- Email: [email protected]
Mechanical and corrosion properties of newly developed biodegradable Zn-based alloys for bone fixation
Acta Biomaterialia 7 (2011) 3515–3522
Contents lists available at ScienceDirect
Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat
Mechanical and corrosion properties of newly developed biodegradable Zn-based alloys for bone fixation D. Vojteˇch ⇑, J. Kubásek, J. Šerák, P. Novák Department of Metals and Corrosion Engineering, Institute of Chemical Technology, Technická 5, 166 28 Prague 6, Czech Republic
a r t i c l e
i n f o
Article history: Received 12 January 2011 Received in revised form 20 April 2011 Accepted 9 May 2011 Available online 14 May 2011 Keywords: Biodegradable material Zinc Bone fixation Mechanical properties Corrosion
a b s t r a c t In the present work Zn–Mg alloys containing up to 3 wt.% Mg were studied as potential biodegradable materials for medical use. The structure, mechanical properties and corrosion behavior of these alloys were investigated and compared with those of pure Mg, AZ91HP and casting Zn–Al–Cu alloys. The structures were examined by light and scanning electron microscopy (SEM), and tensile and hardness testing were used to characterize the mechanical properties of the alloys. The corrosion behavior of the materials in simulated body fluid with pH values of 5, 7 and 10 was determined by immersion tests, potentiodynamic measurements and by monitoring the pH value evolution during corrosion. The surfaces of the corroded alloys were investigated by SEM, energy-dispersive spectrometry and X-ray photoelectron spectroscopy. It was found that a maximum strength and elongation of 150 MPa and 2%, respectively, were achieved at Mg contents of approximately 1 wt.%. These mechanical properties are discussed in relation to the structural features of the alloys. The corrosion rates of the Zn–Mg alloys were determined to be significantly lower than those of Mg and AZ91HP alloys. The former alloys corroded at rates of the order of tens of microns per year, whereas the corrosion rates of the latter were of the order of hundreds of microns per year. Possible zinc doses and toxicity were estimated from the corrosion behavior of the zinc alloys. It was found that these doses are negligible compared with the tolerable biological daily limit of zinc. Ó 2011 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
1. Introduction Biodegradable materials are capable of progressively degrading in the human body to produce non-toxic compounds that can be readily excreted. Biodegradable polymeric materials have been known and used for a long time as, for example, fibers for the reparation of damaged tissue or for winding biodegradable stents. However, for load-bearing applications, such as screws, plates or other fixations of fractured bones, polymeric materials are not suitable due to their low mechanical strength [1]. Among biodegradable metallic materials magnesium has attracted the greatest interest, and its properties related to bioapplications have been studied since the beginning of the 20th century [2]. The reason for this is that magnesium is non-toxic to the human body and excessive amounts of it can be readily excreted by the kidneys. Magnesium is also very important for biological functions of the human body. The main disadvantage of most magnesium alloys is that they corrode too rapidly in physiological environments. The corrosion of magnesium produces hydrogen pockets near the implant, which retard the healing ⇑ Corresponding author. Tel.: +420 220444290; fax: +420 220444400. E-mail address: [email protected] (D. Vojteˇch).
process. In addition, the local increase in alkalinity close to the magnesium implant also has adverse effects on healing [2–6]. Therefore, great effort has been exerted during the last 20 years to find magnesium-based alloys that corrode at acceptably low rates and whose corrosion products, like hydrogen and alkaline materials, can be absorbed by the surrounding tissue without negative effects on the healing process. The magnesium alloys initially considered for bio-applications were based on alloying systems originally developed for engineering applications. Among them AZ, LAE and WE type alloys have been widely studied, but only the WE43 alloy has been used to prepare a biodegradable implant, in this case a vascular stent, used in the human body [7–9]. Besides engineering alloys, some new alloying systems have been developed to provide good corrosion resistance and biocompatibility. Among them Mg–Zn, Mg–Zn–Mn–Ca, Mg–Zn–Y, Mg–Gd, Mg–Zn–Si and other systems have recently been studied [10–15]. In these systems zinc is often the major constituent. Zinc, as a more noble metal than magnesium, is well known to positively affect the corrosion resistance and strength of Mg. From a biological point of view zinc is very important for biological functions in the human body because it is involved in various aspects of cellular metabolism. Zinc is important to the proper function of numerous enzymes and it supports immune
1742-7061/$ - see front matter Ó 2011 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.actbio.2011.05.008
3516
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522
functions, protein and DNA synthesis and wound healing. It also supports normal growth and a proper sense of taste and smell [10,16]. The recommended dietary allowance (RDA) and recommended upper limit for zinc are 15 and 40 mg day–1, respectively [17]. The consumption of zinc in amounts higher than these values is generally considered relatively non-toxic, and amounts approaching 100 mg day–1 can be tolerated for some time [17]. In the Mg–Zn binary phase diagram [18] there is a deep eutectic point at about 51 wt.% Zn. The existence of a deep eutectic point is the basic factor supporting the high glass-forming ability (GFA) of alloying systems. Indeed, amorphous ternary Mg–Zn–Ca alloys whose compositions are close to the eutectic point have already been prepared [19,20]. It was shown recently that amorphous Mg–Zn-based alloys containing about 50 wt.% Zn are promising candidates for biodegradable implants because they show excellent strength, high corrosion resistance, low hydrogen evolution rate and good biocompatibility within animal tissues [19,21]. However, the preparation of bulk amorphous Mg–Zn-based alloys is difficult because it requires rapid cooling rates. The fact that zinc is a biologically tolerable element, even when its content in Mg-based alloys approaches 50 wt.% [21], indicates that Zn-based alloys may also be promising candidates for biodegradable implants. The advantage of Zn-based alloys over Mg–Zn metallic glasses lies in their much easier preparation. Zinc alloys can be prepared by classical routes such as gravity or die casting, hot rolling or hot extrusion. Another advantage is the lower melting point, lower chemical reactivity and better machinability of zinc, compared with magnesium. Therefore, melting of zinc alloys can be performed in air. To our knowledge there are only a few sources in the scientific literature that mention zinc alloys within the context of biodegradable materials [22]. Detailed studies on their mechanical and corrosion performance in simulated biological environments are lacking. For this reason, in our study we focused our attention on the mechanical and corrosion behavior of Zn-based alloys. Commercial die casting zinc alloys commonly contain aluminum and copper as the main alloying elements. These elements improve the castability and strength of zinc. However, as mentioned before, aluminum is not very suitable for biological applications. Therefore, we have selected a new alloying system, Zn–Mg, in our research because magnesium is assumed to improve the strength and biocompatibility of zinc. The bone healing process may also be improved by the addition of magnesium due to the positive effect of magnesium on bone growth [16]. The Zn–Mg equilibrium phase diagram is shown in Fig. 1. One can see that the solid solubility of magnesium in zinc is low
Fig. 1. Zn-rich region of the Zn–Mg equilibrium phase diagram [18].
and that the Mg2Zn11 intermetallic phase forms in the structure even at small amounts of magnesium. There is a eutectic point in the diagram corresponding to about 3 wt.% Mg. Alloys with higher magnesium concentrations would thus contain primary intermetallic phases that would cause brittleness of such alloys. For this reason, we used magnesium concentrations ranging from 1 to 3 wt.% in the investigated binary Zn–Mg alloys. Pure zinc and commercial ZnAl4Cu1 (in wt.%) alloys were also investigated for comparison. In the latter alloy the addition of aluminum and copper led to mechanical strengthening. In contrast to the zinc alloys mentioned above, magnesium alloys are well documented with respect to their corrosion behavior in simulated body fluids [19]. Therefore, the behavior of the zinc alloys was compared with that of pure magnesium and high purity extruded MgAl9Zn1 (in wt.%, AZ91HP) alloy. These materials serve as standards to assess the mechanical and corrosion performance of new zinc alloys. All of the materials investigated, except for the AZ91HP alloy, were used in the as-cast state. The AZ91HP alloy was used in the as-extruded state, because hot extrusion is well known to refine and homogenize the structure, which results in better corrosion resistance and mechanical strength. Therefore, it is reasonable to compare the new Zn–Mg alloys with the corrosion resistant standard.
2. Experimental In this study Zn, Mg, Zn–Mg, Zn–Al–Cu and Mg–Al–Zn alloys were investigated. The designations and chemical compositions of the alloys studied are given in Table 1. Zn-based alloys were prepared by melting pure Zn (99.95%), Mg (99.90%), Cu (99.90%) and Al (99.50%) in a resistance furnace in air. To prevent excessive evaporation of the volatile zinc the melting temperature did not exceed 500 °C, and homogenization was insured by intense mechanical stirring with a graphite rod. After sufficient homogenization the melts were poured into a cast iron mold to prepare cylindrical ingots 20 mm in diameter and 130 mm in length. Their chemical composition was verified by Xray fluorescence spectrometry, as shown in Table 1. Cylindrical ingots of pure magnesium and AZ91HP alloy of the same dimensions as above were prepared by melting Mg, Al and Zn in an induction furnace under an argon atmosphere and by casting the melts into a cast iron mold. The AZ91HP alloy was then hot extruded at 400 °C and an extrusion ratio of 1:11 to produce a 6 mm rod diameter. The mechanical properties of the prepared alloys were characterized by Vickers hardness measurements using a loading of 5 kg. Tensile tests were also carried out in an Instron 5880 machine at a deformation rate of 1 mm min1 to determine the ultimate tensile strength (UTS), yield strength (YS) and elongation (E) of the alloys. The corrosion behavior was studied in an aerated simulated body fluid (SBF) containing 8 g l1 NaCl, 0.4 g l1 KCl, 0.14 g l1 CaCl2, 0.35 g l1 NaHCO3, 1 g l1 glucose, 0.2 g l1 MgSO47H2O, 0.09 g l1 KH2PO4 and 0.08 g l1 Na2HPO412H2O [23]. This chemical composition indicates that the SBF used has an almost neutral pH value, although real fluids in the human body may have pH values slightly different from 7. For example, it is known that the pH level may decrease to about 5 in the case of an inflammatory reaction [1]. It is also well known that the corrosion rates of both Zn and Mg increase with acidity [24]. For these reasons the corrosion rates were also measured in solutions with pH values of 5 and 10, which were obtained by adding small amounts of HCl and NaOH, respectively, to the SBF. Hereafter these solutions will be denoted as SBF (pH 5) and SBF (pH 10) for simplicity. Both immersion tests and electrochemical potentiodynamic measurements were performed to assess corrosion resistance. In the former coupons 6 mm in diameter and 2 mm in thickness were immersed in
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522
3517
Table 1 Chemical compositions of the investigated Zn- and Mg-based alloys. Alloy designation
Zn ZnMg1 ZnMg1.5 ZnMg3 ZnAl4Cu1 Mg AZ91HP
Element (wt.%) Mg
Al
Si
Ti
Fe
Ni
Cu
Zn
<0.01 0.93 1.55 3.39 <0.01 99.90 90.17
<0.01 <0.01 <0.01 <0.01 4.14 0.04 8.85
0.01 0.01 0.02 0.06 0.02 0.03 0.07
<0.005 <0.005 <0.005 <0.005 <0.005 <0.005 <0.005
<0.004 <0.004 <0.004 <0.004 <0.004 <0.004 <0.004
<0.004 <0.004 <0.004 <0.004 <0.004 <0.004 <0.004
<0.004 <0.004 <0.004 <0.004 0.90 <0.004 <0.004
99.95 99.06 98.43 96.55 94.94 0.01 0.89
200 ml of the SBF for 336 h at 37 °C. Afterward the corrosion products were removed by chemical agents according to ISO 8407. For the Zn-based alloys these products were dissolved in a solution containing 200 g l g l1 CrO3. For the Mg-based alloys a solution of 200 g l1 CrO3, 10 g l1 AgNO3 and 20 g l1 Ba(NO3)2 was used for this purpose. The corrosion rates were then calculated (lm year1) using the weight losses measured on a balance with an accuracy of 0.1 mg, according to ASTM G31-72. Changes in the chemical composition of the SBF caused by corrosion of the alloys were also monitored by pH value measurements. The potentiodynamic curves of the alloys were measured in SBF at 37 °C (Gamry FAS1 potentiostat). Experiments were performed in a standard three electrode set-up: with the sample (a coupon 6 mm in diameter and 2 mm in thickness) as the working electrode, two graphite rods as the counter electrodes and Ag/AgCl/KCl (3 mol l1) as the reference electrode (SSCE) with a potential of 0.199 V/SHE. All potentials presented in this paper were measured against SSCE. Potentiodynamic curves were scanned from –0.2 V/Eocp to +0.8 V/Eocp at a rate of 2 mV s1. The microstructures, phase and chemical compositions and surface morphologies of the alloys both before and after corrosion were examined by light microscopy (LM), scanning electron microscopy (SEM) (Tescan Vega 3), energy dispersion spectrometry (EDS) (Oxford Instruments Inca 350), X-ray diffraction (XRD) (X Pert Pro) and X-ray photoelectron spectroscopy (XPS) (ESCA Probe P, pressure in the analytical chamber 2 108 Pa, monochromatic Al-Ka X-ray source, binding energy calibration with respect to the energy of the Au 4f7/2 peak). 3. Results and discussion 3.1. Structures Optical micrographs of the alloys investigated are shown in Fig. 2. One can observe that the pure zinc contains large grains ranging from about 100 lm to above 1 mm in size (Fig. 2a), which correspond to a relatively low cooling rate during casting. Some grains are almost equiaxed, while others are elongated in a direction parallel to the heat removal during casting. Fig. 2b–d present the Zn–Mg alloys. The ZnMg1 and ZnMg1.5 alloys are hypoeutectic because they consist of primary Zn dendrites (light) and a Zn + Mg2Zn11 eutectic mixture (dark), as was proved by XRD. The average thickness of the dendritic branches in both alloys was approximately 30 lm. The only difference between the ZnMg1 and ZnMg1.5 alloys was in the volume fraction of the eutectic mixture, which increases with the Mg content. Fig. 2d shows a purely eutectic structure for the ZnMg3 alloy, containing a very fine micrometer sized mixture of Zn and Mg2Zn11 phases (see also the phase diagram in Fig. 1). The eutectic mixture creates large colonies that differ in spatial orientation of the intermetallic phase. The sizes of the eutectic colonies range between 50 and 500 lm. Fig. 2e illustrates the structure of the ZnAl4Cu1 alloy. This alloy is dominated by primary dendrites of zinc (light) and a Zn + Al eu-
tectic mixture (dark). The dendritic branches of primary zinc have an average thickness of 20 lm. Due to the relatively low copper concentration in the alloy (Table 1) most of the copper remains dissolved in both zinc and aluminum, as XRD did not indicate the presence of any other phases. This is in accordance with the Al– Cu and Zn–Cu phase diagrams [18], which show that the equilibrium solid solubilities of copper in aluminum and copper in zinc are approximately 6 and 3 wt.%, respectively. The structures of the pure Mg and AZ91HP alloy are shown in Fig. 2f and g. Like pure zinc, pure magnesium consists of large, elongated grains whose thickness is about 300 lm and length exceeds 1 mm. The structure of the AZ91HP alloy is recrystallized and contains equiaxed grains of Mg and fine micrometer sized particles of the Al12Mg17 intermetallic phase. The average Mg grain size is approximately 30 lm. Along the longitudinal section in Fig. 2g the extrusion direction is clearly observed from the arrangement of the intermetallic phases. It is known that as-cast Mg–Al–Zn alloys contain primary dendrites of Mg and a Mg + Al12Mg17 interdendritic eutectic network [25]. However, the hot extrusion process breaks the eutectic network and deforms the Mg dendrites. Dynamic recrystallization then results in the formation of new equiaxed grains of Mg and rows of intermetallic particles, as seen in Fig. 2g. 3.2. Mechanical properties Fig. 3 shows the dependence of the mechanical properties of the Zn–Mg alloys on the magnesium content. The mechanical properties of the other materials investigated are summarized in Table 2. Fig. 3 shows that there is a direct relationship between the magnesium content and the mechanical properties. One can observe that the hardness of the Zn–Mg alloys increases with magnesium concentration from approximately 25 HV for pure Zn up to 200 HV for the ZnMg3 alloy. This behavior can be attributed to the increasing volume fraction of the hard Mg2Zn11 intermetallic phase due to magnesium (see Fig. 2a–d). A slightly different behavior is observed for the UTS of the Zn–Mg alloys, which increases up to approximately 1% of Mg and then decreases at higher magnesium concentrations. The purely eutectic ZnMg3 alloy shows a low strength similar to that of the pure Zn. This low strength can be attributed to the high volume fraction of the brittle eutectic in the ZnMg3 alloy and to the coarse grained structure of both materials (Fig. 2a and d). In both cases the fracture crack growth resistance is low. Due to the presence of the fine primary Zn dendrites and eutectic network in the ZnMg1 and ZnMg1.5 alloys (Fig. 2b and c) the resistance to the fracture increases. Both the fine grains and the hard network represent barriers to the growing crack. The optimum volume fractions of both structural components correspond to the ZnMg1 alloy, which achieves the maximum UTS of approximately 150 MPa. At higher magnesium concentrations the brittle eutectic mixture occupies a larger volume in the alloy and, therefore, the fracture occurs more readily. Accordingly, the YS was measured only for the ZnMg1 alloy as the other alloys fractured before the onset of plastic deformation. This trend was also
3518
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522
Fig. 2. Optical micrographs of the investigated alloys: (a) Zn, (b) ZnMg1, (c) ZnMg1.5, (d) ZnMg3, (e) ZnAl4Cu1, (f) Mg, (g) AZ91HP (longitudinal section).
Table 2 Mechanical properties of the ZnAl4Cu1 and Mg-based alloys (average values were obtained from three measurements).
Fig. 3. Mechanical properties of the Zn–Mg alloys versus Mg content (average values were obtained from three measurements).
confirmed by the E values obtained, which reach their maximum at approximately 1% Mg. The E values of the pure zinc, ZnMg1.5 and ZnMg3 alloys do not exceed 0.5%. Table 2 shows that the ZnAl4Cu1 alloy exhibits a slightly higher UTS and YS and a slightly lower E compared with the ZnMg1 alloy. This is in accordance with a higher volume fraction of the eutectic mixture in the ZnAl4Cu1 alloy (Fig. 2e). The low elongation of all the zinc alloys, which does not exceed 2%, is an intrinsic characteristic of the as-cast state of these
Alloy
Tensile strength (MPa)
Yield strength (MPa)
Elongation (%)
Vickers hardness
ZnAl4Cu1 Mg AZ91HP
210 93 322
171 49 208
1 3 15
80 26 72
alloys. This state is always associated with brittle interdendritic eutectics, relatively coarse as-cast structures and the presence of casting defects, such as porosity (not shown). It is well known that hot extruded or hot rolled zinc alloys achieve significantly higher elongations on the order of tens of percent and higher strength [26]. Table 2 also shows the mechanical properties of the Mg-based materials. As expected, the strength and elongation of the hot extruded AZ91HP alloy is the highest among all of the investigated materials. Hot extrusion eliminates casting defects, refines the microstructure and breaks the eutectic networks (Fig. 2g). All of these modifications improve the strength and plasticity of the material. For the Zn–Mg alloys, particularly the ZnMg1 alloy, a similar improvement would be observed. Comparing the strength and plasticity of magnesium and zinc in Fig. 3 and Table 2, one can see that the former material achieves stronger mechanical properties. As is shown in Fig. 4, which shows the fracture surfaces of the materials, the higher strength and plasticity of magnesium can be attributed to finer grains and to a slight difference in its fracture mechanisms.
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522
3519
Fig. 4. Fracture surfaces of the alloys after tensile tests of (a) Zn, (b, c) Mg, (d) ZnMg1, (e) ZnMg1.5, (f, g) ZnMg3, (h) ZnAl4Cu1 and (i) AZ91HP.
Fig. 4 illustrates the fracture morphologies of the alloys after tensile testing. One can observe that the zinc shows a brittle and intercrystalline fracture (Fig. 4a) without any plastic deformation. The coarse and elongated grains are clearly visible in this figure. Similarly, the magnesium also exhibits an almost brittle and intercrystalline fracture (Fig. 4b), although the fracture morphology is slightly finer than in the case of zinc, and a detailed view (Fig. 4c) reveals that the fracture is accompanied by a slight plastic deformation. This is the reason for the higher strength and plasticity of magnesium compared with zinc. Fig. 4d and e shows the fracture surfaces of the ZnMg1 and ZnMg1.5 alloys, respectively. In contrast to pure zinc, these surfaces have a refined morphology in which both the primary zinc and eutectic mixture (see also Fig. 2b and c) can be clearly distinguished. The primary zinc dendrites are characterized by almost brittle fracture that corresponds to flat facets on the fracture surface. The eutectic mixture that surrounds these facets exhibits a refined morphology, and there is an indication of plastic deformation in this area. A completely different fracture morphology is observed for the eutectic ZnMg3 alloy (Fig. 4f). The fracture is intercrystalline and almost brittle, and large eutectic grains are clearly visible. A detailed view illustrates the presence of very fine eutectic particles on a flat fracture facet (Fig. 4g). The ZnAl4Cu1 alloy presented in Fig. 4h has a similar fracture morphology to that
of the hypoeutectic Zn–Mg alloys and consists of flat facets of the primary zinc and refined areas of the Zn + Al eutectic. A refined fracture morphology with a certain degree of plastic deformation is also observed for the AZ91HP alloy (Fig. 4i). The differences between the fracture morphologies of the alloys investigated agree well with the differences in their mechanical properties. The coarse and intercrystalline fractures (Zn, Mg, and ZnMg3) indicate low strength and plasticity. In contrast, refined fracture surfaces containing some portion of plastic deformation are associated with improved tensile strength. In the Zn–Mg system investigated the alloys containing approximately 1 wt.% Mg achieved the best mechanical performance, as is shown in Fig. 3. When considering the application of these alloys as, for example, fixation screws for fractured bones the mechanical properties of the alloys should be compared with those of bones. Table 3 provides a summary of the basic mechanical properties of bone, the new Zn–Mg alloys, a biodegradable polymer, polylactic acid (PLA), and inert Ti-, Co- and Fe-based alloys that are currently used in bone fixations. One can see that the Zn–Mg alloys are characterized as possessing strengths and moduli much closer to those of bone, compared with the inert biomaterials. Moreover, they also show significantly higher strength than the biodegradable polymer. That the strength could be further improved by hot extrusion has already been discussed.
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522
3520
Table 3 Basic mechanical properties of various biomaterials and natural bones [1]. Material
Density (g cm3)
Tensile strength (MPa)
Elastic modulus (GPa)
Elongation (%)
Bone PLA As-cast Zn–Mg Wrought Tibased Wrought Cobased Wrought stainless steel
2 1 7 4.5
30–280 30 150 800–1100
5–20 2 90 110
1–2 – 1–2 10
8.5
700–1200
220
10
8
600–1000
200
20
PLA, polylactic acid.
3.3. Corrosion behavior The corrosion resistance of the alloys investigated was determined by immersion tests, pH value measurements and potentiodynamic measurements. The corrosion rates in the SBF at three pH values obtained from the immersion tests are presented in Fig. 5a– c. The potentiodynamic curves for the alloys are shown in Fig. 6, and the corrosion parameters obtained from these curves are summarized in Table 4. The following trends can be observed from the figures and tables. (1) The corrosion rates of all of the zinc alloys investigated are significantly lower compared with those of the Mg and AZ91HP alloys. The former alloys corrode at rates of the order of tens of microns per year, while the corrosion rates of the latter are of the order of hundreds of microns per year
Fig. 6. Potentiodynamic curves for the investigated alloys measured in SBF at pH 7.
Table 4 Corrosion parameters of the alloys determined from the potentiodynamic curves in Fig. 6. Alloy
EC (V/SSCE)
j (mA cm2)
Zn ZnMg1 ZnMg1.5 ZnMg3 ZnAl4Cu1 Mg AZ91HP
–0.89 –0.98 –0.93 –0.93 –0.92 –1.64 –1.51
9.7 103 1.2 103 8.8 103 7.4 103 5.2 103 5.3 102 4.8 103
EC, corrosion potential; j, corrosion current density.
Fig. 5. Corrosion rates in SBF at three pH values obtained from the immersion tests: (a) pH 5, (b) pH 7, (c) pH 10.
(Fig. 5). This difference is in accordance with the fact that zinc is more noble in nature than magnesium. The standard potentials of zinc and magnesium are –0.762 and –2.372 V (versus SHE), respectively [24]. The corrosion potentials measured in SBF are higher for the Zn-based alloys than for the Mg-based alloys by about 0.5–0.7 V (Fig. 6 and Table 4). (2) The corrosion rates of all of the alloys increase with decreasing pH (Fig. 5). For the zinc alloys a pH value reduction from 10 to 5 leads to an approximately twofold increase in the corrosion rate. This behavior has been well known for a long time and is schematically expressed by the Pourbaix diagrams of both metals. According to these diagrams magnesium corrodes in acidic, neutral and slightly alkaline environments, irrespective of their oxidation properties. In a strongly alkaline environment a passive magnesium hydroxide (Mg(OH)2) layer forms, which significantly slows down corrosion [24]. The Pourbaix diagram of zinc indicates that it corrodes in acidic and strongly alkaline liquids, while in neutral and slightly alkaline liquids it is passivated [24]. (3) The influence of Mg on the corrosion rate and corrosion parameters of the Zn–Mg alloys is small (Fig. 5 and Table 4). This could be explained by the low Mg concentrations used in this study. Moreover, magnesium forms the Mg2Zn11 phase in the Zn–Mg alloys (Fig. 2b–d), and it can be assumed that the standard potential of this phase is not very different from that of pure zinc. Therefore, the formation of galvanic micro-cells between Zn and Mg2Zn11, which would accelerate corrosion and shift the corrosion parameters, is unlikely.
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522 Table 5 Surface chemical compositions of the alloys corroded in SBF at pH 7 for 336 h (EDS).a Alloy
Element (wt.%)
Zn ZnMg1 ZnMg1.5 ZnMg3 ZnAl4Cu1 Mg AZ91HP a
Mg
Zn
5.0 3.0 5.8 6.5 5.2 47.8 54.9
64.5 86.0 37.0 63.6 33.8
Al
Na
1.1 0.6
0.6
10.3
Ca
P
14.2 4.2 31.0 14.1 32.3 26.8 14.6
16.3 6.7 26.2 15.9 27.1 24.9 19.6
Cu
0.5
C, H and O are not included.
Fig. 7. Changes in pH value during exposure of the alloys to SBF.
It should be noted that the corrosion rates of the zinc alloys investigated in this study are considerably lower compared with the highly corrosion-resistant Mg-based biodegradable alloys. As reported recently by Gu et al. [19], these corrosion-resistant alloys are based on pure Mg, Mg–Zn, Mg–rare earth or Mg–Zn–Ca bulk metallic glasses. In SBF with a pH of about 7 all of these alloys show corrosion rates of approximately 200 lm year1 or higher, which is at least a sixfold increase relative to the Zn–Mg alloys (Fig. 5b). Fig. 7 shows the evolution of the pH value during immersion of the alloys in SBF pH 7. It is evident that the pH value progressively increases as corrosion proceeds. The observed pH value change is attributable to the base corrosion reactions that produce OH anions:
Mg þ 2H2 O ! Mg2þ þ H2 þ 2OH
ð1Þ
Zn þ 2H2 O ! Zn2þ þ H2 þ 2OH
ð2Þ
3521
As can be observed in Fig. 7, the reaction of the Zn and ZnMg1 alloys is significantly slower, as it only causes the pH value to increase to approximately 7.4–7.5 after 90 h immersion. For the Mg and AZ91HP alloys the pH value increases to about 7.8 due to the corrosion reaction. The difference between the final pH values is seemingly small. However, it should be noted that this small difference indicates a relatively large difference in the numbers of OH– anions produced by reactions (1) and (2). A simple calculation reveals that the concentrations of OH– anions after 90 h immersion of Mg and Zn are 6 107 and 2.5 107 mol l1, respectively. Taking the stoichiometry of the reactions expressed by Eqs. (1) and (2), one can estimate that there is a fivefold difference between Zn and Mg in terms of their volume reduction due to corrosion. This difference agrees well with the corrosion rates shown in Fig. 5b, in which the corrosion rates of Mg and Zn are shown to be approximately 220 and 50 lm year1, respectively. The surfaces of the alloys exposed to SBF at pH 7 for 336 h are similar and they are illustrated in Fig. 8. The surfaces of all of the alloys are covered by relatively compact layers of corrosion products. Due to internal stress these layers are cracked and locally detached from the surface in some areas. The phase compositions of the corrosion products were analyzed by XRD. However, this method revealed only those phases present in the original alloys (Fig. 2) and no new phases were accurately determined after corrosion. This suggests that the layers of corrosion products that remain adherent on the surface are very thin. The chemical compositions of the corroded surfaces measured by EDS and XPS are summarized in Tables 5 and 6, respectively. The difference between the two methods is the depth to which each method is capable of analyzing material. EDS analyzes an approximately 1 lm thick surface layer, while the XPS signal is acquired from a surface layer with a thickness of only several interatomic distances. Besides the basic components of the alloys, one can observe in Table 5 that the corrosion products mainly contain phosphorus and calcium, both originating from the SBF. It should be noted that light elements like O, C and H are not included in this analysis. XPS analysis (Table 6) indicates that the Ca2+, Mg2+, Zn2+ and P5+ oxidation states predominate on the surface. Therefore, it can be assumed that the surfaces of the corroded zinc and magnesium alloys are dominated by insoluble calcium–magnesium phosphates. In the zinc alloys the phosphate contains a low concentration of zinc. Carbonates may also exist on the corroded surfaces, although their presence cannot be proved by EDS and XPS analyses. The formation of insoluble phosphates can be attributed to the presence of hydrogen phosphate and dihydrogen phosphate in the SBF. As indicated by Eqs. (1) and (2) and by Fig. 7, the corrosion reaction is accompanied
Fig. 8. The surfaces of the alloys exposed to the SBF at pH 7 for 336 h: (a) AZ91HP, (b) ZnMg1.5.
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522
3522
Table 6 Surface chemical compositions of two selected alloys corroded in SBF at pH 7 for 336 h (XPS).a Alloy
AZ91HP ZnMg1
Element (wt.%) (oxidation state)
of the Czech Republic (Projects Nos. MSM6046137302 and MSMT 21/2010). Appendix A. Figures with essential color discrimination
O
Ca
P
Mg
Zn
37.5 38.7
25.2 (Ca2+) 27.8 (Ca2+)
25.4 (P5+) 26.7 (P5+)
11.9 (Mg2+) 4.8 (Mg2+)
– 1.9 (Zn2+)
a C and H are not included. The high oxygen concentrations are caused by adsorption on the surface.
Certain figures in this article, particularly Figs. 5–7, are difficult to interpret in black and white. The full color images can be found in the on-line version, at doi:10.1016/j.actbio.2011.05.008. References
by alkalization of the SBF. As a result, insoluble phosphates are formed by the following simplified reaction: 6Mg2þ ðCa2þ Þ þ 2HPO2 4 þ 2H2 PO4 þ 6OH
! 2ðMg; CaÞ3 ðPO4 Þ2 ðinsolubleÞ þ 6H2 O
ð3Þ
Due to the presence of chloride anions in the SBF the insoluble corrosion products may be partially transformed to soluble chlorides, which accelerates corrosion. Chlorides were not detected in the insoluble corrosion products, either by EDS or by XPS, as is indicated in Tables 5 and 6. From a biocompatibility point of view the possible toxicity of zinc alloys should be taken into account. Compared with magnesium, which has also been considered for biodegradable implants, the maximum acceptable doses of zinc are lower, and overdoses may cause medical problems. It is, therefore, important to estimate the possible intake of zinc from the corrosion of a biodegradable implant. Let us consider a typical screw 5 mm in diameter and 50 mm in length made of zinc for the fixation of a fractured cortical bone. As indicated in Fig. 5a, the corrosion rate of this screw in a slightly acidic environment could be about 80 lm year1. A simple calculation reveals that the average daily intake of zinc from one screw is approximately 1.5 mg Zn. This value is negligible compared with the RDA and recommended upper limit for zinc, which are 15 and 40 mg day1, respectively [17]. 4. Conclusions New Zn–Mg alloys containing approximately 1 wt.% Mg are proposed for use in biodegradable medical implants, such as bone fixations. These alloys can be considered alternatives to Mg-based biodegradable alloys. In the as-cast state Zn–Mg alloys have a relatively fine structure composed of primary zinc and an interdendritic eutectic mixture. Such a structure results in good mechanical strength, comparable with that of bone, which is significantly higher than that of biodegradable polymeric materials. It is assumed that additional improvements in the strength of Zn–Mg alloys can be achieved by the use of hot extrusion. The main advantage of zinc alloys over magnesium alloys lies in their significantly better corrosion resistance in SBF. The low corrosion rate results in low rates of pH value increase and hydrogen evolution. Both factors are very important to prevent negative reactions in tissue during healing. In addition, the possible doses of zinc ions released from implants are significantly below zinc tolerable limit. Acknowledgments The research on biodegradable metallic materials was financially supported by the Ministry of Education, Youth and Sports
[1] Davis JR. Handbook of materials for medical devices. Materials Park, OH: ASM International; 2003. [2] Witte F. The history of biodegradable magnesium implants. Acta Biomater 2010;6:1680–92. [3] Zhang ChY, Zeng RCh, Liu ChL, Gao JCh. Comparison of calcium phosphate coatings on Mg–Al and Mg–Ca alloys and their corrosion behavior in Hank’s solution. Surf Coat Technol 2010;204:3636–40. [4] Perada MD, Alonso C, Burgos-Asperilla L, del Valle JA, Ruano OA, Perez P, et al. Corrosion inhibition of powder metallurgy Mg by fluoride treatments. Acta Biomater 2010;6:1772–82. [5] Gray-Munro JE, Seguin Ch, Strong M. Influence of surface modification on the in vitro corrosion rate of magnesium alloy AZ31. J Biomed Mater Res Part A 2009;91:221–30. [6] Hiromoto S, Shishido T, Yamamoto A, Maruyama N, Somekawa H, Mukai T. Precipitation control of calcium phosphate on pure magnesium by anodization. Corrosion Sci 2008;50:2906–13. [7] Erbel L et al. Temporary scaffolding of coronary arteries with bioabsorbable magnesium stents: a prospective, non-randomised multicentre trial. Lancet 2007;369:1869–75. [8] Di Mario C et al. Drug-eluting bioabsorbable magnesium stent. J Interven Cardiol 2004;17:391–5. [9] Peeters P, Bosiers M, Verbist J, Deloose K, Heublein B. Preliminary results after application of absorbable metal stents in patients with critical limb ischemia. J Endovasc Ther 2005;12:1–5. [10] Zhang S. At al. Research of Mg–Zn alloy as a degradable biomaterial. Acta Biomater 2010;6:626–40. [11] Li Z, Gu X, Lou S, Zheng Y. The development of binary Mg–Ca alloys for use as biodegradable materials within bone. Biomaterials 2008;29:1329–44. [12] Zhang E, Yang L, Xu J, Chen H. Microstructure, mechanical properties and biocorrosion properties of Mg–Si(–Ca, Zn) alloy for biomedical application. Acta Biomater 2010;6:1756–62. [13] Zhang E, Yang L. Microstructure, mechanical properties and bio-corrosion properties of Mg–Zn–Mn–Ca alloy for biomedical application. Mater Sci Eng A 2008;497:111–8. [14] Gu XN, Zheng W, Cheng Y, Zheng YF. A study on alkaline heat treated Mg–Ca alloy for the control of the biocorrosion rate. Acta Biomater 2009;5:2790–9. [15] Huan ZG, Leeflang MA, Zhou J, Fratila-Apachitei LE, Duszczyk J. In vitro degradation behaviour and cytocompatibility of Mg–Zn–Zr alloys. J Mater Sci Mater Med 2010;21:2623–35. [16] Witte F, Hort N, Vogt C, Cohen S, Kainer KU, Willumeit R, et al. Degradable biomaterials based on magnesium corrosion. Curr Opin Solid State Mater Sci 2008;12:63–72. [17] Fosmire GJ. Zinc toxicity. Am J Clin Nutr 1990;51:225–7. [18] Gale WF, Totemeier TC. Smithells metals reference book. 8th ed. Oxford: Elsevier; 2004. [19] Gu X, Zheng Y, Zhong S, Xi T, Wang J, Wang W. Corrosion of, and cellular responses to Mg–Zn–Ca bulk metallic glasses. Biomaterials 2010;31:1093–103. [20] Li QF, Weng HR, Suo ZY, Ren YL, Yuan XG, Qiu KQ. Microstructure and mechanical properties of bulk Mg–Zn–Ca amorphous alloys and amorphous matrix composites. Mater Sci Eng A 2008;487:301–8. [21] Zberg B, Uggowitzer PJ, Loffler JF. MgZnCa glasses without clinically observable hydrogen evolution for biodegradable implants. Nat Mater 2009;8:887–91. [22] Wang X, Lu H, Li X, Li L, Zheng Y. Effect of cooling rate and composition on microstructures and properties of Zn–Mg alloys. Trans Nonferrous Met Soc China 2007;17:122–5. [23] Wang H, Shi ZM, Yang K. Magnesium and magnesium alloys as degradable metallic biomaterials. Adv Mater Res 2008;32:207–10. [24] Talbot DEJ, Talbot JDR. Corrosion science and technology. London: CRC Press; 2007. ˇ ízˇová H, Volenec K. Investigation of magnesium-based alloys for [25] Vojteˇch D, C biomedical applications. Kovove Mater 2006;44:211–23. [26] Davis JR, editor. Properties and selection: nonferrous alloys and specialpurpose materials. ASM handbook, vol. 2. Materials Park, OH: ASM International; 1990.
Contents lists available at ScienceDirect
Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat
Mechanical and corrosion properties of newly developed biodegradable Zn-based alloys for bone fixation D. Vojteˇch ⇑, J. Kubásek, J. Šerák, P. Novák Department of Metals and Corrosion Engineering, Institute of Chemical Technology, Technická 5, 166 28 Prague 6, Czech Republic
a r t i c l e
i n f o
Article history: Received 12 January 2011 Received in revised form 20 April 2011 Accepted 9 May 2011 Available online 14 May 2011 Keywords: Biodegradable material Zinc Bone fixation Mechanical properties Corrosion
a b s t r a c t In the present work Zn–Mg alloys containing up to 3 wt.% Mg were studied as potential biodegradable materials for medical use. The structure, mechanical properties and corrosion behavior of these alloys were investigated and compared with those of pure Mg, AZ91HP and casting Zn–Al–Cu alloys. The structures were examined by light and scanning electron microscopy (SEM), and tensile and hardness testing were used to characterize the mechanical properties of the alloys. The corrosion behavior of the materials in simulated body fluid with pH values of 5, 7 and 10 was determined by immersion tests, potentiodynamic measurements and by monitoring the pH value evolution during corrosion. The surfaces of the corroded alloys were investigated by SEM, energy-dispersive spectrometry and X-ray photoelectron spectroscopy. It was found that a maximum strength and elongation of 150 MPa and 2%, respectively, were achieved at Mg contents of approximately 1 wt.%. These mechanical properties are discussed in relation to the structural features of the alloys. The corrosion rates of the Zn–Mg alloys were determined to be significantly lower than those of Mg and AZ91HP alloys. The former alloys corroded at rates of the order of tens of microns per year, whereas the corrosion rates of the latter were of the order of hundreds of microns per year. Possible zinc doses and toxicity were estimated from the corrosion behavior of the zinc alloys. It was found that these doses are negligible compared with the tolerable biological daily limit of zinc. Ó 2011 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
1. Introduction Biodegradable materials are capable of progressively degrading in the human body to produce non-toxic compounds that can be readily excreted. Biodegradable polymeric materials have been known and used for a long time as, for example, fibers for the reparation of damaged tissue or for winding biodegradable stents. However, for load-bearing applications, such as screws, plates or other fixations of fractured bones, polymeric materials are not suitable due to their low mechanical strength [1]. Among biodegradable metallic materials magnesium has attracted the greatest interest, and its properties related to bioapplications have been studied since the beginning of the 20th century [2]. The reason for this is that magnesium is non-toxic to the human body and excessive amounts of it can be readily excreted by the kidneys. Magnesium is also very important for biological functions of the human body. The main disadvantage of most magnesium alloys is that they corrode too rapidly in physiological environments. The corrosion of magnesium produces hydrogen pockets near the implant, which retard the healing ⇑ Corresponding author. Tel.: +420 220444290; fax: +420 220444400. E-mail address: [email protected] (D. Vojteˇch).
process. In addition, the local increase in alkalinity close to the magnesium implant also has adverse effects on healing [2–6]. Therefore, great effort has been exerted during the last 20 years to find magnesium-based alloys that corrode at acceptably low rates and whose corrosion products, like hydrogen and alkaline materials, can be absorbed by the surrounding tissue without negative effects on the healing process. The magnesium alloys initially considered for bio-applications were based on alloying systems originally developed for engineering applications. Among them AZ, LAE and WE type alloys have been widely studied, but only the WE43 alloy has been used to prepare a biodegradable implant, in this case a vascular stent, used in the human body [7–9]. Besides engineering alloys, some new alloying systems have been developed to provide good corrosion resistance and biocompatibility. Among them Mg–Zn, Mg–Zn–Mn–Ca, Mg–Zn–Y, Mg–Gd, Mg–Zn–Si and other systems have recently been studied [10–15]. In these systems zinc is often the major constituent. Zinc, as a more noble metal than magnesium, is well known to positively affect the corrosion resistance and strength of Mg. From a biological point of view zinc is very important for biological functions in the human body because it is involved in various aspects of cellular metabolism. Zinc is important to the proper function of numerous enzymes and it supports immune
1742-7061/$ - see front matter Ó 2011 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.actbio.2011.05.008
3516
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522
functions, protein and DNA synthesis and wound healing. It also supports normal growth and a proper sense of taste and smell [10,16]. The recommended dietary allowance (RDA) and recommended upper limit for zinc are 15 and 40 mg day–1, respectively [17]. The consumption of zinc in amounts higher than these values is generally considered relatively non-toxic, and amounts approaching 100 mg day–1 can be tolerated for some time [17]. In the Mg–Zn binary phase diagram [18] there is a deep eutectic point at about 51 wt.% Zn. The existence of a deep eutectic point is the basic factor supporting the high glass-forming ability (GFA) of alloying systems. Indeed, amorphous ternary Mg–Zn–Ca alloys whose compositions are close to the eutectic point have already been prepared [19,20]. It was shown recently that amorphous Mg–Zn-based alloys containing about 50 wt.% Zn are promising candidates for biodegradable implants because they show excellent strength, high corrosion resistance, low hydrogen evolution rate and good biocompatibility within animal tissues [19,21]. However, the preparation of bulk amorphous Mg–Zn-based alloys is difficult because it requires rapid cooling rates. The fact that zinc is a biologically tolerable element, even when its content in Mg-based alloys approaches 50 wt.% [21], indicates that Zn-based alloys may also be promising candidates for biodegradable implants. The advantage of Zn-based alloys over Mg–Zn metallic glasses lies in their much easier preparation. Zinc alloys can be prepared by classical routes such as gravity or die casting, hot rolling or hot extrusion. Another advantage is the lower melting point, lower chemical reactivity and better machinability of zinc, compared with magnesium. Therefore, melting of zinc alloys can be performed in air. To our knowledge there are only a few sources in the scientific literature that mention zinc alloys within the context of biodegradable materials [22]. Detailed studies on their mechanical and corrosion performance in simulated biological environments are lacking. For this reason, in our study we focused our attention on the mechanical and corrosion behavior of Zn-based alloys. Commercial die casting zinc alloys commonly contain aluminum and copper as the main alloying elements. These elements improve the castability and strength of zinc. However, as mentioned before, aluminum is not very suitable for biological applications. Therefore, we have selected a new alloying system, Zn–Mg, in our research because magnesium is assumed to improve the strength and biocompatibility of zinc. The bone healing process may also be improved by the addition of magnesium due to the positive effect of magnesium on bone growth [16]. The Zn–Mg equilibrium phase diagram is shown in Fig. 1. One can see that the solid solubility of magnesium in zinc is low
Fig. 1. Zn-rich region of the Zn–Mg equilibrium phase diagram [18].
and that the Mg2Zn11 intermetallic phase forms in the structure even at small amounts of magnesium. There is a eutectic point in the diagram corresponding to about 3 wt.% Mg. Alloys with higher magnesium concentrations would thus contain primary intermetallic phases that would cause brittleness of such alloys. For this reason, we used magnesium concentrations ranging from 1 to 3 wt.% in the investigated binary Zn–Mg alloys. Pure zinc and commercial ZnAl4Cu1 (in wt.%) alloys were also investigated for comparison. In the latter alloy the addition of aluminum and copper led to mechanical strengthening. In contrast to the zinc alloys mentioned above, magnesium alloys are well documented with respect to their corrosion behavior in simulated body fluids [19]. Therefore, the behavior of the zinc alloys was compared with that of pure magnesium and high purity extruded MgAl9Zn1 (in wt.%, AZ91HP) alloy. These materials serve as standards to assess the mechanical and corrosion performance of new zinc alloys. All of the materials investigated, except for the AZ91HP alloy, were used in the as-cast state. The AZ91HP alloy was used in the as-extruded state, because hot extrusion is well known to refine and homogenize the structure, which results in better corrosion resistance and mechanical strength. Therefore, it is reasonable to compare the new Zn–Mg alloys with the corrosion resistant standard.
2. Experimental In this study Zn, Mg, Zn–Mg, Zn–Al–Cu and Mg–Al–Zn alloys were investigated. The designations and chemical compositions of the alloys studied are given in Table 1. Zn-based alloys were prepared by melting pure Zn (99.95%), Mg (99.90%), Cu (99.90%) and Al (99.50%) in a resistance furnace in air. To prevent excessive evaporation of the volatile zinc the melting temperature did not exceed 500 °C, and homogenization was insured by intense mechanical stirring with a graphite rod. After sufficient homogenization the melts were poured into a cast iron mold to prepare cylindrical ingots 20 mm in diameter and 130 mm in length. Their chemical composition was verified by Xray fluorescence spectrometry, as shown in Table 1. Cylindrical ingots of pure magnesium and AZ91HP alloy of the same dimensions as above were prepared by melting Mg, Al and Zn in an induction furnace under an argon atmosphere and by casting the melts into a cast iron mold. The AZ91HP alloy was then hot extruded at 400 °C and an extrusion ratio of 1:11 to produce a 6 mm rod diameter. The mechanical properties of the prepared alloys were characterized by Vickers hardness measurements using a loading of 5 kg. Tensile tests were also carried out in an Instron 5880 machine at a deformation rate of 1 mm min1 to determine the ultimate tensile strength (UTS), yield strength (YS) and elongation (E) of the alloys. The corrosion behavior was studied in an aerated simulated body fluid (SBF) containing 8 g l1 NaCl, 0.4 g l1 KCl, 0.14 g l1 CaCl2, 0.35 g l1 NaHCO3, 1 g l1 glucose, 0.2 g l1 MgSO47H2O, 0.09 g l1 KH2PO4 and 0.08 g l1 Na2HPO412H2O [23]. This chemical composition indicates that the SBF used has an almost neutral pH value, although real fluids in the human body may have pH values slightly different from 7. For example, it is known that the pH level may decrease to about 5 in the case of an inflammatory reaction [1]. It is also well known that the corrosion rates of both Zn and Mg increase with acidity [24]. For these reasons the corrosion rates were also measured in solutions with pH values of 5 and 10, which were obtained by adding small amounts of HCl and NaOH, respectively, to the SBF. Hereafter these solutions will be denoted as SBF (pH 5) and SBF (pH 10) for simplicity. Both immersion tests and electrochemical potentiodynamic measurements were performed to assess corrosion resistance. In the former coupons 6 mm in diameter and 2 mm in thickness were immersed in
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522
3517
Table 1 Chemical compositions of the investigated Zn- and Mg-based alloys. Alloy designation
Zn ZnMg1 ZnMg1.5 ZnMg3 ZnAl4Cu1 Mg AZ91HP
Element (wt.%) Mg
Al
Si
Ti
Fe
Ni
Cu
Zn
<0.01 0.93 1.55 3.39 <0.01 99.90 90.17
<0.01 <0.01 <0.01 <0.01 4.14 0.04 8.85
0.01 0.01 0.02 0.06 0.02 0.03 0.07
<0.005 <0.005 <0.005 <0.005 <0.005 <0.005 <0.005
<0.004 <0.004 <0.004 <0.004 <0.004 <0.004 <0.004
<0.004 <0.004 <0.004 <0.004 <0.004 <0.004 <0.004
<0.004 <0.004 <0.004 <0.004 0.90 <0.004 <0.004
99.95 99.06 98.43 96.55 94.94 0.01 0.89
200 ml of the SBF for 336 h at 37 °C. Afterward the corrosion products were removed by chemical agents according to ISO 8407. For the Zn-based alloys these products were dissolved in a solution containing 200 g l g l1 CrO3. For the Mg-based alloys a solution of 200 g l1 CrO3, 10 g l1 AgNO3 and 20 g l1 Ba(NO3)2 was used for this purpose. The corrosion rates were then calculated (lm year1) using the weight losses measured on a balance with an accuracy of 0.1 mg, according to ASTM G31-72. Changes in the chemical composition of the SBF caused by corrosion of the alloys were also monitored by pH value measurements. The potentiodynamic curves of the alloys were measured in SBF at 37 °C (Gamry FAS1 potentiostat). Experiments were performed in a standard three electrode set-up: with the sample (a coupon 6 mm in diameter and 2 mm in thickness) as the working electrode, two graphite rods as the counter electrodes and Ag/AgCl/KCl (3 mol l1) as the reference electrode (SSCE) with a potential of 0.199 V/SHE. All potentials presented in this paper were measured against SSCE. Potentiodynamic curves were scanned from –0.2 V/Eocp to +0.8 V/Eocp at a rate of 2 mV s1. The microstructures, phase and chemical compositions and surface morphologies of the alloys both before and after corrosion were examined by light microscopy (LM), scanning electron microscopy (SEM) (Tescan Vega 3), energy dispersion spectrometry (EDS) (Oxford Instruments Inca 350), X-ray diffraction (XRD) (X Pert Pro) and X-ray photoelectron spectroscopy (XPS) (ESCA Probe P, pressure in the analytical chamber 2 108 Pa, monochromatic Al-Ka X-ray source, binding energy calibration with respect to the energy of the Au 4f7/2 peak). 3. Results and discussion 3.1. Structures Optical micrographs of the alloys investigated are shown in Fig. 2. One can observe that the pure zinc contains large grains ranging from about 100 lm to above 1 mm in size (Fig. 2a), which correspond to a relatively low cooling rate during casting. Some grains are almost equiaxed, while others are elongated in a direction parallel to the heat removal during casting. Fig. 2b–d present the Zn–Mg alloys. The ZnMg1 and ZnMg1.5 alloys are hypoeutectic because they consist of primary Zn dendrites (light) and a Zn + Mg2Zn11 eutectic mixture (dark), as was proved by XRD. The average thickness of the dendritic branches in both alloys was approximately 30 lm. The only difference between the ZnMg1 and ZnMg1.5 alloys was in the volume fraction of the eutectic mixture, which increases with the Mg content. Fig. 2d shows a purely eutectic structure for the ZnMg3 alloy, containing a very fine micrometer sized mixture of Zn and Mg2Zn11 phases (see also the phase diagram in Fig. 1). The eutectic mixture creates large colonies that differ in spatial orientation of the intermetallic phase. The sizes of the eutectic colonies range between 50 and 500 lm. Fig. 2e illustrates the structure of the ZnAl4Cu1 alloy. This alloy is dominated by primary dendrites of zinc (light) and a Zn + Al eu-
tectic mixture (dark). The dendritic branches of primary zinc have an average thickness of 20 lm. Due to the relatively low copper concentration in the alloy (Table 1) most of the copper remains dissolved in both zinc and aluminum, as XRD did not indicate the presence of any other phases. This is in accordance with the Al– Cu and Zn–Cu phase diagrams [18], which show that the equilibrium solid solubilities of copper in aluminum and copper in zinc are approximately 6 and 3 wt.%, respectively. The structures of the pure Mg and AZ91HP alloy are shown in Fig. 2f and g. Like pure zinc, pure magnesium consists of large, elongated grains whose thickness is about 300 lm and length exceeds 1 mm. The structure of the AZ91HP alloy is recrystallized and contains equiaxed grains of Mg and fine micrometer sized particles of the Al12Mg17 intermetallic phase. The average Mg grain size is approximately 30 lm. Along the longitudinal section in Fig. 2g the extrusion direction is clearly observed from the arrangement of the intermetallic phases. It is known that as-cast Mg–Al–Zn alloys contain primary dendrites of Mg and a Mg + Al12Mg17 interdendritic eutectic network [25]. However, the hot extrusion process breaks the eutectic network and deforms the Mg dendrites. Dynamic recrystallization then results in the formation of new equiaxed grains of Mg and rows of intermetallic particles, as seen in Fig. 2g. 3.2. Mechanical properties Fig. 3 shows the dependence of the mechanical properties of the Zn–Mg alloys on the magnesium content. The mechanical properties of the other materials investigated are summarized in Table 2. Fig. 3 shows that there is a direct relationship between the magnesium content and the mechanical properties. One can observe that the hardness of the Zn–Mg alloys increases with magnesium concentration from approximately 25 HV for pure Zn up to 200 HV for the ZnMg3 alloy. This behavior can be attributed to the increasing volume fraction of the hard Mg2Zn11 intermetallic phase due to magnesium (see Fig. 2a–d). A slightly different behavior is observed for the UTS of the Zn–Mg alloys, which increases up to approximately 1% of Mg and then decreases at higher magnesium concentrations. The purely eutectic ZnMg3 alloy shows a low strength similar to that of the pure Zn. This low strength can be attributed to the high volume fraction of the brittle eutectic in the ZnMg3 alloy and to the coarse grained structure of both materials (Fig. 2a and d). In both cases the fracture crack growth resistance is low. Due to the presence of the fine primary Zn dendrites and eutectic network in the ZnMg1 and ZnMg1.5 alloys (Fig. 2b and c) the resistance to the fracture increases. Both the fine grains and the hard network represent barriers to the growing crack. The optimum volume fractions of both structural components correspond to the ZnMg1 alloy, which achieves the maximum UTS of approximately 150 MPa. At higher magnesium concentrations the brittle eutectic mixture occupies a larger volume in the alloy and, therefore, the fracture occurs more readily. Accordingly, the YS was measured only for the ZnMg1 alloy as the other alloys fractured before the onset of plastic deformation. This trend was also
3518
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522
Fig. 2. Optical micrographs of the investigated alloys: (a) Zn, (b) ZnMg1, (c) ZnMg1.5, (d) ZnMg3, (e) ZnAl4Cu1, (f) Mg, (g) AZ91HP (longitudinal section).
Table 2 Mechanical properties of the ZnAl4Cu1 and Mg-based alloys (average values were obtained from three measurements).
Fig. 3. Mechanical properties of the Zn–Mg alloys versus Mg content (average values were obtained from three measurements).
confirmed by the E values obtained, which reach their maximum at approximately 1% Mg. The E values of the pure zinc, ZnMg1.5 and ZnMg3 alloys do not exceed 0.5%. Table 2 shows that the ZnAl4Cu1 alloy exhibits a slightly higher UTS and YS and a slightly lower E compared with the ZnMg1 alloy. This is in accordance with a higher volume fraction of the eutectic mixture in the ZnAl4Cu1 alloy (Fig. 2e). The low elongation of all the zinc alloys, which does not exceed 2%, is an intrinsic characteristic of the as-cast state of these
Alloy
Tensile strength (MPa)
Yield strength (MPa)
Elongation (%)
Vickers hardness
ZnAl4Cu1 Mg AZ91HP
210 93 322
171 49 208
1 3 15
80 26 72
alloys. This state is always associated with brittle interdendritic eutectics, relatively coarse as-cast structures and the presence of casting defects, such as porosity (not shown). It is well known that hot extruded or hot rolled zinc alloys achieve significantly higher elongations on the order of tens of percent and higher strength [26]. Table 2 also shows the mechanical properties of the Mg-based materials. As expected, the strength and elongation of the hot extruded AZ91HP alloy is the highest among all of the investigated materials. Hot extrusion eliminates casting defects, refines the microstructure and breaks the eutectic networks (Fig. 2g). All of these modifications improve the strength and plasticity of the material. For the Zn–Mg alloys, particularly the ZnMg1 alloy, a similar improvement would be observed. Comparing the strength and plasticity of magnesium and zinc in Fig. 3 and Table 2, one can see that the former material achieves stronger mechanical properties. As is shown in Fig. 4, which shows the fracture surfaces of the materials, the higher strength and plasticity of magnesium can be attributed to finer grains and to a slight difference in its fracture mechanisms.
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522
3519
Fig. 4. Fracture surfaces of the alloys after tensile tests of (a) Zn, (b, c) Mg, (d) ZnMg1, (e) ZnMg1.5, (f, g) ZnMg3, (h) ZnAl4Cu1 and (i) AZ91HP.
Fig. 4 illustrates the fracture morphologies of the alloys after tensile testing. One can observe that the zinc shows a brittle and intercrystalline fracture (Fig. 4a) without any plastic deformation. The coarse and elongated grains are clearly visible in this figure. Similarly, the magnesium also exhibits an almost brittle and intercrystalline fracture (Fig. 4b), although the fracture morphology is slightly finer than in the case of zinc, and a detailed view (Fig. 4c) reveals that the fracture is accompanied by a slight plastic deformation. This is the reason for the higher strength and plasticity of magnesium compared with zinc. Fig. 4d and e shows the fracture surfaces of the ZnMg1 and ZnMg1.5 alloys, respectively. In contrast to pure zinc, these surfaces have a refined morphology in which both the primary zinc and eutectic mixture (see also Fig. 2b and c) can be clearly distinguished. The primary zinc dendrites are characterized by almost brittle fracture that corresponds to flat facets on the fracture surface. The eutectic mixture that surrounds these facets exhibits a refined morphology, and there is an indication of plastic deformation in this area. A completely different fracture morphology is observed for the eutectic ZnMg3 alloy (Fig. 4f). The fracture is intercrystalline and almost brittle, and large eutectic grains are clearly visible. A detailed view illustrates the presence of very fine eutectic particles on a flat fracture facet (Fig. 4g). The ZnAl4Cu1 alloy presented in Fig. 4h has a similar fracture morphology to that
of the hypoeutectic Zn–Mg alloys and consists of flat facets of the primary zinc and refined areas of the Zn + Al eutectic. A refined fracture morphology with a certain degree of plastic deformation is also observed for the AZ91HP alloy (Fig. 4i). The differences between the fracture morphologies of the alloys investigated agree well with the differences in their mechanical properties. The coarse and intercrystalline fractures (Zn, Mg, and ZnMg3) indicate low strength and plasticity. In contrast, refined fracture surfaces containing some portion of plastic deformation are associated with improved tensile strength. In the Zn–Mg system investigated the alloys containing approximately 1 wt.% Mg achieved the best mechanical performance, as is shown in Fig. 3. When considering the application of these alloys as, for example, fixation screws for fractured bones the mechanical properties of the alloys should be compared with those of bones. Table 3 provides a summary of the basic mechanical properties of bone, the new Zn–Mg alloys, a biodegradable polymer, polylactic acid (PLA), and inert Ti-, Co- and Fe-based alloys that are currently used in bone fixations. One can see that the Zn–Mg alloys are characterized as possessing strengths and moduli much closer to those of bone, compared with the inert biomaterials. Moreover, they also show significantly higher strength than the biodegradable polymer. That the strength could be further improved by hot extrusion has already been discussed.
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522
3520
Table 3 Basic mechanical properties of various biomaterials and natural bones [1]. Material
Density (g cm3)
Tensile strength (MPa)
Elastic modulus (GPa)
Elongation (%)
Bone PLA As-cast Zn–Mg Wrought Tibased Wrought Cobased Wrought stainless steel
2 1 7 4.5
30–280 30 150 800–1100
5–20 2 90 110
1–2 – 1–2 10
8.5
700–1200
220
10
8
600–1000
200
20
PLA, polylactic acid.
3.3. Corrosion behavior The corrosion resistance of the alloys investigated was determined by immersion tests, pH value measurements and potentiodynamic measurements. The corrosion rates in the SBF at three pH values obtained from the immersion tests are presented in Fig. 5a– c. The potentiodynamic curves for the alloys are shown in Fig. 6, and the corrosion parameters obtained from these curves are summarized in Table 4. The following trends can be observed from the figures and tables. (1) The corrosion rates of all of the zinc alloys investigated are significantly lower compared with those of the Mg and AZ91HP alloys. The former alloys corrode at rates of the order of tens of microns per year, while the corrosion rates of the latter are of the order of hundreds of microns per year
Fig. 6. Potentiodynamic curves for the investigated alloys measured in SBF at pH 7.
Table 4 Corrosion parameters of the alloys determined from the potentiodynamic curves in Fig. 6. Alloy
EC (V/SSCE)
j (mA cm2)
Zn ZnMg1 ZnMg1.5 ZnMg3 ZnAl4Cu1 Mg AZ91HP
–0.89 –0.98 –0.93 –0.93 –0.92 –1.64 –1.51
9.7 103 1.2 103 8.8 103 7.4 103 5.2 103 5.3 102 4.8 103
EC, corrosion potential; j, corrosion current density.
Fig. 5. Corrosion rates in SBF at three pH values obtained from the immersion tests: (a) pH 5, (b) pH 7, (c) pH 10.
(Fig. 5). This difference is in accordance with the fact that zinc is more noble in nature than magnesium. The standard potentials of zinc and magnesium are –0.762 and –2.372 V (versus SHE), respectively [24]. The corrosion potentials measured in SBF are higher for the Zn-based alloys than for the Mg-based alloys by about 0.5–0.7 V (Fig. 6 and Table 4). (2) The corrosion rates of all of the alloys increase with decreasing pH (Fig. 5). For the zinc alloys a pH value reduction from 10 to 5 leads to an approximately twofold increase in the corrosion rate. This behavior has been well known for a long time and is schematically expressed by the Pourbaix diagrams of both metals. According to these diagrams magnesium corrodes in acidic, neutral and slightly alkaline environments, irrespective of their oxidation properties. In a strongly alkaline environment a passive magnesium hydroxide (Mg(OH)2) layer forms, which significantly slows down corrosion [24]. The Pourbaix diagram of zinc indicates that it corrodes in acidic and strongly alkaline liquids, while in neutral and slightly alkaline liquids it is passivated [24]. (3) The influence of Mg on the corrosion rate and corrosion parameters of the Zn–Mg alloys is small (Fig. 5 and Table 4). This could be explained by the low Mg concentrations used in this study. Moreover, magnesium forms the Mg2Zn11 phase in the Zn–Mg alloys (Fig. 2b–d), and it can be assumed that the standard potential of this phase is not very different from that of pure zinc. Therefore, the formation of galvanic micro-cells between Zn and Mg2Zn11, which would accelerate corrosion and shift the corrosion parameters, is unlikely.
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522 Table 5 Surface chemical compositions of the alloys corroded in SBF at pH 7 for 336 h (EDS).a Alloy
Element (wt.%)
Zn ZnMg1 ZnMg1.5 ZnMg3 ZnAl4Cu1 Mg AZ91HP a
Mg
Zn
5.0 3.0 5.8 6.5 5.2 47.8 54.9
64.5 86.0 37.0 63.6 33.8
Al
Na
1.1 0.6
0.6
10.3
Ca
P
14.2 4.2 31.0 14.1 32.3 26.8 14.6
16.3 6.7 26.2 15.9 27.1 24.9 19.6
Cu
0.5
C, H and O are not included.
Fig. 7. Changes in pH value during exposure of the alloys to SBF.
It should be noted that the corrosion rates of the zinc alloys investigated in this study are considerably lower compared with the highly corrosion-resistant Mg-based biodegradable alloys. As reported recently by Gu et al. [19], these corrosion-resistant alloys are based on pure Mg, Mg–Zn, Mg–rare earth or Mg–Zn–Ca bulk metallic glasses. In SBF with a pH of about 7 all of these alloys show corrosion rates of approximately 200 lm year1 or higher, which is at least a sixfold increase relative to the Zn–Mg alloys (Fig. 5b). Fig. 7 shows the evolution of the pH value during immersion of the alloys in SBF pH 7. It is evident that the pH value progressively increases as corrosion proceeds. The observed pH value change is attributable to the base corrosion reactions that produce OH anions:
Mg þ 2H2 O ! Mg2þ þ H2 þ 2OH
ð1Þ
Zn þ 2H2 O ! Zn2þ þ H2 þ 2OH
ð2Þ
3521
As can be observed in Fig. 7, the reaction of the Zn and ZnMg1 alloys is significantly slower, as it only causes the pH value to increase to approximately 7.4–7.5 after 90 h immersion. For the Mg and AZ91HP alloys the pH value increases to about 7.8 due to the corrosion reaction. The difference between the final pH values is seemingly small. However, it should be noted that this small difference indicates a relatively large difference in the numbers of OH– anions produced by reactions (1) and (2). A simple calculation reveals that the concentrations of OH– anions after 90 h immersion of Mg and Zn are 6 107 and 2.5 107 mol l1, respectively. Taking the stoichiometry of the reactions expressed by Eqs. (1) and (2), one can estimate that there is a fivefold difference between Zn and Mg in terms of their volume reduction due to corrosion. This difference agrees well with the corrosion rates shown in Fig. 5b, in which the corrosion rates of Mg and Zn are shown to be approximately 220 and 50 lm year1, respectively. The surfaces of the alloys exposed to SBF at pH 7 for 336 h are similar and they are illustrated in Fig. 8. The surfaces of all of the alloys are covered by relatively compact layers of corrosion products. Due to internal stress these layers are cracked and locally detached from the surface in some areas. The phase compositions of the corrosion products were analyzed by XRD. However, this method revealed only those phases present in the original alloys (Fig. 2) and no new phases were accurately determined after corrosion. This suggests that the layers of corrosion products that remain adherent on the surface are very thin. The chemical compositions of the corroded surfaces measured by EDS and XPS are summarized in Tables 5 and 6, respectively. The difference between the two methods is the depth to which each method is capable of analyzing material. EDS analyzes an approximately 1 lm thick surface layer, while the XPS signal is acquired from a surface layer with a thickness of only several interatomic distances. Besides the basic components of the alloys, one can observe in Table 5 that the corrosion products mainly contain phosphorus and calcium, both originating from the SBF. It should be noted that light elements like O, C and H are not included in this analysis. XPS analysis (Table 6) indicates that the Ca2+, Mg2+, Zn2+ and P5+ oxidation states predominate on the surface. Therefore, it can be assumed that the surfaces of the corroded zinc and magnesium alloys are dominated by insoluble calcium–magnesium phosphates. In the zinc alloys the phosphate contains a low concentration of zinc. Carbonates may also exist on the corroded surfaces, although their presence cannot be proved by EDS and XPS analyses. The formation of insoluble phosphates can be attributed to the presence of hydrogen phosphate and dihydrogen phosphate in the SBF. As indicated by Eqs. (1) and (2) and by Fig. 7, the corrosion reaction is accompanied
Fig. 8. The surfaces of the alloys exposed to the SBF at pH 7 for 336 h: (a) AZ91HP, (b) ZnMg1.5.
D. Vojteˇch et al. / Acta Biomaterialia 7 (2011) 3515–3522
3522
Table 6 Surface chemical compositions of two selected alloys corroded in SBF at pH 7 for 336 h (XPS).a Alloy
AZ91HP ZnMg1
Element (wt.%) (oxidation state)
of the Czech Republic (Projects Nos. MSM6046137302 and MSMT 21/2010). Appendix A. Figures with essential color discrimination
O
Ca
P
Mg
Zn
37.5 38.7
25.2 (Ca2+) 27.8 (Ca2+)
25.4 (P5+) 26.7 (P5+)
11.9 (Mg2+) 4.8 (Mg2+)
– 1.9 (Zn2+)
a C and H are not included. The high oxygen concentrations are caused by adsorption on the surface.
Certain figures in this article, particularly Figs. 5–7, are difficult to interpret in black and white. The full color images can be found in the on-line version, at doi:10.1016/j.actbio.2011.05.008. References
by alkalization of the SBF. As a result, insoluble phosphates are formed by the following simplified reaction: 6Mg2þ ðCa2þ Þ þ 2HPO2 4 þ 2H2 PO4 þ 6OH
! 2ðMg; CaÞ3 ðPO4 Þ2 ðinsolubleÞ þ 6H2 O
ð3Þ
Due to the presence of chloride anions in the SBF the insoluble corrosion products may be partially transformed to soluble chlorides, which accelerates corrosion. Chlorides were not detected in the insoluble corrosion products, either by EDS or by XPS, as is indicated in Tables 5 and 6. From a biocompatibility point of view the possible toxicity of zinc alloys should be taken into account. Compared with magnesium, which has also been considered for biodegradable implants, the maximum acceptable doses of zinc are lower, and overdoses may cause medical problems. It is, therefore, important to estimate the possible intake of zinc from the corrosion of a biodegradable implant. Let us consider a typical screw 5 mm in diameter and 50 mm in length made of zinc for the fixation of a fractured cortical bone. As indicated in Fig. 5a, the corrosion rate of this screw in a slightly acidic environment could be about 80 lm year1. A simple calculation reveals that the average daily intake of zinc from one screw is approximately 1.5 mg Zn. This value is negligible compared with the RDA and recommended upper limit for zinc, which are 15 and 40 mg day1, respectively [17]. 4. Conclusions New Zn–Mg alloys containing approximately 1 wt.% Mg are proposed for use in biodegradable medical implants, such as bone fixations. These alloys can be considered alternatives to Mg-based biodegradable alloys. In the as-cast state Zn–Mg alloys have a relatively fine structure composed of primary zinc and an interdendritic eutectic mixture. Such a structure results in good mechanical strength, comparable with that of bone, which is significantly higher than that of biodegradable polymeric materials. It is assumed that additional improvements in the strength of Zn–Mg alloys can be achieved by the use of hot extrusion. The main advantage of zinc alloys over magnesium alloys lies in their significantly better corrosion resistance in SBF. The low corrosion rate results in low rates of pH value increase and hydrogen evolution. Both factors are very important to prevent negative reactions in tissue during healing. In addition, the possible doses of zinc ions released from implants are significantly below zinc tolerable limit. Acknowledgments The research on biodegradable metallic materials was financially supported by the Ministry of Education, Youth and Sports
[1] Davis JR. Handbook of materials for medical devices. Materials Park, OH: ASM International; 2003. [2] Witte F. The history of biodegradable magnesium implants. Acta Biomater 2010;6:1680–92. [3] Zhang ChY, Zeng RCh, Liu ChL, Gao JCh. Comparison of calcium phosphate coatings on Mg–Al and Mg–Ca alloys and their corrosion behavior in Hank’s solution. Surf Coat Technol 2010;204:3636–40. [4] Perada MD, Alonso C, Burgos-Asperilla L, del Valle JA, Ruano OA, Perez P, et al. Corrosion inhibition of powder metallurgy Mg by fluoride treatments. Acta Biomater 2010;6:1772–82. [5] Gray-Munro JE, Seguin Ch, Strong M. Influence of surface modification on the in vitro corrosion rate of magnesium alloy AZ31. J Biomed Mater Res Part A 2009;91:221–30. [6] Hiromoto S, Shishido T, Yamamoto A, Maruyama N, Somekawa H, Mukai T. Precipitation control of calcium phosphate on pure magnesium by anodization. Corrosion Sci 2008;50:2906–13. [7] Erbel L et al. Temporary scaffolding of coronary arteries with bioabsorbable magnesium stents: a prospective, non-randomised multicentre trial. Lancet 2007;369:1869–75. [8] Di Mario C et al. Drug-eluting bioabsorbable magnesium stent. J Interven Cardiol 2004;17:391–5. [9] Peeters P, Bosiers M, Verbist J, Deloose K, Heublein B. Preliminary results after application of absorbable metal stents in patients with critical limb ischemia. J Endovasc Ther 2005;12:1–5. [10] Zhang S. At al. Research of Mg–Zn alloy as a degradable biomaterial. Acta Biomater 2010;6:626–40. [11] Li Z, Gu X, Lou S, Zheng Y. The development of binary Mg–Ca alloys for use as biodegradable materials within bone. Biomaterials 2008;29:1329–44. [12] Zhang E, Yang L, Xu J, Chen H. Microstructure, mechanical properties and biocorrosion properties of Mg–Si(–Ca, Zn) alloy for biomedical application. Acta Biomater 2010;6:1756–62. [13] Zhang E, Yang L. Microstructure, mechanical properties and bio-corrosion properties of Mg–Zn–Mn–Ca alloy for biomedical application. Mater Sci Eng A 2008;497:111–8. [14] Gu XN, Zheng W, Cheng Y, Zheng YF. A study on alkaline heat treated Mg–Ca alloy for the control of the biocorrosion rate. Acta Biomater 2009;5:2790–9. [15] Huan ZG, Leeflang MA, Zhou J, Fratila-Apachitei LE, Duszczyk J. In vitro degradation behaviour and cytocompatibility of Mg–Zn–Zr alloys. J Mater Sci Mater Med 2010;21:2623–35. [16] Witte F, Hort N, Vogt C, Cohen S, Kainer KU, Willumeit R, et al. Degradable biomaterials based on magnesium corrosion. Curr Opin Solid State Mater Sci 2008;12:63–72. [17] Fosmire GJ. Zinc toxicity. Am J Clin Nutr 1990;51:225–7. [18] Gale WF, Totemeier TC. Smithells metals reference book. 8th ed. Oxford: Elsevier; 2004. [19] Gu X, Zheng Y, Zhong S, Xi T, Wang J, Wang W. Corrosion of, and cellular responses to Mg–Zn–Ca bulk metallic glasses. Biomaterials 2010;31:1093–103. [20] Li QF, Weng HR, Suo ZY, Ren YL, Yuan XG, Qiu KQ. Microstructure and mechanical properties of bulk Mg–Zn–Ca amorphous alloys and amorphous matrix composites. Mater Sci Eng A 2008;487:301–8. [21] Zberg B, Uggowitzer PJ, Loffler JF. MgZnCa glasses without clinically observable hydrogen evolution for biodegradable implants. Nat Mater 2009;8:887–91. [22] Wang X, Lu H, Li X, Li L, Zheng Y. Effect of cooling rate and composition on microstructures and properties of Zn–Mg alloys. Trans Nonferrous Met Soc China 2007;17:122–5. [23] Wang H, Shi ZM, Yang K. Magnesium and magnesium alloys as degradable metallic biomaterials. Adv Mater Res 2008;32:207–10. [24] Talbot DEJ, Talbot JDR. Corrosion science and technology. London: CRC Press; 2007. ˇ ízˇová H, Volenec K. Investigation of magnesium-based alloys for [25] Vojteˇch D, C biomedical applications. Kovove Mater 2006;44:211–23. [26] Davis JR, editor. Properties and selection: nonferrous alloys and specialpurpose materials. ASM handbook, vol. 2. Materials Park, OH: ASM International; 1990.