of stainless steel or titanium alloy exhibit stress shielding effect [3]. The density and elastic modulus of magnesium alloys (density: 1.7 g/cm3; Young's modulus: 45 MPa) are close to those of human cortical bone [2], and thus the stress shielding effect of magnesium alloys can be effectively minimized or avoided. In ideal conditions, magnesium alloys provide appropriate mechanical support, and maintain the mechanical strength during the degradation period. Thus, the risk of secondary surgery can be prevented on condition that bone implant is made by magnesium alloys. Magnesium is one of the major essential elements in the human body. A lack of magnesium leads to significant cardiovascular disease and osteoporosis [4]. Magnesium exhibits unique biodegradability behavior, as the degradation product (Mg2+ ions) can be regulated by the metabolism of the kidney, with excess magnesium ions eliminated via renal excretion [5]. Magnesium alloys have many advantages as biomedical materials, and thus have high potential for orthopedic applications. (See Table 1.) One of the useful tools to modify the applicability of magnesium alloys is controlling their alloying composition. Researchers have attempted to solve the problems of low strength and poor corrosion ⁎ Corresponding author. E-mail address: [email protected] (F.-Y. Hung).
http://dx.doi.org/10.1016/j.msec.2015.03.004 0928-4931/© 2015 Elsevier B.V. All rights reserved.
resistance of Mg alloys. Aluminum has been widely reported to improve the mechanical properties and corrosion resistance of magnesium alloys. Many studies have used aluminum and/or rare earth elements (REEs) as alloying elements [6–8]. Magnesium alloys that contain aluminum (e.g., AZ91D) can form a dense passivation film that protects the alloy. Unfortunately, it is well known that a high concentration of aluminum is harmful to neuron cells and osteoblasts [9,10], and is associated with Alzheimer's disease and osteomalacia. The use of RE could lead to hepatotoxicity or have adverse effect on genes. [11,12]. Consequently, aluminum and RE are widely acknowledged as being potentially harmful alloying elements for magnesium materials. To minimize toxicity, magnesium alloys with a low content of harmful alloying elements need to be developed for practical applications. Zinc (Zn) is the second most abundant essential trace element in the human body, with 85% of Zn in the body being in the muscle and bone [13]. Furthermore, Zn alloying can improve the anti-corrosion properties and mechanical properties of magnesium alloys [14]. Moreover, for low aluminum content magnesium alloys, the ratio of Zn to Al needs to increase in improving casting properties and avoiding hot crack effect [15]. Magnesium alloys with low aluminum content provided lower neurotoxicity, they are relatively highly safe material. Notably, scientists tend to add calcium into magnesium alloys to enhance biocompatibility in recent years [16]. Ca is also one of the alloying elements that can modify the microstructure of magnesium alloys. Due to the high GRF (grain restriction factor) value of calcium in magnesium alloy system, the grain refinement effect in the as-cast condition can
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308 Table 1 Results of the electrochemical polarization tests in r-SBF solution.
AE T4 S355 S370
Ecorr (V)
Icorr (μA/cm2)
Rp (Ω/cm2)
CR (mm/year)
−1.51 −1.49 −1.48 −1.46
28.92 18.47 167.82 99.98
459.37 803.92 173.34 440.74
0.78 0.50 4.54 2.46
301
resolution transmission electron microscopy (HR-TEM) analysis was used to investigate the crystalline structures of the major secondary phase by electron diffraction pattern. Tensile test used a mechanical testing system with a tensile speed of 1 mm/min. A typical specimen expressed with a gauge length of 20 mm, a gauge width of 7 mm and a thickness of 3 mm was selected for tensile test. At least three samples are tested for each test. The result is the average of three samples. 2.3. Electrochemical test
be significantly improved. Studies have reported that the grain size decreased with increasing calcium content for content levels of 0 to 0.5 wt.% [17,18]. However, excessive calcium in Mg–Zn–Ca alloy led to the formation of Mg2Ca and Ca2Mg6Zn3, which are brittle phases that decrease ductility and strength. In general, the suggested amount of calcium is below 0.5 wt.% to prevent excess brittle phases in the matrix. Recent studies have reported the influence of microstructure on the corrosion behavior and mechanical properties of Mg alloys. Zhang et al. found that adding Ca and Zn into Mg–Si alloy led to a microstructural transformation of the as-cast matrix [19]. Zheng et al. reported that the tensile strength and corrosion resistance of ZK60 were significantly different for as-cast and as-extrude microstructures, and that the biocompatibility could be enhanced through grain refinement and post-processing [20]. However, little attention has been given to the influences of metallurgical segregation and the distribution of secondary phases. This study investigates the primary research results of Mg–Zn–Al–Ca alloy (ZAX1330) with high Zn (12.83 wt.%), low Al (3.35 wt.%), and trace Ca (0.20 wt.%) on the evolution of microstructure, using SIMA (strain induced melting activation) heat treatment [21] creates an extremely segregated microstructure (Ca atoms gather) compared with the solid-solution (T4) microstructure and the transitional microstructure to investigate the biodegradable characteristics. In addition, the mechanical properties, corrosion behavior and cytocompatibility of Mg–Zn–Al–Ca alloy (ZAX1330) have also been studied in four kinds of microstructure by using different heattreatments. The related outcomes can be expected to provide the preliminary suggestions for medical applications of Bio-Mg alloys.
A conventional three-electrode electrochemical cell was used for the polarization test with a platinum counter electrode and a saturated calomel electrode (SCE) as the reference electrode. The tests were performed in revised simulated body fluid (r-SBF) solution (per liter, dissolved 5.403 g of NaCl, 0.736 g of NaHCO3, 2.036 g of Na2CO3, 0.225 g of KCl, 0.182 g of K2HPO4, 0.310 g of MgCl2 · 6H2O, 11.928 g of 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 0.293 g of CaCl2, 0.072 g of Na2SO4 in deionized water. The r-SBF solution was then buffered to pH 7.4 at 37 °C by adding HEPES and NaOH) at 37 °C by using a water bath. The polarization test used PARSTAT 2273 electrochemistry workstation with a scanning rate of 1 mV s−1 from −1.65 V to −1.3 V. The exposed area of the working electrode to the electrolyte was controlled by a Teflon holder within 1 cm2. 2.4. Immersion test The immersion tests were employed to realize the corrosion mechanism in r-SBF solution according to ASTM G31-72 [23]. The r-SBF volume to surface area ratio was fixed at 20 mL/cm2. The experimental samples were removed after 24 h of immersion, rinsed with distilled water and cleaned with boiled chromic acid/silver nitrate solution (200 g/L CrO3 + 10 g/L AgNO3) to remove the corrosion products. The corrosion surfaces were examined by SEM. The degradation rates were obtained according to the following equation: C = (m1 − m2) / St (m1: weight before corrosion, m2: weight after corrosion, S: surface area, t: immersion time)
2. Materials and methods
2.5. Cell culture and in-vitro test
2.1. Material preparation
Human osteoblast-like cells (MG63) were used to examine the cytocompatibility of ZAX1330. Before cytotoxicity tests, the MG63 cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) with high glucose, supplemented with 10% fetal bovine serum (FBS), and 1% antibiotics. Cells were sub-cultured at least twice a week and were kept in a humidified atmosphere of 95% air and 5% CO2. The cytotoxicity tests were carried out by indirect method. Extracts were prepared according to ISO 10993–5 [24]. After 24 h of incubation in the cell incubator at 37 °C, each group of the extraction medium was passed through a 0.22 μm filter paper to remove corrosion particles. The negative control group used culture medium and the positive control group used culture medium with 5% dimethyl sulfoxide (DMSO). Cells were cultured in 96-well plates at 5 × 103 cells/100 μL medium in each well and incubated for 24 h to allow for initial cell adhesion. The medium was replaced with 100 μL of experimental extracts. After the cells were incubated for 1, 3, and 6 days, respectively, 10 μL of CellTiter 96 solution was added into each well. All solutions were incubated for 4 h in a cell incubator, and then 100 μL of CellTiter-96-treated medium solution was sucked out and placed into new 96-well plates. The absorbance of the CellTiter-96-treated solution was measured using an enzyme-linked immunosorbent assay (ELISA) reader at a wavelength of 490 nm. After the experiment, the cell relative growth rate (RGR) was calculated according to the following formula:
The extruded ZAX1330 alloy (12.83 wt.% Zn, 3.35 wt.% Al, 0.20 wt.% Ca) was used in this study. ZAX1330 was processed into a circular plate (12.7 mm in diameter) using a computer numerical control (CNC) wire cutting machine. The as-extruded alloy is denoted as the AE specimen. Solution treatment (T4) of the alloy was carried out at 345 °C for 10 h in an electrical furnace, followed by water quenching at room temperature. The SIMA process was used to create a Ca-rich zone microstructure. The process was performed at 370 °C for 1 h to produce coarse grains (spheroidal α-Mg grains) with a high fraction of the liquid phase (Ca-rich zone) in the matrix (S370 specimen). A medium temperature of 355 °C was chosen to investigate the transition microstructure from equiaxed grains to spheroidal grains. This temperature was held for 1 h (S355 specimen). 2.2. Microstructure and mechanical test All ZAX1330 samples (AE, T4, S355, S370) for microstructural observation were polished up to 0.05 μm and etched. Optical microscopy (OM) and scanning electron microscopy (SEM) with an energy dispersive spectrometer (EDS) were used for microstructural observation. The grain size was determined by the linear intercept method according to ASTM E112-96 standard [22]. The constituent phases were characterized by X-ray diffraction meter (XRD) with Cu-Kα radiation. Diffraction patterns were obtained between 2θ values of 20–80°. The high
h i RGR ¼ ðODtest −ODblank Þ= ODnegative −ODblank 100%:
302
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
Cell cytotoxicity measurements were given as the mean RGR value ± standard deviation, and the analysis of one-way variance (ANOVA) was conducted to evaluate the statistical significance of differences. Difference at p ≦ 0.05 was considered to be statistically significant. For live/dead staining, the cell culture media were removed after 3 days of incubation and the wells were rinsed with phosphate buffer solution (PBS), and 200 μL of dye was settled in 96 wells to stain the cells at 37 °C. After staining for 30 min, the dye was cleaned by PBS to remove unreacted dye and non-adhered cells. The labeled cells were observed using a fluorescence microscopy at wavelengths of 530 nm (live cells) and 645 nm (dead cells). 3. Results and discussion 3.1. Microstructural characteristics The microstructures of all specimens are shown in Fig. 1. In asextruded ZAX1330 (Fig. 1a, AE), the average grain size of the primary α-Mg was about 2 μm and the clusters of the secondary phases were aligned in the direction of extrusion. After solid-solution treatment (Fig. 1b, T4), the grain size of the matrix increased to 5 μm. The secondary phases of the T4 specimen were distributed uniformly. EDS analysis shows a zinc content of 3.84 at.% and an aluminum content of 1.83 at.% in the secondary phase. Fig. 1c shows the microstructure of S355 for heat treatment at 1 h at 355 °C, and the grain size significantly increased and partially melted secondary phases re-aggregated at the grain boundaries to form initial liquid phases. When the heat treatment temperature was increased to 370 °C (Fig. 1d, S370), the secondary phases disappeared during heat treatment, and metallurgical segregation was
observed along the grain boundary, which led to the formation of a liquid zone (LZ). All equiaxed grains evolved into coarse α-Mg spheroidal grains. The EDS analysis of Fig. 2 shows that the LZ consisted of non-equilibrium phases and intermetallic compounds (IMC). There were several Ca-rich needle-like compounds which were confirmed as Ca2Mg6Zn3 in the LZ of the S370 matrix. Most of the secondary phases of S370 aggregated at the grain boundaries, there were still few Al8Mn5 and Mg5Zn2Al2 in grains. Fig. 3 shows the X-ray diffraction patterns of all samples. The main secondary phase is Mg32(Al, Zn)49. According to TEM of T4 (Fig. 4), there was 0.8 at.% calcium in α-Mg, Mg32(Al, Zn)49 was found to be in a ternary icosahedral quasi-crystalline phase (Q phase) instead of the τ phase, which has been reported to be a body-centered cubic structure (space group Im3, a = 1.416 nm) [25]. Notably, the Q-phase in S355 and S370 specimens was significantly lower than that in the T4 specimen, but the intensity of α-Mg showed no significant difference. This result can be explained by the Q-phase melting at high temperatures. From the results of XRD, Ca2Mg6Zn3 was in AE and S370 specimen, whereas X-ray didn't detect other intermetallic phases consisted with Ca (e.g., Al2Ca). In the ternary phase diagram of Mg–Zn–Ca, the eutectic phases are α-Mg + Ca2Mg6Zn3 + Mg2Ca when the Zn/Ca atomic ratio is less than 1.2 [26]. On the other hand, other kinds of eutectic phases are αMg + Ca2Mg6Zn3 which form while the Zn/Ca atomic ratio is more than 1.2. The results presented here confirm that α-Mg and Ca2Mg6Zn3 coexisted in the matrix (the Zn/Ca ratio for ZAX1330 is higher than 1.2). 3.2. Effect of metallurgical segregation on mechanical properties Fig. 5 shows the stress–strain curves of AE, T4, S355 and S370 specimens. The elastic modulus decreased with the increase of heat
Fig. 1. Typical microstructures of each groups (a): AE (b): T4 (c): S355 (d): S370. (ED: extrusion direction, ND: normal direction and TD: transverse direction).
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
303
Fig. 2. SEM images of secondary phase distribution (a) AE (b) T4 (c) S355 (d) S370. EDS analysis results for (e) Al8Mn5 and (f) Ca2Mg6Zn3.
treatment temperatures, AE and T4 can be classified as belonging to one category, with S355 and S370 belonging to another category. It can be clearly seen that the elastic modulus decreased when the temperature was above 355 °C. All fractured behaviors show no necking effect, the strain work-hardening effect was only observed in T4 specimen. There was no obvious work-hardening effect for the other three groups, because the fracture points of these groups were very close to plastic deformation point. Fig. 6 shows a comparison of tensile mechanical properties. The results show that the AE specimen had quite high UTS and YS. The YS value decreased after T4 treatment, but UTS and elongation were significantly increased. After T4 heat treatment, the Ca2Mg6Zn3 precipitates almost disappeared. The calcium element can effectively dissolve into the matrix via the solid solution effect. Calcium was evenly distributed in the αMg grains and helped the dispersion of the Mg-Zn precipitates. This uniformity of the matrix improves mechanical performance. For the SIMA process under 370 °C, the matrix started melting in the alloying zone, leading to a high volume fraction of segregation at grain boundaries
[27]. At the same time, the solid-solution calcium in the matrix diffused into the liquid phase and segregated to form α-Mg + Ca2Mg6Zn3. Therefore, it was difficult to find Ca in the grains of S370, but there still remained a little amount of calcium in the grains of S355, indicating a better precipitation strengthening effect than S370. After the sample was heat-treated at high semisolid temperature (370 °C), the elongation of S355 and S370 specimens had no apparent increase, and the S370 specimen showed the lowest elongation of all the groups which was only 2.4%. The presence of brittle secondary phases and the LZ significantly affected the mechanical properties of the tested materials. During tensile testing, tensile fracture initiated in the area adjacent to the secondary phase particles and the LZ. The volume fraction and spatial distribution of the secondary phases and the LZ determined the strength and ductility. The elongation and UTS of T4 increased with decreasing volume fraction of secondary phases due to the finer dispersion of the Mg–Zn precipitate in the grains. According to the microstructural results, the S370 specimen tended to form continuous liquid phases with secondary phases along the grain
304
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
400
S355
9
350
AE
T4
Stress (MPa)
300
S370 S355
S370
250 200 150
65 20
100 13
7
T4 50
R5
Unit: mm
3
AE
0 0
20
30
40
50
60
70
2
4
6
8
10
12
14
16
18
20
22
24
Strain (%)
80
2θ Fig. 5. Tensile curves of different microstructures of ZAX1330. Fig. 3. Phase analysis for ZAX1330 and heat-treated samples.
3.3. Effect of microstructure and calcium segregation on corrosion behavior The in-vitro corrosion is commonly used to estimate the corrosion performance of biodegradable magnesium alloy [29]. The corrosion reaction of magnesium can be express as: 2þ
−
Mg→Mg þ 2e ðanodic reactionÞ − − 2H2 O þ 2e →H2 þ 2OH ðcathodic reactionÞ : Mg þ 2H2 O→H2 þ MgðOHÞ2 ðoverall reactionÞ In general, the cathodic reaction as mentioned above is the main reaction of water decomposition to form hydrogen gas. The anodic reaction represents the dissolution of the α-Mg matrix. Fig. 7 shows that the
specimens with liquid phases have relatively high cathodic current density. The AE and T4 specimens had similar cathodic polarization curves. Their cathodic current densities were lower than those of the S355 and S370 specimens. The hydrogen evolution for the AE and T4 specimens was lower than that for the S355 and S370 specimens. The corrosion potential (Ecorr) of AE (−1.51 V) was relatively cathodic. The corrosion potential of T4 was similar to that of AE. Increasing the heat treatment temperature increased the corrosion potential. Notably, the S355 and S370 specimens showed less negative values of Ecorr. A reasonable explanation for this change is that the temperature in the SIMA region (above 355 °C) has led to a higher liquid content. S355 specimen possessed ~ 6% initial liquid phase and S370 possessed ~15% liquid phase (LZ). The shift of corrosion potential was correlated to the fraction of liquid phases to the matrix. The corrosion current density of the different groups presented the following order: T4 (18.47 μA/cm2) b AE (28.92 μA/cm2) b S370 (99.98 μA/cm2) b S355 (167.82 μA/cm2). Lower Icorr means better corrosion resistance. In the anodic polarization curves, no passivation behavior appears for AE, S355 and S370, but a change of current slope at −1.39 V which is called the breakdown potential (Eb) was found for the T4 specimen. This indicates the protective film on the surface of T4 specimen and that the localization corrosion occurred more difficult on the T4 specimen. This protection mechanism has been previously reported [30]. In order to understand the corrosion behavior of different microstructures, all specimens were immersed in r-SBF at 37 °C for 24 h. Fig. 8 shows the corrosion rate and the sequence of corrosion rate for four different specimens from low to high: T4 (1.12 mg/cm2 day) b AE (1.44 mg/cm2 day) b S370 (1.64 mg/cm2 day) b S355 (2.06 mg/cm2 day). 18
400
UTS YS EL
350
Strength (MPa)
300
14 12
250 10 200 8 150 6 100
4
50
2
0
0
AE Fig. 4. TEM observation for Q phase and precipitates in grain interior.
16
T4
S355
S370
Fig. 6. Mechanical properties of ZAX1330 alloys at room temperature.
Elongation (%)
boundaries, and these liquid phases negatively affected YS and ductility. Therefore, the tensile strengths of the T4 and S355 specimens, whose structure was uniform, were higher than those of the AE and S370 specimens. The higher tensile strength can be attributed to the calcium solid solution and the distribution of secondary phases. Furthermore, it is well known that intermetallic morphology affects tensile properties [28]. For example, the needle-shaped particles decreased the tensile properties because cracks were induced easily by the stress concentration at the interface of particles and the matrix. The needle-shaped Ca2Mg6Zn3 segregation also reduces the tensile strength and the ductility of the AE and S370 specimens.
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
Fig. 7. Potentiodynamic polarization curves obtained in r-SBF solution at 37 °C.
The same results were obtained from the electrochemical test, which showed that the corrosion resistance of T4 was much higher than those of the other three groups. To examine the in-vitro degradation mechanisms, the samples were immersed in r-SBF solution for 24 h. Fig. 9 shows the surfaces cleaned with CrO3/AgNO3 solution, which removed the corrosion products after the immersion test. The results show an obvious difference in the surface corrosion morphology. For AE, the corrosion pits were relatively abundant along the clusters of secondary phases, the deep corrosion pits are always nearby Al8Mn5, it can clearly explain that the existence of Al–Mn particle causes the severe localized corrosion. For T4, there was no obvious pitting, and the secondary phases still remained in the matrix even though no pitting occurred near the Al8Mn5 particles. For the S355 and S370 specimens, several pitting appeared in the area surrounding the liquid phase, showing the typical localized corrosion morphology. Comparing Fig. 9 (e and g), it can be clearly seen that the corrosion rate of S370 is lower than that of S355 and that the corrosion path of S370 stopped at the network-like LZ. S355 did not form a continuous barrier against localized corrosion and thus lost a large amount of mass. Moreover, due to the heterogeneous and discontinuous distribution of the initial liquid phases, the corrosion tended to propagate from the corroded regions to the un-corroded regions along the initial liquid phase, showing the typical intergranular corrosion morphology. It is worthy to note that only S355 coexisted the intergranular corrosion and severe pitting corrosion on the surface. In the previous reports on the corrosion mechanism in magnesium alloys, the secondary phase β-Mg17Al12 was found to play the two
2.0
2
Corrosion rate (mg/cm -day)
2.5
1.5
1.0
0.5
0.0
AE
T4
S355
S370
Fig. 8. Corrosion rate calculated from immersion test for ZAX1330.
305
roles. (1) β-Mg17Al12 has high potential, making it an effective corrosion barrier [31]. (2) The theoretical potential of α-Mg is only −2.3 V, a large difference compared to that of β-Mg17Al12; β-Mg17Al12 thus acts as a cathode and enhances the micro-galvanic corrosion effect. For a T6treated AZ91 specimen, the β-Mg17Al12 phase might form a continuous network against corrosion propagation. However, if the distribution and configuration of the β-Mg17Al12 phase do not lead to a connected network, the β-Mg17Al12 phase acts as a galvanic cathode [32]. The Q phase, Al8Mn5, and Ca2Mg6Zn3 perform the same function in ZAX1330 as that of β-Mg17Al12 in AZ91. The inner grains of S355 and S370 had many tiny holes, indicating that the Q phase and Ca2Mg6Zn3 were emptied by the micro-galvanic corrosion effect. Similar evidence has been reported for AM60 magnesium alloy [33]. Corrosion pits initiate near Al8Mn5 or other noble secondary phases. The corrosion path propagates along the secondary phase particles until the secondary phase is removed. Notably, only T4 avoided severe localized corrosion. The uniform corrosion morphology on T4 can be attributed to the lower galvanic current density (Ig), which can be expressed as [34]: Ig ¼ ðφc −φa Þ=ðRa þ Rc þ Rs þ Rm Þ: galvanic current, φc and φa: potentials of cathode and anode, respectively Ra, Rc and Rs resistances of cathode, anode and solution surface, respectively resistance of metallic path from anode to cathode Rm Ig
In this equation, the value of Ig is proportional to (φc − φa). Ig can be decreased by increasing φa or decreasing φc. Furthermore, the TEM images for the T4 specimen and previous reports reveal that the calcium solid solution directly helps the precipitation effect [16]. The evenly distributed high potential Mg–Zn precipitates in the α-Mg matrix not only increased the tensile strength but also increased the anode potential of α-Mg grains. A similar phenomenon was reported by Chang et al., who found that the presence of a precipitate in the Mg matrix reduced the anode-to-cathode potential difference [35]. In the present study, the T4 specimen absorbs the Ca element and other constituent phases to form the nano-precipitates in α-Mg grain, and it can change the nature of α-Mg grain. This phenomenon is referred to as anode precipitate modification (APM). The function of APM was aimed to raise φa and Ra and decrease the potential difference between the matrix and intermetallic phases. Compared to other groups (AE and S370), the calcium segregated. Therefore, calcium dose not participate in APM, and thus the corrosion rate and morphology cannot be improved. According to the references [36,37] reported here, the corrosion mechanism of ZAX1330 in r-SBF is listed systematically in Fig. 10. (1) Grain size: for fine-grained microstructure, the grain boundaries act as a physical corrosion barrier. The fine-grained microstructure contained more grain boundaries, showing good corrosion protection property. The coarse-grained microstructure lacked protection, and the secondary phase detached when micro-galvanic corrosion occurred in the matrix. Thus, pittings initiated and propagated at the corroded sites, which would accelerate the oxidation reaction and corrosion (Fig. 9e). (2) Configuration of secondary phases: for fine-grained microstructure, the uniformity must be improved just like T4 specimen due to more resistance in galvanic corrosion. That is because a great amount of the Zn, Al and Ca atoms of the secondary phases diffuse into the matrix during T4 treatment, balancing the potential difference between matrix and grain boundaries. The Zn and Al atoms stabilize the surface oxide layer, resulting in passivation behavior [38,39]. The Ca atoms are beneficial for improving the APM and herein the resistance of α-Mg is elevated while the nano-sized Mg–Zn precipitates exist in the matrix. Hence, the corrosion would occur on the whole surface instead of the
306
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
Fig. 9. Corrosion morphology of the CrO3/AgNO3 solution cleaned surface: (a, b) AE (c, d)T4 (e, f) S355 (g, h) S370. The circle marked areas were the locations of Al8Mn5 particles.
certain site (Fig.9c), finally resulting in the homogenous corrosion morphology. AE maintained an extrusion texture and didn't treat by any heat treatment process. The grain of AE could not be modified by APM. The secondary phases in the AE specimen aggregated to form a secondary phase region, which caused localized corrosion on the surface. For the coarsed grain microstructure, the S355 specimen possessed the discontinuous liquid phase and lacked corrosion protection, and thus had the highest corrosion rate and the worst corrosion
morphology. In contrast, S370 possessed continuous LZ along the grain boundary to create a new barrier. The network-like LZ can be seen as a new effective barrier to prevent corrosion propagation. 3.4. Cytocompatibility test The colors of culture medium after extraction showed obvious differences, and each group was relatively lighter than negative control. Specifically, the extract of S355 specimen was almost colorless. Because
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
307
Fig. 10. Schematic representation of the relation between corrosion behavior and microstructures.
the high pH value of culture medium affects phenol-red in culture medium which is also called “ deprotonation”[40] (Fig. 11). Fig. 12 shows the MG63 cell viability of cells cultured in the extraction medium for 1, 3, and 6 days. According to the RGR value of MG63 cells, the T4 specimen was the best condition. It can be seen that the cells cultured in T4 extract had the highest absorbance, indicating that T4 extract provided the most suitable environment for MG63 cells. For the S355 extract, the RGR value was lower than that for the negative control. The difference in RGR value between the S355 and T4 extracts was confirmed statistically (p b 0.05) in this study. The RGR values of MG63 cells cultured with AE and S370 extracts were higher than that for S355 in culture period. All the RGR% values were higher than 75%, indicating that the cell cytotoxicity was Grade 1 (no toxicity) according to ISO 10993–5: 1999 [24]. This indicated that all the tested microstructures of ZAX1330 were bio-safe for implantable magnesium alloy. The results of the live/dead cytotoxicity assay show that the AE, T4, and S370 specimens exhibit excellent biocompatibility (Fig. 13). The promotion of cell adhesion and proliferation may be attributed to the proper concentrations of magnesium and zinc ions, which are the main degradation products of ZAX1330 alloy. Zreiqat et al. [41] found that Al2O3 doped with magnesium ions enhanced the adhesion behavior of human-bone-derived cells, which were found to express significantly higher levels of α5β1, α3β1, and β1 integrin receptors than control group, which enhanced the secretion of type I collagen and mediated cell adhesion to biomaterial surfaces. Salih et al. [42] reported that zinc ions may be implicated in cellular transport and proliferation. Furthermore, it has been reported that zinc increases the specific
activity of skeletal ALP in osteoblast-like cells [43]. A healthy morphology and cell adhesion with a high cell area were found in stained AE, T4, and S370 specimens (Fig. 13). In contrast, the green fluorescence of live cells was markedly lower in the S355 extracts compared to that in the negative control and other experimental extracts. S355 had the highest corrosion rate, and thus it is assumed that its non-ideal ion concentrations inhibit cells adhesion and proliferation. In the present study, the cytotoxicity test was applied to ZAX1330 with various microstructures. The results all show acceptable cytocompatibility. Furthermore, the cell morphology and RGR values for MG63 cells in T4 extract were the best. Therefore, the microstructure of ZAX1330 plays an important role in mechanical properties, corrosion behavior, and cytocompatibility. The reliability of microstructure performance can be regulated by controlling the preparation conditions (e.g., heat treatment temperature and uniformity of microstructures).
4. Conclusion Heat treatment process regulated the position of Ca in the ZAX1330. The mechanical properties and corrosion behavior were found to be influenced by microstructure. T4 and S355 (Ca separated in matrix) were chosen according to the homogeneity in the microstructural design. The solid solution of Ca atoms directly helped the uniformity of the distribution of precipitates, leading to good tensile properties. AE and S370 (Ca gathered in specific locations) had significantly reduced tensile elongation. 120
Day-3
Day-6
100 80
RGR(%)
pH 7.8
pH 8.0
AE
Day-1
T4
60 40 20 0
Fig. 11. Experimental extracts for cytotoxicity test and the relation between color and pH value.
Fig. 12. MG63 cell viability expressed as percentage of RGR of cells in control group. All data are presented as means ± SD of six replicate cultures.
308
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
Positive
Negative
AE
T4
S355
S370
Fig. 13. MG63 cultured for 3 days and stained with live/dead cytotoxicity assay. Few apoptotic cells were observed for the S355 and positive control.
Electrochemical polarization curves show that the T4 microstructure had the lowest corrosion density. A short-term immersion test showed that T4 had the best anti-corrosion property. The participation of Ca in the T4 matrix modified the potential difference between the matrix and the secondary phase via APM, which improved the corrosion rate and corrosion morphology. Notably, there was only a 10 °C difference in the heat treatment process of the best microstructure (T4) and the worst microstructure (S355). The in vitro analysis by indirect cytotoxicity and live/dead staining using osteoblast-like MG63 cells showed an inverse correlation with corrosion rate; a lower corrosion rate led to a higher RGR value. A T4heat-treated ZAX1330 has good mechanical properties and biocorrosion resistance, making it a promising biodegradable material. Acknowledgments The authors are grateful to The Instrument Center of National Cheng Kung University, 103-2221-E-006-066 for the financial support of this research. References [1] B. Heublein, R. Rohde, V. Kaese, M. Niemeyer, W. Hartung, A. Haverich, Heart 89 (2003) 651–656. [2] M.P. Staiger, A.M. Pietak, J. Huadmai, G. Dias, Biomaterials 27 (2006) 1728–1734. [3] D.R. Sumner, J.O. Galante, Clin. Orthop. Relat. Res. (1992) 202–212. [4] R.K. Rude, J. Bone Miner. Res. 13 (1998) 749–758. [5] P. Schuck, G. Gammelin, K.L. Resch, Lancet 352 (1998) 1474–1475. [6] L. Yang, Y.D. Huang, F. Feyerabend, R. Willumeit, K.U. Kainer, N. Hort, J. Mech. Behav. Biomed. Mater. 13 (2012) 36–44. [7] J. Kubasek, D. Vojtech, J. Mater. Sci. Mater. Med. 24 (2013) 1615–1626. [8] C.L. Liu, Y.C. Xin, G.Y. Tang, P.K. Chu, Mater. Sci. Eng., A 456 (2007) 350–357. [9] M. Kawahara, M. Kato, Y. Kuroda, Brain Res. Bull. 55 (2001) 211–217. [10] Z.L. Sun, J.C. Wataha, C.T. Hanks, J. Biomed. Mater. Res. 34 (1997) 29–37. [11] W.D. Yang, P. Zhang, J.S. Liu, Y.F. Xue, J. Rare Earths 24 (2006) 369–373. [12] F. Feyerabend, J. Fischer, J. Holtz, F. Witte, R. Willumeit, H. Drucker, C. Vogt, N. Hort, Acta Biomater. 6 (2010) 1834–1842. [13] H. Tapiero, K.D. Tew, Biomed. Pharmacother. 57 (2003) 399–411. [14] E.L. Zhang, D.S. Yin, L.P. Xu, L. Yang, K. Yang, Mater. Sci. Eng. C Biomim. Supramol. Syst. 29 (2009) 987–993.
[15] J. Zhang, Z.X. Guo, F.S. Pan, Z.S. Li, X.D. Luo, Mater. Sci. Eng., A 456 (2007) 43–51. [16] C.J. Bettles, M.A. Gibson, K. Venkatesan, Scr. Mater. 51 (2004) 193–197. [17] Y.C. Lee, A.K. Dahle, D.H. StJohn, Metall. Mater. Trans. A Phys. Metall. Mater. Sci. 31 (2000) 2895–2906. [18] T. Mukai, H. Watanabe, K. Higashi, Magnes. Alloys 2000 (350–3) (2000) 159–170. [19] E.L. Zhang, L. Yang, J.W. Xu, H.Y. Chen, Acta Biomater. 6 (2010) 1756–1762. [20] X.N. Gu, N. Li, Y.F. Zheng, L.Q. Ruan, Mater. Sci. Eng., B 176 (2011) 1778–1784. [21] Z.D. Zhao, Q.A. Chen, Z.J. Tang, C.K. Hu, J. Alloys Compd. 497 (2010) 402–411. [22] ASTM, E112-96 standard test methods for determining average grain size, Annual Book of ASTM Standards, 3.01, 2003, pp. 243–266. [23] ASTM-G31-72, Standard practice for laboratory immersion corrosion testing of metals, Annual Book of ASTM Standards, American Society for Testing and Materials, Philadelphia, PA, 2004. [24] ISO 10993–5, Biological evaluation of medical devices, Part 5. Tests for Cytotoxicity: In Vitro Methods, ANSI/AAMI, Arlington, VA, 1999. [25] Z. Zhang, R. Tremblay, D. Dube, Mater. Sci. Eng., A 385 (2004) 286–291. [26] E.L. Zhang, L. Yang, Mater. Sci. Eng., A 497 (2008) 111–118. [27] S. Kleiner, O. Beffort, P.J. Uggowitzer, Scr. Mater. 51 (2004) 405–410. [28] D. Lin, L. Wang, F.Q. Meng, J.Z. Cui, Q.C. Le, Trans. Nonferrous Metals Soc. China 20 (2010) S421–S425. [29] F. Witte, J. Fischer, J. Nellesen, H.A. Crostack, V. Kaese, A. Pisch, F. Beckmann, H. Windhagen, Biomaterials 27 (2006) 1013–1018. [30] S.H. Cai, T. Lei, N.F. Li, F.F. Feng, Mater. Sci. Eng., C 32 (2012) 2570–2577. [31] O. Lunder, J.E. Lein, T.K. Aune, K. Nisancioglu, Corrosion 45 (1989) 741–748. [32] S. Mathieu, C. Rapin, J. Steinmetz, P. Steinmetz, Corros. Sci. 45 (2003) 2741–2755. [33] R.C. Zeng, J. Zhang, W.J. Huang, W. Dietzel, K.U. Kainer, C. Blawert, W. Ke, Trans. Nonferrous Metals Soc. China 16 (2006) S763–S771. [34] G.S. Wu, J.M. Ibrahim, P.K. Chu, Surf. Coat. Technol. 233 (2013) 2–12. [35] J.W. Chang, X.W. Guo, P.H. Fu, L.M. Peng, W.J. Ding, Electrochim. Acta 52 (2007) 3160–3167. [36] N.N. Aung, W. Zhou, Corros. Sci. 52 (2010) 589–594. [37] M. Alvarez-Lopez, M.D. Pereda, J.A. del Valle, M. Fernandez-Lorenzo, M.C. Garcia-Alonso, O.A. Ruano, M.L. Escudero, Acta Biomater. 6 (2010) 1763–1771. [38] W. Zhou, T. Shen, N.N. Aung, Corros. Sci. 52 (2010) 1035–1041. [39] X.B. Liu, D.Y. Shan, Y.W. Song, E.H. Han, Trans. Nonferrous Metals Soc. China 20 (2010) 1345–1350. [40] J. Fischer, D. Profrock, N. Hort, R. Willumeit, F. Feyerabend, Mater. Sci. Eng., B 176 (2011) 830–834. [41] H. Zreiqat, C.R. Howlett, A. Zannettino, P. Evans, G. Schulze-Tanzil, C. Knabe, M. Shakibaei, J. Biomed. Mater. Res. 62 (2002) 175–184. [42] V. Salih, A. Patel, J.C. Knowles, Biomed. Mater. 2 (2007) 11–20. [43] W.Q. Yan, X. Huang, X.C. Zhong, S.D. Miao, W.J. Weng, K. Cheng, Bioceramics 18 (Pts 1 and 2 309–311) (2006) 735–738.
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Heat treatment mechanism and biodegradable characteristics of ZAX1330 Mg alloy Da-Jun Lin a, Fei-Yi Hung a,⁎, Truan-Sheng Lui a, Ming-Long Yeh b a b
Department of Materials Science and Engineering, National Cheng Kung University, Tainan 701, Taiwan Department of Biomedical Engineering, National Cheng Kung University, Tainan 701, Taiwan
a r t i c l e
i n f o
Article history: Received 26 July 2014 Received in revised form 25 February 2015 Accepted 7 March 2015 Available online 10 March 2015 Keywords: Magnesium alloy Biodegradable Heat treatment Mechanical properties Corrosion Cytocompatibility
a b s t r a c t Heat treatments are key processes in the development of biodegradable magnesium implants. The aim of this study is to investigate the factors of microstructures and metallurgical segregation on the functionality of biodegradable magnesium alloy. The solid solution heat treatment and strain induced melting activation heat treatment were employed to alter the microstructures of ZAX1330 alloy in this study. Heat treatments caused a significant change on grain size and distribution of secondary phases. The fine-grained microstructure enhanced the mechanical strength, corrosion resistance and achieved the lowest degradation rate in simulated body fluid solution. In coarse-grained microstructure systems, grain growth followed liquid phase formation. The corrosion rate increased due to a larger cathodic region. The status of micro-alloyed calcium (in solid solution or segregated) influenced the microstructural evolution mechanisms, mechanical strength, and degradation properties. A cytotoxicity test and a live/dead assay showed that ZAX1330 had good cytocompatibility, which varied with heat treatment, and no cell toxicity. The results suggest that heat treatment should be controlled precisely in order to improve the cytocompatibility of magnesium alloys for application in orthopedic implants. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Magnesium alloys have recently received substantial scientific and clinical interests for use as new generation implant materials in biomedical application [1,2]. Metallic implants made of stainless steel or titanium alloy exhibit stress shielding effect [3]. The density and elastic modulus of magnesium alloys (density: 1.7 g/cm3; Young's modulus: 45 MPa) are close to those of human cortical bone [2], and thus the stress shielding effect of magnesium alloys can be effectively minimized or avoided. In ideal conditions, magnesium alloys provide appropriate mechanical support, and maintain the mechanical strength during the degradation period. Thus, the risk of secondary surgery can be prevented on condition that bone implant is made by magnesium alloys. Magnesium is one of the major essential elements in the human body. A lack of magnesium leads to significant cardiovascular disease and osteoporosis [4]. Magnesium exhibits unique biodegradability behavior, as the degradation product (Mg2+ ions) can be regulated by the metabolism of the kidney, with excess magnesium ions eliminated via renal excretion [5]. Magnesium alloys have many advantages as biomedical materials, and thus have high potential for orthopedic applications. (See Table 1.) One of the useful tools to modify the applicability of magnesium alloys is controlling their alloying composition. Researchers have attempted to solve the problems of low strength and poor corrosion ⁎ Corresponding author. E-mail address: [email protected] (F.-Y. Hung).
http://dx.doi.org/10.1016/j.msec.2015.03.004 0928-4931/© 2015 Elsevier B.V. All rights reserved.
resistance of Mg alloys. Aluminum has been widely reported to improve the mechanical properties and corrosion resistance of magnesium alloys. Many studies have used aluminum and/or rare earth elements (REEs) as alloying elements [6–8]. Magnesium alloys that contain aluminum (e.g., AZ91D) can form a dense passivation film that protects the alloy. Unfortunately, it is well known that a high concentration of aluminum is harmful to neuron cells and osteoblasts [9,10], and is associated with Alzheimer's disease and osteomalacia. The use of RE could lead to hepatotoxicity or have adverse effect on genes. [11,12]. Consequently, aluminum and RE are widely acknowledged as being potentially harmful alloying elements for magnesium materials. To minimize toxicity, magnesium alloys with a low content of harmful alloying elements need to be developed for practical applications. Zinc (Zn) is the second most abundant essential trace element in the human body, with 85% of Zn in the body being in the muscle and bone [13]. Furthermore, Zn alloying can improve the anti-corrosion properties and mechanical properties of magnesium alloys [14]. Moreover, for low aluminum content magnesium alloys, the ratio of Zn to Al needs to increase in improving casting properties and avoiding hot crack effect [15]. Magnesium alloys with low aluminum content provided lower neurotoxicity, they are relatively highly safe material. Notably, scientists tend to add calcium into magnesium alloys to enhance biocompatibility in recent years [16]. Ca is also one of the alloying elements that can modify the microstructure of magnesium alloys. Due to the high GRF (grain restriction factor) value of calcium in magnesium alloy system, the grain refinement effect in the as-cast condition can
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308 Table 1 Results of the electrochemical polarization tests in r-SBF solution.
AE T4 S355 S370
Ecorr (V)
Icorr (μA/cm2)
Rp (Ω/cm2)
CR (mm/year)
−1.51 −1.49 −1.48 −1.46
28.92 18.47 167.82 99.98
459.37 803.92 173.34 440.74
0.78 0.50 4.54 2.46
301
resolution transmission electron microscopy (HR-TEM) analysis was used to investigate the crystalline structures of the major secondary phase by electron diffraction pattern. Tensile test used a mechanical testing system with a tensile speed of 1 mm/min. A typical specimen expressed with a gauge length of 20 mm, a gauge width of 7 mm and a thickness of 3 mm was selected for tensile test. At least three samples are tested for each test. The result is the average of three samples. 2.3. Electrochemical test
be significantly improved. Studies have reported that the grain size decreased with increasing calcium content for content levels of 0 to 0.5 wt.% [17,18]. However, excessive calcium in Mg–Zn–Ca alloy led to the formation of Mg2Ca and Ca2Mg6Zn3, which are brittle phases that decrease ductility and strength. In general, the suggested amount of calcium is below 0.5 wt.% to prevent excess brittle phases in the matrix. Recent studies have reported the influence of microstructure on the corrosion behavior and mechanical properties of Mg alloys. Zhang et al. found that adding Ca and Zn into Mg–Si alloy led to a microstructural transformation of the as-cast matrix [19]. Zheng et al. reported that the tensile strength and corrosion resistance of ZK60 were significantly different for as-cast and as-extrude microstructures, and that the biocompatibility could be enhanced through grain refinement and post-processing [20]. However, little attention has been given to the influences of metallurgical segregation and the distribution of secondary phases. This study investigates the primary research results of Mg–Zn–Al–Ca alloy (ZAX1330) with high Zn (12.83 wt.%), low Al (3.35 wt.%), and trace Ca (0.20 wt.%) on the evolution of microstructure, using SIMA (strain induced melting activation) heat treatment [21] creates an extremely segregated microstructure (Ca atoms gather) compared with the solid-solution (T4) microstructure and the transitional microstructure to investigate the biodegradable characteristics. In addition, the mechanical properties, corrosion behavior and cytocompatibility of Mg–Zn–Al–Ca alloy (ZAX1330) have also been studied in four kinds of microstructure by using different heattreatments. The related outcomes can be expected to provide the preliminary suggestions for medical applications of Bio-Mg alloys.
A conventional three-electrode electrochemical cell was used for the polarization test with a platinum counter electrode and a saturated calomel electrode (SCE) as the reference electrode. The tests were performed in revised simulated body fluid (r-SBF) solution (per liter, dissolved 5.403 g of NaCl, 0.736 g of NaHCO3, 2.036 g of Na2CO3, 0.225 g of KCl, 0.182 g of K2HPO4, 0.310 g of MgCl2 · 6H2O, 11.928 g of 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 0.293 g of CaCl2, 0.072 g of Na2SO4 in deionized water. The r-SBF solution was then buffered to pH 7.4 at 37 °C by adding HEPES and NaOH) at 37 °C by using a water bath. The polarization test used PARSTAT 2273 electrochemistry workstation with a scanning rate of 1 mV s−1 from −1.65 V to −1.3 V. The exposed area of the working electrode to the electrolyte was controlled by a Teflon holder within 1 cm2. 2.4. Immersion test The immersion tests were employed to realize the corrosion mechanism in r-SBF solution according to ASTM G31-72 [23]. The r-SBF volume to surface area ratio was fixed at 20 mL/cm2. The experimental samples were removed after 24 h of immersion, rinsed with distilled water and cleaned with boiled chromic acid/silver nitrate solution (200 g/L CrO3 + 10 g/L AgNO3) to remove the corrosion products. The corrosion surfaces were examined by SEM. The degradation rates were obtained according to the following equation: C = (m1 − m2) / St (m1: weight before corrosion, m2: weight after corrosion, S: surface area, t: immersion time)
2. Materials and methods
2.5. Cell culture and in-vitro test
2.1. Material preparation
Human osteoblast-like cells (MG63) were used to examine the cytocompatibility of ZAX1330. Before cytotoxicity tests, the MG63 cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) with high glucose, supplemented with 10% fetal bovine serum (FBS), and 1% antibiotics. Cells were sub-cultured at least twice a week and were kept in a humidified atmosphere of 95% air and 5% CO2. The cytotoxicity tests were carried out by indirect method. Extracts were prepared according to ISO 10993–5 [24]. After 24 h of incubation in the cell incubator at 37 °C, each group of the extraction medium was passed through a 0.22 μm filter paper to remove corrosion particles. The negative control group used culture medium and the positive control group used culture medium with 5% dimethyl sulfoxide (DMSO). Cells were cultured in 96-well plates at 5 × 103 cells/100 μL medium in each well and incubated for 24 h to allow for initial cell adhesion. The medium was replaced with 100 μL of experimental extracts. After the cells were incubated for 1, 3, and 6 days, respectively, 10 μL of CellTiter 96 solution was added into each well. All solutions were incubated for 4 h in a cell incubator, and then 100 μL of CellTiter-96-treated medium solution was sucked out and placed into new 96-well plates. The absorbance of the CellTiter-96-treated solution was measured using an enzyme-linked immunosorbent assay (ELISA) reader at a wavelength of 490 nm. After the experiment, the cell relative growth rate (RGR) was calculated according to the following formula:
The extruded ZAX1330 alloy (12.83 wt.% Zn, 3.35 wt.% Al, 0.20 wt.% Ca) was used in this study. ZAX1330 was processed into a circular plate (12.7 mm in diameter) using a computer numerical control (CNC) wire cutting machine. The as-extruded alloy is denoted as the AE specimen. Solution treatment (T4) of the alloy was carried out at 345 °C for 10 h in an electrical furnace, followed by water quenching at room temperature. The SIMA process was used to create a Ca-rich zone microstructure. The process was performed at 370 °C for 1 h to produce coarse grains (spheroidal α-Mg grains) with a high fraction of the liquid phase (Ca-rich zone) in the matrix (S370 specimen). A medium temperature of 355 °C was chosen to investigate the transition microstructure from equiaxed grains to spheroidal grains. This temperature was held for 1 h (S355 specimen). 2.2. Microstructure and mechanical test All ZAX1330 samples (AE, T4, S355, S370) for microstructural observation were polished up to 0.05 μm and etched. Optical microscopy (OM) and scanning electron microscopy (SEM) with an energy dispersive spectrometer (EDS) were used for microstructural observation. The grain size was determined by the linear intercept method according to ASTM E112-96 standard [22]. The constituent phases were characterized by X-ray diffraction meter (XRD) with Cu-Kα radiation. Diffraction patterns were obtained between 2θ values of 20–80°. The high
h i RGR ¼ ðODtest −ODblank Þ= ODnegative −ODblank 100%:
302
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
Cell cytotoxicity measurements were given as the mean RGR value ± standard deviation, and the analysis of one-way variance (ANOVA) was conducted to evaluate the statistical significance of differences. Difference at p ≦ 0.05 was considered to be statistically significant. For live/dead staining, the cell culture media were removed after 3 days of incubation and the wells were rinsed with phosphate buffer solution (PBS), and 200 μL of dye was settled in 96 wells to stain the cells at 37 °C. After staining for 30 min, the dye was cleaned by PBS to remove unreacted dye and non-adhered cells. The labeled cells were observed using a fluorescence microscopy at wavelengths of 530 nm (live cells) and 645 nm (dead cells). 3. Results and discussion 3.1. Microstructural characteristics The microstructures of all specimens are shown in Fig. 1. In asextruded ZAX1330 (Fig. 1a, AE), the average grain size of the primary α-Mg was about 2 μm and the clusters of the secondary phases were aligned in the direction of extrusion. After solid-solution treatment (Fig. 1b, T4), the grain size of the matrix increased to 5 μm. The secondary phases of the T4 specimen were distributed uniformly. EDS analysis shows a zinc content of 3.84 at.% and an aluminum content of 1.83 at.% in the secondary phase. Fig. 1c shows the microstructure of S355 for heat treatment at 1 h at 355 °C, and the grain size significantly increased and partially melted secondary phases re-aggregated at the grain boundaries to form initial liquid phases. When the heat treatment temperature was increased to 370 °C (Fig. 1d, S370), the secondary phases disappeared during heat treatment, and metallurgical segregation was
observed along the grain boundary, which led to the formation of a liquid zone (LZ). All equiaxed grains evolved into coarse α-Mg spheroidal grains. The EDS analysis of Fig. 2 shows that the LZ consisted of non-equilibrium phases and intermetallic compounds (IMC). There were several Ca-rich needle-like compounds which were confirmed as Ca2Mg6Zn3 in the LZ of the S370 matrix. Most of the secondary phases of S370 aggregated at the grain boundaries, there were still few Al8Mn5 and Mg5Zn2Al2 in grains. Fig. 3 shows the X-ray diffraction patterns of all samples. The main secondary phase is Mg32(Al, Zn)49. According to TEM of T4 (Fig. 4), there was 0.8 at.% calcium in α-Mg, Mg32(Al, Zn)49 was found to be in a ternary icosahedral quasi-crystalline phase (Q phase) instead of the τ phase, which has been reported to be a body-centered cubic structure (space group Im3, a = 1.416 nm) [25]. Notably, the Q-phase in S355 and S370 specimens was significantly lower than that in the T4 specimen, but the intensity of α-Mg showed no significant difference. This result can be explained by the Q-phase melting at high temperatures. From the results of XRD, Ca2Mg6Zn3 was in AE and S370 specimen, whereas X-ray didn't detect other intermetallic phases consisted with Ca (e.g., Al2Ca). In the ternary phase diagram of Mg–Zn–Ca, the eutectic phases are α-Mg + Ca2Mg6Zn3 + Mg2Ca when the Zn/Ca atomic ratio is less than 1.2 [26]. On the other hand, other kinds of eutectic phases are αMg + Ca2Mg6Zn3 which form while the Zn/Ca atomic ratio is more than 1.2. The results presented here confirm that α-Mg and Ca2Mg6Zn3 coexisted in the matrix (the Zn/Ca ratio for ZAX1330 is higher than 1.2). 3.2. Effect of metallurgical segregation on mechanical properties Fig. 5 shows the stress–strain curves of AE, T4, S355 and S370 specimens. The elastic modulus decreased with the increase of heat
Fig. 1. Typical microstructures of each groups (a): AE (b): T4 (c): S355 (d): S370. (ED: extrusion direction, ND: normal direction and TD: transverse direction).
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
303
Fig. 2. SEM images of secondary phase distribution (a) AE (b) T4 (c) S355 (d) S370. EDS analysis results for (e) Al8Mn5 and (f) Ca2Mg6Zn3.
treatment temperatures, AE and T4 can be classified as belonging to one category, with S355 and S370 belonging to another category. It can be clearly seen that the elastic modulus decreased when the temperature was above 355 °C. All fractured behaviors show no necking effect, the strain work-hardening effect was only observed in T4 specimen. There was no obvious work-hardening effect for the other three groups, because the fracture points of these groups were very close to plastic deformation point. Fig. 6 shows a comparison of tensile mechanical properties. The results show that the AE specimen had quite high UTS and YS. The YS value decreased after T4 treatment, but UTS and elongation were significantly increased. After T4 heat treatment, the Ca2Mg6Zn3 precipitates almost disappeared. The calcium element can effectively dissolve into the matrix via the solid solution effect. Calcium was evenly distributed in the αMg grains and helped the dispersion of the Mg-Zn precipitates. This uniformity of the matrix improves mechanical performance. For the SIMA process under 370 °C, the matrix started melting in the alloying zone, leading to a high volume fraction of segregation at grain boundaries
[27]. At the same time, the solid-solution calcium in the matrix diffused into the liquid phase and segregated to form α-Mg + Ca2Mg6Zn3. Therefore, it was difficult to find Ca in the grains of S370, but there still remained a little amount of calcium in the grains of S355, indicating a better precipitation strengthening effect than S370. After the sample was heat-treated at high semisolid temperature (370 °C), the elongation of S355 and S370 specimens had no apparent increase, and the S370 specimen showed the lowest elongation of all the groups which was only 2.4%. The presence of brittle secondary phases and the LZ significantly affected the mechanical properties of the tested materials. During tensile testing, tensile fracture initiated in the area adjacent to the secondary phase particles and the LZ. The volume fraction and spatial distribution of the secondary phases and the LZ determined the strength and ductility. The elongation and UTS of T4 increased with decreasing volume fraction of secondary phases due to the finer dispersion of the Mg–Zn precipitate in the grains. According to the microstructural results, the S370 specimen tended to form continuous liquid phases with secondary phases along the grain
304
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
400
S355
9
350
AE
T4
Stress (MPa)
300
S370 S355
S370
250 200 150
65 20
100 13
7
T4 50
R5
Unit: mm
3
AE
0 0
20
30
40
50
60
70
2
4
6
8
10
12
14
16
18
20
22
24
Strain (%)
80
2θ Fig. 5. Tensile curves of different microstructures of ZAX1330. Fig. 3. Phase analysis for ZAX1330 and heat-treated samples.
3.3. Effect of microstructure and calcium segregation on corrosion behavior The in-vitro corrosion is commonly used to estimate the corrosion performance of biodegradable magnesium alloy [29]. The corrosion reaction of magnesium can be express as: 2þ
−
Mg→Mg þ 2e ðanodic reactionÞ − − 2H2 O þ 2e →H2 þ 2OH ðcathodic reactionÞ : Mg þ 2H2 O→H2 þ MgðOHÞ2 ðoverall reactionÞ In general, the cathodic reaction as mentioned above is the main reaction of water decomposition to form hydrogen gas. The anodic reaction represents the dissolution of the α-Mg matrix. Fig. 7 shows that the
specimens with liquid phases have relatively high cathodic current density. The AE and T4 specimens had similar cathodic polarization curves. Their cathodic current densities were lower than those of the S355 and S370 specimens. The hydrogen evolution for the AE and T4 specimens was lower than that for the S355 and S370 specimens. The corrosion potential (Ecorr) of AE (−1.51 V) was relatively cathodic. The corrosion potential of T4 was similar to that of AE. Increasing the heat treatment temperature increased the corrosion potential. Notably, the S355 and S370 specimens showed less negative values of Ecorr. A reasonable explanation for this change is that the temperature in the SIMA region (above 355 °C) has led to a higher liquid content. S355 specimen possessed ~ 6% initial liquid phase and S370 possessed ~15% liquid phase (LZ). The shift of corrosion potential was correlated to the fraction of liquid phases to the matrix. The corrosion current density of the different groups presented the following order: T4 (18.47 μA/cm2) b AE (28.92 μA/cm2) b S370 (99.98 μA/cm2) b S355 (167.82 μA/cm2). Lower Icorr means better corrosion resistance. In the anodic polarization curves, no passivation behavior appears for AE, S355 and S370, but a change of current slope at −1.39 V which is called the breakdown potential (Eb) was found for the T4 specimen. This indicates the protective film on the surface of T4 specimen and that the localization corrosion occurred more difficult on the T4 specimen. This protection mechanism has been previously reported [30]. In order to understand the corrosion behavior of different microstructures, all specimens were immersed in r-SBF at 37 °C for 24 h. Fig. 8 shows the corrosion rate and the sequence of corrosion rate for four different specimens from low to high: T4 (1.12 mg/cm2 day) b AE (1.44 mg/cm2 day) b S370 (1.64 mg/cm2 day) b S355 (2.06 mg/cm2 day). 18
400
UTS YS EL
350
Strength (MPa)
300
14 12
250 10 200 8 150 6 100
4
50
2
0
0
AE Fig. 4. TEM observation for Q phase and precipitates in grain interior.
16
T4
S355
S370
Fig. 6. Mechanical properties of ZAX1330 alloys at room temperature.
Elongation (%)
boundaries, and these liquid phases negatively affected YS and ductility. Therefore, the tensile strengths of the T4 and S355 specimens, whose structure was uniform, were higher than those of the AE and S370 specimens. The higher tensile strength can be attributed to the calcium solid solution and the distribution of secondary phases. Furthermore, it is well known that intermetallic morphology affects tensile properties [28]. For example, the needle-shaped particles decreased the tensile properties because cracks were induced easily by the stress concentration at the interface of particles and the matrix. The needle-shaped Ca2Mg6Zn3 segregation also reduces the tensile strength and the ductility of the AE and S370 specimens.
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
Fig. 7. Potentiodynamic polarization curves obtained in r-SBF solution at 37 °C.
The same results were obtained from the electrochemical test, which showed that the corrosion resistance of T4 was much higher than those of the other three groups. To examine the in-vitro degradation mechanisms, the samples were immersed in r-SBF solution for 24 h. Fig. 9 shows the surfaces cleaned with CrO3/AgNO3 solution, which removed the corrosion products after the immersion test. The results show an obvious difference in the surface corrosion morphology. For AE, the corrosion pits were relatively abundant along the clusters of secondary phases, the deep corrosion pits are always nearby Al8Mn5, it can clearly explain that the existence of Al–Mn particle causes the severe localized corrosion. For T4, there was no obvious pitting, and the secondary phases still remained in the matrix even though no pitting occurred near the Al8Mn5 particles. For the S355 and S370 specimens, several pitting appeared in the area surrounding the liquid phase, showing the typical localized corrosion morphology. Comparing Fig. 9 (e and g), it can be clearly seen that the corrosion rate of S370 is lower than that of S355 and that the corrosion path of S370 stopped at the network-like LZ. S355 did not form a continuous barrier against localized corrosion and thus lost a large amount of mass. Moreover, due to the heterogeneous and discontinuous distribution of the initial liquid phases, the corrosion tended to propagate from the corroded regions to the un-corroded regions along the initial liquid phase, showing the typical intergranular corrosion morphology. It is worthy to note that only S355 coexisted the intergranular corrosion and severe pitting corrosion on the surface. In the previous reports on the corrosion mechanism in magnesium alloys, the secondary phase β-Mg17Al12 was found to play the two
2.0
2
Corrosion rate (mg/cm -day)
2.5
1.5
1.0
0.5
0.0
AE
T4
S355
S370
Fig. 8. Corrosion rate calculated from immersion test for ZAX1330.
305
roles. (1) β-Mg17Al12 has high potential, making it an effective corrosion barrier [31]. (2) The theoretical potential of α-Mg is only −2.3 V, a large difference compared to that of β-Mg17Al12; β-Mg17Al12 thus acts as a cathode and enhances the micro-galvanic corrosion effect. For a T6treated AZ91 specimen, the β-Mg17Al12 phase might form a continuous network against corrosion propagation. However, if the distribution and configuration of the β-Mg17Al12 phase do not lead to a connected network, the β-Mg17Al12 phase acts as a galvanic cathode [32]. The Q phase, Al8Mn5, and Ca2Mg6Zn3 perform the same function in ZAX1330 as that of β-Mg17Al12 in AZ91. The inner grains of S355 and S370 had many tiny holes, indicating that the Q phase and Ca2Mg6Zn3 were emptied by the micro-galvanic corrosion effect. Similar evidence has been reported for AM60 magnesium alloy [33]. Corrosion pits initiate near Al8Mn5 or other noble secondary phases. The corrosion path propagates along the secondary phase particles until the secondary phase is removed. Notably, only T4 avoided severe localized corrosion. The uniform corrosion morphology on T4 can be attributed to the lower galvanic current density (Ig), which can be expressed as [34]: Ig ¼ ðφc −φa Þ=ðRa þ Rc þ Rs þ Rm Þ: galvanic current, φc and φa: potentials of cathode and anode, respectively Ra, Rc and Rs resistances of cathode, anode and solution surface, respectively resistance of metallic path from anode to cathode Rm Ig
In this equation, the value of Ig is proportional to (φc − φa). Ig can be decreased by increasing φa or decreasing φc. Furthermore, the TEM images for the T4 specimen and previous reports reveal that the calcium solid solution directly helps the precipitation effect [16]. The evenly distributed high potential Mg–Zn precipitates in the α-Mg matrix not only increased the tensile strength but also increased the anode potential of α-Mg grains. A similar phenomenon was reported by Chang et al., who found that the presence of a precipitate in the Mg matrix reduced the anode-to-cathode potential difference [35]. In the present study, the T4 specimen absorbs the Ca element and other constituent phases to form the nano-precipitates in α-Mg grain, and it can change the nature of α-Mg grain. This phenomenon is referred to as anode precipitate modification (APM). The function of APM was aimed to raise φa and Ra and decrease the potential difference between the matrix and intermetallic phases. Compared to other groups (AE and S370), the calcium segregated. Therefore, calcium dose not participate in APM, and thus the corrosion rate and morphology cannot be improved. According to the references [36,37] reported here, the corrosion mechanism of ZAX1330 in r-SBF is listed systematically in Fig. 10. (1) Grain size: for fine-grained microstructure, the grain boundaries act as a physical corrosion barrier. The fine-grained microstructure contained more grain boundaries, showing good corrosion protection property. The coarse-grained microstructure lacked protection, and the secondary phase detached when micro-galvanic corrosion occurred in the matrix. Thus, pittings initiated and propagated at the corroded sites, which would accelerate the oxidation reaction and corrosion (Fig. 9e). (2) Configuration of secondary phases: for fine-grained microstructure, the uniformity must be improved just like T4 specimen due to more resistance in galvanic corrosion. That is because a great amount of the Zn, Al and Ca atoms of the secondary phases diffuse into the matrix during T4 treatment, balancing the potential difference between matrix and grain boundaries. The Zn and Al atoms stabilize the surface oxide layer, resulting in passivation behavior [38,39]. The Ca atoms are beneficial for improving the APM and herein the resistance of α-Mg is elevated while the nano-sized Mg–Zn precipitates exist in the matrix. Hence, the corrosion would occur on the whole surface instead of the
306
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
Fig. 9. Corrosion morphology of the CrO3/AgNO3 solution cleaned surface: (a, b) AE (c, d)T4 (e, f) S355 (g, h) S370. The circle marked areas were the locations of Al8Mn5 particles.
certain site (Fig.9c), finally resulting in the homogenous corrosion morphology. AE maintained an extrusion texture and didn't treat by any heat treatment process. The grain of AE could not be modified by APM. The secondary phases in the AE specimen aggregated to form a secondary phase region, which caused localized corrosion on the surface. For the coarsed grain microstructure, the S355 specimen possessed the discontinuous liquid phase and lacked corrosion protection, and thus had the highest corrosion rate and the worst corrosion
morphology. In contrast, S370 possessed continuous LZ along the grain boundary to create a new barrier. The network-like LZ can be seen as a new effective barrier to prevent corrosion propagation. 3.4. Cytocompatibility test The colors of culture medium after extraction showed obvious differences, and each group was relatively lighter than negative control. Specifically, the extract of S355 specimen was almost colorless. Because
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
307
Fig. 10. Schematic representation of the relation between corrosion behavior and microstructures.
the high pH value of culture medium affects phenol-red in culture medium which is also called “ deprotonation”[40] (Fig. 11). Fig. 12 shows the MG63 cell viability of cells cultured in the extraction medium for 1, 3, and 6 days. According to the RGR value of MG63 cells, the T4 specimen was the best condition. It can be seen that the cells cultured in T4 extract had the highest absorbance, indicating that T4 extract provided the most suitable environment for MG63 cells. For the S355 extract, the RGR value was lower than that for the negative control. The difference in RGR value between the S355 and T4 extracts was confirmed statistically (p b 0.05) in this study. The RGR values of MG63 cells cultured with AE and S370 extracts were higher than that for S355 in culture period. All the RGR% values were higher than 75%, indicating that the cell cytotoxicity was Grade 1 (no toxicity) according to ISO 10993–5: 1999 [24]. This indicated that all the tested microstructures of ZAX1330 were bio-safe for implantable magnesium alloy. The results of the live/dead cytotoxicity assay show that the AE, T4, and S370 specimens exhibit excellent biocompatibility (Fig. 13). The promotion of cell adhesion and proliferation may be attributed to the proper concentrations of magnesium and zinc ions, which are the main degradation products of ZAX1330 alloy. Zreiqat et al. [41] found that Al2O3 doped with magnesium ions enhanced the adhesion behavior of human-bone-derived cells, which were found to express significantly higher levels of α5β1, α3β1, and β1 integrin receptors than control group, which enhanced the secretion of type I collagen and mediated cell adhesion to biomaterial surfaces. Salih et al. [42] reported that zinc ions may be implicated in cellular transport and proliferation. Furthermore, it has been reported that zinc increases the specific
activity of skeletal ALP in osteoblast-like cells [43]. A healthy morphology and cell adhesion with a high cell area were found in stained AE, T4, and S370 specimens (Fig. 13). In contrast, the green fluorescence of live cells was markedly lower in the S355 extracts compared to that in the negative control and other experimental extracts. S355 had the highest corrosion rate, and thus it is assumed that its non-ideal ion concentrations inhibit cells adhesion and proliferation. In the present study, the cytotoxicity test was applied to ZAX1330 with various microstructures. The results all show acceptable cytocompatibility. Furthermore, the cell morphology and RGR values for MG63 cells in T4 extract were the best. Therefore, the microstructure of ZAX1330 plays an important role in mechanical properties, corrosion behavior, and cytocompatibility. The reliability of microstructure performance can be regulated by controlling the preparation conditions (e.g., heat treatment temperature and uniformity of microstructures).
4. Conclusion Heat treatment process regulated the position of Ca in the ZAX1330. The mechanical properties and corrosion behavior were found to be influenced by microstructure. T4 and S355 (Ca separated in matrix) were chosen according to the homogeneity in the microstructural design. The solid solution of Ca atoms directly helped the uniformity of the distribution of precipitates, leading to good tensile properties. AE and S370 (Ca gathered in specific locations) had significantly reduced tensile elongation. 120
Day-3
Day-6
100 80
RGR(%)
pH 7.8
pH 8.0
AE
Day-1
T4
60 40 20 0
Fig. 11. Experimental extracts for cytotoxicity test and the relation between color and pH value.
Fig. 12. MG63 cell viability expressed as percentage of RGR of cells in control group. All data are presented as means ± SD of six replicate cultures.
308
D.-J. Lin et al. / Materials Science and Engineering C 51 (2015) 300–308
Positive
Negative
AE
T4
S355
S370
Fig. 13. MG63 cultured for 3 days and stained with live/dead cytotoxicity assay. Few apoptotic cells were observed for the S355 and positive control.
Electrochemical polarization curves show that the T4 microstructure had the lowest corrosion density. A short-term immersion test showed that T4 had the best anti-corrosion property. The participation of Ca in the T4 matrix modified the potential difference between the matrix and the secondary phase via APM, which improved the corrosion rate and corrosion morphology. Notably, there was only a 10 °C difference in the heat treatment process of the best microstructure (T4) and the worst microstructure (S355). The in vitro analysis by indirect cytotoxicity and live/dead staining using osteoblast-like MG63 cells showed an inverse correlation with corrosion rate; a lower corrosion rate led to a higher RGR value. A T4heat-treated ZAX1330 has good mechanical properties and biocorrosion resistance, making it a promising biodegradable material. Acknowledgments The authors are grateful to The Instrument Center of National Cheng Kung University, 103-2221-E-006-066 for the financial support of this research. References [1] B. Heublein, R. Rohde, V. Kaese, M. Niemeyer, W. Hartung, A. Haverich, Heart 89 (2003) 651–656. [2] M.P. Staiger, A.M. Pietak, J. Huadmai, G. Dias, Biomaterials 27 (2006) 1728–1734. [3] D.R. Sumner, J.O. Galante, Clin. Orthop. Relat. Res. (1992) 202–212. [4] R.K. Rude, J. Bone Miner. Res. 13 (1998) 749–758. [5] P. Schuck, G. Gammelin, K.L. Resch, Lancet 352 (1998) 1474–1475. [6] L. Yang, Y.D. Huang, F. Feyerabend, R. Willumeit, K.U. Kainer, N. Hort, J. Mech. Behav. Biomed. Mater. 13 (2012) 36–44. [7] J. Kubasek, D. Vojtech, J. Mater. Sci. Mater. Med. 24 (2013) 1615–1626. [8] C.L. Liu, Y.C. Xin, G.Y. Tang, P.K. Chu, Mater. Sci. Eng., A 456 (2007) 350–357. [9] M. Kawahara, M. Kato, Y. Kuroda, Brain Res. Bull. 55 (2001) 211–217. [10] Z.L. Sun, J.C. Wataha, C.T. Hanks, J. Biomed. Mater. Res. 34 (1997) 29–37. [11] W.D. Yang, P. Zhang, J.S. Liu, Y.F. Xue, J. Rare Earths 24 (2006) 369–373. [12] F. Feyerabend, J. Fischer, J. Holtz, F. Witte, R. Willumeit, H. Drucker, C. Vogt, N. Hort, Acta Biomater. 6 (2010) 1834–1842. [13] H. Tapiero, K.D. Tew, Biomed. Pharmacother. 57 (2003) 399–411. [14] E.L. Zhang, D.S. Yin, L.P. Xu, L. Yang, K. Yang, Mater. Sci. Eng. C Biomim. Supramol. Syst. 29 (2009) 987–993.
[15] J. Zhang, Z.X. Guo, F.S. Pan, Z.S. Li, X.D. Luo, Mater. Sci. Eng., A 456 (2007) 43–51. [16] C.J. Bettles, M.A. Gibson, K. Venkatesan, Scr. Mater. 51 (2004) 193–197. [17] Y.C. Lee, A.K. Dahle, D.H. StJohn, Metall. Mater. Trans. A Phys. Metall. Mater. Sci. 31 (2000) 2895–2906. [18] T. Mukai, H. Watanabe, K. Higashi, Magnes. Alloys 2000 (350–3) (2000) 159–170. [19] E.L. Zhang, L. Yang, J.W. Xu, H.Y. Chen, Acta Biomater. 6 (2010) 1756–1762. [20] X.N. Gu, N. Li, Y.F. Zheng, L.Q. Ruan, Mater. Sci. Eng., B 176 (2011) 1778–1784. [21] Z.D. Zhao, Q.A. Chen, Z.J. Tang, C.K. Hu, J. Alloys Compd. 497 (2010) 402–411. [22] ASTM, E112-96 standard test methods for determining average grain size, Annual Book of ASTM Standards, 3.01, 2003, pp. 243–266. [23] ASTM-G31-72, Standard practice for laboratory immersion corrosion testing of metals, Annual Book of ASTM Standards, American Society for Testing and Materials, Philadelphia, PA, 2004. [24] ISO 10993–5, Biological evaluation of medical devices, Part 5. Tests for Cytotoxicity: In Vitro Methods, ANSI/AAMI, Arlington, VA, 1999. [25] Z. Zhang, R. Tremblay, D. Dube, Mater. Sci. Eng., A 385 (2004) 286–291. [26] E.L. Zhang, L. Yang, Mater. Sci. Eng., A 497 (2008) 111–118. [27] S. Kleiner, O. Beffort, P.J. Uggowitzer, Scr. Mater. 51 (2004) 405–410. [28] D. Lin, L. Wang, F.Q. Meng, J.Z. Cui, Q.C. Le, Trans. Nonferrous Metals Soc. China 20 (2010) S421–S425. [29] F. Witte, J. Fischer, J. Nellesen, H.A. Crostack, V. Kaese, A. Pisch, F. Beckmann, H. Windhagen, Biomaterials 27 (2006) 1013–1018. [30] S.H. Cai, T. Lei, N.F. Li, F.F. Feng, Mater. Sci. Eng., C 32 (2012) 2570–2577. [31] O. Lunder, J.E. Lein, T.K. Aune, K. Nisancioglu, Corrosion 45 (1989) 741–748. [32] S. Mathieu, C. Rapin, J. Steinmetz, P. Steinmetz, Corros. Sci. 45 (2003) 2741–2755. [33] R.C. Zeng, J. Zhang, W.J. Huang, W. Dietzel, K.U. Kainer, C. Blawert, W. Ke, Trans. Nonferrous Metals Soc. China 16 (2006) S763–S771. [34] G.S. Wu, J.M. Ibrahim, P.K. Chu, Surf. Coat. Technol. 233 (2013) 2–12. [35] J.W. Chang, X.W. Guo, P.H. Fu, L.M. Peng, W.J. Ding, Electrochim. Acta 52 (2007) 3160–3167. [36] N.N. Aung, W. Zhou, Corros. Sci. 52 (2010) 589–594. [37] M. Alvarez-Lopez, M.D. Pereda, J.A. del Valle, M. Fernandez-Lorenzo, M.C. Garcia-Alonso, O.A. Ruano, M.L. Escudero, Acta Biomater. 6 (2010) 1763–1771. [38] W. Zhou, T. Shen, N.N. Aung, Corros. Sci. 52 (2010) 1035–1041. [39] X.B. Liu, D.Y. Shan, Y.W. Song, E.H. Han, Trans. Nonferrous Metals Soc. China 20 (2010) 1345–1350. [40] J. Fischer, D. Profrock, N. Hort, R. Willumeit, F. Feyerabend, Mater. Sci. Eng., B 176 (2011) 830–834. [41] H. Zreiqat, C.R. Howlett, A. Zannettino, P. Evans, G. Schulze-Tanzil, C. Knabe, M. Shakibaei, J. Biomed. Mater. Res. 62 (2002) 175–184. [42] V. Salih, A. Patel, J.C. Knowles, Biomed. Mater. 2 (2007) 11–20. [43] W.Q. Yan, X. Huang, X.C. Zhong, S.D. Miao, W.J. Weng, K. Cheng, Bioceramics 18 (Pts 1 and 2 309–311) (2006) 735–738.