This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and

Transcription

1 This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier s archiving and manuscript policies are encouraged to visit:

2 Acta Biomaterialia 6 (2010) Contents lists available at ScienceDirect Acta Biomaterialia journal homepage: Corrosion fatigue behaviors of two biomedical Mg alloys AZ91D and WE43 In simulated body fluid X.N. Gu a,b, W.R. Zhou a,b, Y.F. Zheng a,b,c, *, Y. Cheng c, S.C. Wei c, S.P. Zhong c, T.F. Xi c, L.J. Chen d a State Key Laboratory for Turbulence and Complex System, College of Engineering, Peking University, Beijing , China b Department of Advanced Materials and Nanotechnology, College of Engineering, Peking University, Beijing , China c Center for Biomedical Materials and Tissue Engineering, Academy for Advanced Interdisciplinary Studies, Peking University, Beijing , China d School of Material Science and Engineering, Shengyang University of Technology, Shenyang , China article info abstract Article history: Received 27 March 2010 Received in revised form 16 July 2010 Accepted 20 July 2010 Available online 23 July 2010 Keywords: Fatigue Corrosion fatigue Magnesium alloy Biomedical metallic materials Magnesium alloys have been recently developed as biodegradable implant materials, yet there has been no study concerning their corrosion fatigue properties under cyclic loading. In this study the die-cast AZ91D (A for aluminum 9%, Z for zinc 1% and D for a fourth phase) and extruded WE43 (W for yttrium 4%, E for rare earth mischmetal 3%) alloys were chosen to evaluate their fatigue and corrosion fatigue behaviors in simulated body fluid (SBF). The die-cast AZ91D alloy indicated a fatigue limit of 50 MPa at 10 7 cycles in air compared to 20 MPa at 10 6 cycles tested in SBF at 37 C. A fatigue limit of 110 MPa at 10 7 cycles in air was observed for extruded WE43 alloy compared to 40 MPa at 10 7 cycles tested in SBF at 37 C. The fatigue cracks initiated from the micropores when tested in air and from corrosion pits when tested in SBF, respectively. The overload zone of the extruded WE43 alloy exhibited a ductile fracture mode with deep dimples, in comparison to a brittle fracture mode for the die-cast AZ91D. The corrosion rate of the two experimental alloys increased under cyclic loading compared to that in the static immersion test. Ó 2010 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. 1. Introduction Magnesium alloys have been widely studied recently for their potential applications as implant materials within bone or blood vessels [1 10]. Extensive investigations indicated that two types of magnesium alloy systems might satisfy the general requirements for biodegradable biomaterials, including good mechanical properties, biocompatibilities and biodegradation properties: (1) The AZ series alloy system: Witte et al. [2] first showed the gradual degradation process of gravity-cast AZ91 alloy within guinea pig femora and observed increased bone mass around the magnesium rods. Since then, many investigations for bulk AZ91 alloy [3,5], porous AZ91D alloy [6,7] and surface modified AZ91D alloy [8,9] have been carried out for the orthopedic application. For example, the degrading AZ91D alloy scaffold was found to promote both bone formation and resorption in a rabbit model [7], and even the fast degrading of an AZ91D scaffold exhibited good biocompatibility with an appropriate inflammatory host response [6]. * Corresponding author at: Department of Advanced Materials and Nanotechnology, College of Engineering, Peking University, Beijing , China. Tel./fax: address: yfzheng@pku.edu.cn (Y.F. Zheng). (2) WE series alloy system: Mario et al. [10] studied the in vivo corrosion of magnesium stents made of WE43 alloy, and their animal and clinical experimental results showed that the WE43 alloy stents possessed sufficient antiproliferative properties and drastically reduced the restenosis rate in comparison to conventional stainless steel stents [10]. The PROGRESS-AMS (Clinical Performance and Angiographic Results of Coronary Stenting with Absorbable Metal Stents) clinical trial results [11] showed that magnesium alloy stents could achieve an immediate angiographic result similar to the result of other metal stents and could be safely degraded after 4 months. Although the recipe for AMS was a commercial secret of BIOTRONIK, the work of Waksman et al. [12] and Loos et al. [13] revealed that the AMS stent was constructed from magnesium alloy containing zirconium, yttrium and rare earth metals, which was similar to the composition of WE43 alloy. Implants that function as bone and blood vessels, including bone plates, screws and vascular stents, are usually used under severe cyclic loading conditions, such as tension, compression and bending. It has been found that a material very often fails at a cyclic load well below the threshold to produce failure by single application, which means a fatigue failure [14] /$ - see front matter Ó 2010 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi: /j.actbio

3 4606 X.N. Gu et al. / Acta Biomaterialia 6 (2010) For orthopedic implants, the ultimate implant failure is usually associated with the corrosion fatigue, which is the synergetic effect of electrochemical corrosion and cyclic mechanical loading [15,16]. For example, Azevedo [15] found the premature fracture of a titanium reconstruction plate for osteosynthesis was associated with a corrosion fatigue mechanism. For vascular devices, the failure of a stent due to fatigue may result in loss of radial support of the stented vessel or in perforation of the vessel by the stent struts. It was estimated that approximately two-thirds of the failures of vascular devices would result in patient death [14,17]. As documented in FDA Guidance Document 1545 [18] and ASTM-F [19], the in vitro fatigue of vascular stents should be measured using the fully processed and implant quality product through pulsatile methods which simulate the loading that stent will experience in vivo. Clearly, the fatigue of vascular stent is strongly dependent on the fatigue performance of the actual materials it is made of, as well as its pattern design. The corrosion fatigue of magnesium alloys have been studied concerning their engineering application in 3.5 wt.% NaCl (ph 5) [20,21], 5 wt.% NaCl (ph 6.59) [22] and 0.1N Na 2 B 4 O 7 (ph 9.3) [20]. Yet as a new type of potential biomedical metallic material, the corrosion fatigue behavior of magnesium alloys examined by the standard metallic biomaterial testing method in a physiological environment has not been done. In the present study, the fatigue and corrosion fatigue behaviors in SBF of the die-cast AZ91D (A for aluminum 9%, Z for zinc 1% and D for fourth phase) and extruded WE43 (W for yttrium 4%, E for rare earth mischmetal 3%) alloys were systematically investigated. 2. Experimental procedure 2.1. Preparation of materials The die-cast AZ91D alloy (89.59 wt.% Mg, 9.21 wt.% Al, 0.80 wt.% Zn, 0.34 wt.% Mn, 0.06 wt.% Si) was provided by Shengyang University of Technology, China. The extruded WE43 alloy (91.35 wt.% Mg, 4.16 wt.% Y, 3.80 wt.% RE, 0.36 wt.% Zr, 0.20 wt.% Zn, 0.13 wt.% Mn) was provided by Changchun Zhong-Ke-Xi-Mei Magnesium Alloy Co. Ltd., China, with an extrusion ratio of 10. The asreceived material was further machined into standard surfacesmooth fatigue samples according to ASTM-E [23] with a circular cross-section 5 mm in diameter and a gage length of 10 mm Microstructural characterization The experimental samples were polished and etched with a mixture of 1 g of oxalic acid, 1 ml of acetic acid, 1 ml of nitric acid and 98 ml of water, and observed with an environmental scanning electron microscope (ESEM; Quanta 200FEG). X-ray diffractometry (XRD; Rigaku DMAX 2400) using Cu Ka radiation was employed for the identification of the constituent phases in the experimental samples Tensile test The standard surface-smooth tensile samples of the experimental alloys were machined according to ASTM-E8-04 [24]. The tensile tests were carried out at a displacement rate of 1 mm min 1 by an Instron 3365 universal test machine. An average of at least three measurements was taken for each group Fatigue test Axial fatigue tests were conducted using a computer-controlled servo-hydraulic PLD-50KN testing machine with sinusoidal loading control, with a stress ratio of 1 and a frequency of 10 Hz. The fatigue test was continued until complete failure of the sample, unless the test was stopped when the sample did not fail up to 10 7 cycles. The maximum stress at which the sample has not failed at 10 7 cycles is defined as a fatigue limit in this study. At every load condition at least three samples were tested. The fatigue fracture surfaces were ultrasonically cleaned in ethanol and dried in air prior to the morphology observation. The ESEM was used to observe the fracture surfaces to locate the fatigue crack initiation site and help determining the fatigue fracture mechanism Corrosion fatigue test The medium used in the corrosion fatigue test was SBF (containing g l 1 NaCl, g l 1 NaHCO 3, g l 1 KCl, g l 1 K 2 HPO 4?3H 2 O, g l 1 MgCl 2?6H 2 O, g l 1 CaCl 2, g l 1 Na 2 SO 4 and g l 1 Tris (HOCH 2 ) 3 CNH 2, following the preparation procedure of Ref. [25]). An acrylic chamber attachment was self-designed and assembled on the PLD-50KN testing machine, which guaranteed the gage length of the sample was immersed in the SBF throughout the testing procedure. The SBF solution (2 l in total volume) was kept at 37 ± 1 C by a water bath, and circulated through the chamber by a pump. During the corrosion fatigue tests, the electrolyte was changed every 24 h. The electrolyte would be replaced for each new test. The release of magnesium and the alloying element into the electrolyte during the corrosion fatigue tests was examined by inductively coupled plasma atomic emission spectrometry (Leeman, Profile ICP-AES). The fatigue-fractured samples were immersed in 10 g l 1 CrO 3 solution to clean the corrosion product and then ultrasonically cleaned in distilled water. The corrosion fatigue fracture surfaces after the cleaning were observed with the ESEM. The static immersion test in SBF (2 l) under no load was also carried out and the ionreleasing concentration with the same immersion period as those of the corrosion fatigue samples were measured. The corrosion rate was calculated using the equation: v corr ¼ C Mg V At where C Mg is the Mg ion concentration, V is the electrolyte volume, A is the original surface area exposed to the electrolyte and t is the exposure time Statistical analysis Analysis of variance was used to evaluate the difference in the corrosion rate. Statistical significance was defined as Results 3.1. Mechanical properties, fatigue and corrosion fatigue of the diecast AZ91D alloy Fig. 1 shows the typical microstructure and tensile strength of the die-cast AZ91D alloy. As can be seen from Fig. 1a, the microstructure of the die-cast AZ91D alloy is composed of a(mg) matrix and Mg 17 Al 12 phase precipitating along the grain boundaries, with the average grain size of 250 lm. The yield strength, ultimate strength and elongation of the die-cast AZ91D alloy during tension are ± MPa, ± 4.85 MPa and 4.3 ± 0.48%, respectively, as deduced from Fig. 1c. Fig. 2 shows the stress life (S N) behaviors for the die-cast AZ91D alloy tested in air and in SBF at 37 C. It can be seen that the fatigue limit of the die-cast AZ91D alloy at 10 7 cycles is 50 MPa in air compared to 20 MPa at 10 6 cycles tested in SBF.

4 X.N. Gu et al. / Acta Biomaterialia 6 (2010) Fig. 1. XRD patterns of (a) die-cast AZ91D and (b) extruded WE43 alloys (with SEM image of the microstructure as inset), and tensile stress strain curves of (c) die-cast AZ91D and (d) extruded WE43 alloys. Fig. 3 shows the typical high-cycle fatigue fracture surfaces of the die-cast AZ91D alloy tested in air, which are typically comprised of three morphologically distinct zones: the crack initiation zone Fig. 2. S N curves for die-cast AZ91D at room temperature in air and in SBF at 37 C. The experiment was run for cycles. (Fig. 3b), the crack propagation zone (Fig. 3c) and the final stage overload zone (Fig. 3d). It shows similar crack initiation morphology (Fig. 3b) to that observed for AZ91D [26] and as-rolled AZ31 alloy [27], described as transgranular or intergranular cracks. In addition, the extremely low number of cycles to failure of the die-cast AZ91D alloy samples is linked to the entrapped inclusions or casting pores, some of which can even be seen by eye on the fracture surface. Fig. 4 shows a typical high-cycle fatigue fracture surface after cleaning the corrosion products on the surface of the die-cast AZ91D alloy samples tested in SBF. Different from the fatigue fracture surface in air, the failure of the samples tested in SBF is attributed to the formation of the corrosion pit, which is visible in the crack nucleation region on the sample surface (arrowed in Fig. 4b). Various sizes of deep corrosion pits, ranging from 10 to 200 lm (arrowed in Fig. 4b), are observed on both the fracture surface and the sample cylinder surface. The crack growth images for die-cast AZ91D tested in SBF are similar to those tested in air. The overload zone of the die-cast AZ91D alloy shows a brittle fracture surface feature, as shown in Fig. 4d. Fig. 5 shows the corrosion rate of the die-cast AZ91D alloy immersed in static electrolyte and under a cyclic load in circulating SBF. The corrosion rate of the die-cast AZ91D alloy sample is calculated from the measured Mg concentration in electrolyte incubating samples under cyclic loads or no load for the same immersion period. It can be seen that the die-cast AZ91D alloy exhibits a

5 4608 X.N. Gu et al. / Acta Biomaterialia 6 (2010) Fig. 3. SEM images showing that the fatigue fracture surface morphology of the die-cast AZ91D alloy failed over 10 6 cycles in air. (a) Overall fracture surface; (b) crack initiation site; (c) crack propagation site; (d) overload site. significantly increasing corrosion rate (p < 0.05) under a cyclic load, while the corrosion rate increases with the increasing cyclic load Mechanical properties, fatigue and corrosion fatigue of the extruded WE43 alloy Fig. 1b shows the microstructure of the extruded WE43 alloy, indicating that the grain size ranges from 10 to 30 lm. Besides the matrix a(mg) phase identified from the XRD pattern of the extruded WE43 alloy, many Nd-rich and Y-rich precipitation was visible in the SEM microstructure (insets in Fig. 1b), which was identified as Mg 24 Y 5,Mg 41 Nd 5 and Mg 12 Nd phases in a previous investigation [28]. The tensile test results (Fig. 1d) indicate that the yield strength, ultimate tensile strength and elongation are ± 2.89 MPa, ± 4.39 MPa and ± 5.3%, respectively. Fig. 6 shows the S N behavior for the extruded WE43 alloy tested in air and in SBF at 37 C. The corrosion fatigue limit of the extruded WE43 alloy is obviously lower than the fatigue limit of the extruded WE43 alloy (40 vs. 110 MPa). Fig. 7 shows SEM images of typical high-cycle fatigue fracture surfaces of the extruded WE43 alloy samples. It can be seen that the crack nucleation region of the extruded WE43 alloy is relatively flat with a pore (marked by an arrow in Fig. 7b). The crack growth picture (as shown in Fig. 7c) indicates a transgranular fracture mode with a plastically deformed zone. In the overload zone, relative flat regions as well as dimples with sizes of a few microns are observed (Fig. 7d), indicating the ductile fracture mode of the extruded WE43 alloy sample. Fig. 8 shows typical high-cycle fatigue fracture surfaces after cleaning the corrosion products on the surface of the extruded WE43 alloy sample tested in SBF at 37 C. Clearly, the failure of the extruded WE43 alloy samples can be attributed to the formation of the corrosion pit, resulting in the crack nucleation region visible in Fig. 8b. The corrosion pits on the extruded WE43 alloy sample are relatively shallow and the overload zone of the extruded WE43 alloy sample exhibits a dimpled fracture surface (Fig. 8d), the dimples of which are more shallow than on that tested in air. Fig. 9 shows the corrosion rate of the extruded WE43 alloy immersed in static electrolyte and tested under a cyclic load in circulating SBF. It can be seen that the corrosion rate of fatigue samples increases significantly (p < 0.05) under a cyclic load and that this variation becomes more obvious under higher cyclic loads, e.g. under 120 MPa. 4. Discussion 4.1. Factors influencing the fatigue and corrosion fatigue of biomedical magnesium alloys Alloying Alloying influences the mechanical and corrosion properties of magnesium and thus influences the fatigue and corrosion fatigue behaviors. Al and Zn increase the tensile strength and alloying with Al in general improves the corrosion resistance due to the effect of continuous b-phase barriers [29]. In this study, severe localized corrosion was observed for die-cast AZ91D alloy during the corrosion fatigue test (Fig. 4b) and serious pitting corrosion was also reported for its in vivo corrosion measurement [30]. Y and Nd elements form eutectic systems of limited solubility with magnesium, which imparts a significant increase in strength through precipitation hardening [31]. In addition, the presence of Y and Nd

6 X.N. Gu et al. / Acta Biomaterialia 6 (2010) Fig. 4. SEM images showing that the fatigue fracture surface morphology of the die-cast AZ91D alloy after removing the surface corrosion product failed over 10 5 cycles in SBF at 37 C. (a) Overall fracture surface; (b) crack initiation site; (c) crack propagation site; (d) overload site. The inset in (b) shows a profile SEM image of the corrosion fatigue sample examined at a tilt angle of 45. Fig. 5. The corrosion rate of the die-cast AZ91D alloy immersion in static electrolyte and under a cyclic load in circulating SBF. * p < improves the tendency of magnesium to passivation in Na 2 SO 4 and NaCl solutions [32]. However, the amount of these alloying elements added should be conservative since elemental Al is a neurotoxicant [33] and releasing Al ions leads to defective bone mineralization [34], which as a consequence retards the formation of chemical bonds to implants [35]. In addition, it was reported that elements Nd and Y distributed mainly at the implantation site [2], which might be not tolerated by the human body. Fig. 6. S N curves for extruded WE43 alloy at room temperature in air and in SBF at 37 C. The experiment was run for cycles Microstructure The fatigue life usually correlates inversely with the fatigue crack nucleation pore in both as-cast [36,37] and wrought Mg alloy materials [38], and this acts as stress concentration. The release of stress may result in slip bands and microcracks, and may thus accelerate the initiation and growth of cracks [26,37 39]. In addition, the fatigue life is also influenced by the grain size of the magnesium alloy. The extruded WE43 alloy has a substantially smaller

7 4610 X.N. Gu et al. / Acta Biomaterialia 6 (2010) Fig. 7. SEM images showing that the fatigue fracture surface morphology of the extruded WE43 alloy failed over 10 6 cycles in air. (a) Overall fracture surface; (b) crack initiation site; (c) crack propagation site; (d) overload site. The inset in (d) shows an enlarged SEM image of the overload site. grain size than does the die-cast AZ91D alloy, and thus greater fatigue strength. This may be attributed to the reduced grain size after extrusion, leading to inhibited crack initiation, inhibited dislocation motion and an increase in the number of barriers to early crack propagation [40]. This finding is in agreement with the results that the fatigue life of AZ91E and AZ31 alloys decreases with increasing grain size [37,40] The surrounding environment In this work, the fatigue strength of the experimental samples tested in SBF was much lower than those tested in air. This phenomenon was also observed for AZ31, AZ61 and AM50 alloys, which showed decreased fatigue strength tested in NaCl than those tested in air at the same stress level [18,20]. Previous work [41,42] suggested that the corrosion behavior of magnesium alloy in SBF was much more complicated considering the rapid passivation effect of HCO 3 and the influence of the Tris buffer. Rettig and Virtanen [43] indicated that the corrosion rate of extruded WE43 alloy in m-sbf was significantly higher than that in NaCl. Also, the dynamic circulating electrolyte contributes to the corrosion of magnesium alloy. It was reported that the degradation behavior of as-cast AM60B alloy was different when tested under static and dynamic conditions [44] Cyclic loading The magnesium alloy sample under constraining conditions has a higher dissolution rate than that in the free state because of the well-known stress corrosion phenomenon. Here we observed that the corrosion rate of the die-cast AZ91D alloy in the corrosion fatigue test was about 7 8 times that tested in static SBF, and the corrosion rate of the extruded WE43 alloy in the corrosion fatigue test was about 4 12 times that tested in static SBF. On the one hand, this can be attributed to the deformation state of alloys under fatigue test. Eliezer et al. [45] found a 7-fold and a 1.5-fold increase in the dissolution rate of strained AZ91D and AM50 alloy, respectively, with the plastic strain growth from 0% to 4%. On the other hand, under cyclic loading, cumulative plastic strains, the formation of extrusions and intrusions due to persistent slip bands would break the corrosion product film, and thus the condensed electrolyte droplets would penetrate into the underlying sample through the crevice caused by the breakdown of the layer and the crevice corrosion that occurred [46,47]. The stress concentration occurs at the root of the pit and enhances the growth of the corrosion pit to the critical pit size at which the stress intensity factor reaches the threshold for fatigue cracking [48]. Furthermore, the fatigue testing mode (e.g. axial and rotating fatigue) also influences the fatigue behavior of magnesium alloy. Miyashita et al. [49] found that at 10 7 cycles extruded AZ61 alloy exhibits about 30 MPa higher fatigue strength when tested by rotating bending fatigue mode than when tested by axial loading fatigue; the ratio between fatigue life under axial loading and that under rotating bending was about 0.77, which was attributed to the small crack growth behavior and the difference in strain Comparison of fatigue property among magnesium alloys and other biomaterials Fig. 10a illustrates the fatigue strength of the die-cast AZ91D, extruded WE43 and other Mg alloys for biomedical application, including AZ31 [2], AZ61 [5] and AM50 [50] at 10 6 cycles tested

8 X.N. Gu et al. / Acta Biomaterialia 6 (2010) Fig. 8. SEM images showing that the fatigue fracture surface morphology of the extruded WE43 alloy after removing the surface corrosion product failed over 10 5 cycles in SBF at 37 C. (a) Overall fracture surface; (b) crack initiation site; (c) crack propagation site; (d) overload site. The inset in (b) shows an SEM profile image of the corrosion fatigue sample examined at a tilt angle of 45. Fig. 9. The corrosion rate of the extruded WE43 alloy immersion in static electrolyte and under a cyclic load in circulating electrolyte. * p < in air and corrosion medium. Bhuiyan et al. [22] evaluated the reduction rate of fatigue strength for extruded AZ61 magnesium alloy by the ratio of (r LH r CE )/r LH, where r LH is the fatigue limit under low humidity condition and r CE is that under corrosion medium. Similarly, the effect of corrosion medium on fatigue strength of the present Mg alloys can be evaluated by using the reduction rate of fatigue strength (RRFS), as (r air r SBF )/r air, where r air is the fatigue strength tested in air and r SBF is the fatigue strength tested in SBF at the same number of cycles to failure. Correspondingly, the RRFS values are 67% for die-cast AZ91D in SBF, 33% for extruded WE43 in SBF, 48% for extruded AZ61 in 5 wt.% NaCl solution, and 30% for both extruded AM50 and AZ31 in 3.5 wt.% NaCl solution. From these comparisons, the fatigue data for Mg alloys under corrosion medium falls in a widely scatter band (RRFS = 30 67%) and the die-cast AZ91D seems to be more sensitive to the corrosion medium. Fig. 10b compares the fatigue strength of magnesium alloys with some clinically used biomaterials tested in a physiological environment. It can be seen that magnesium alloy shows a relatively wide fatigue strength range, which is much lower than that of alumina [51], Ti alloy [52,53] and ISO stainless steel [54] but indicates a higher fatigue strength than that of calcium phosphate bone cement [55]. Moreover, the high range of fatigue strength for magnesium alloy is comparable with that of 316L stainless steel [52] and also much greater than that of polymer [56 58]. For permanent implants, many fatigue fracture failure cases have been reported over the years of clinical application [15,16]. Biodegradable biomaterials should exist in vivo only for a certain period, thus it can be assumed that the biomaterial should possess the desired strength until it has served its purpose. For bone, typical strains are microstrain for normal walking, which equates to MPa, assuming the Young s modulus to be 17GPa [59]. It was reported that the mechanical properties of implants should last for 1 3 months. Since the normal walking cycle

9 4612 X.N. Gu et al. / Acta Biomaterialia 6 (2010) Fig. 10. (a) Comparison of the fatigue strength of die-cast AZ91D, extruded WE43 and other Mg alloys for biomedical application at 10 6 cycles tested in air and in corrosion medium. Note that uniaxial tension compression fatigue was conducted for extruded AZ61 alloy in Ref. [22] with a frequency of 20 Hz and a stress ratio of 1 in 5 wt.% NaCl solution, whereas a rotating beam-type fatigue was conducted for extruded AM50 and AZ31 alloy in Ref. [21] with a frequency of 30 Hz and a stress ratio of 1 in 3.5 wt.% NaCl solution. (b) Fatigue strength of Mg alloys and some clinically used biomaterials tested in a physiological environment [21,22,51 59]. Note that the fatigue strength of the polymer was tested in air. is repeated about 2 million times per year [59], the important service period ranges from to cycles. The predicted fatigue strength of the extruded WE43 alloy ranges from 103 MPa for cycles to 99 MPa for cycles, with a corrosion fatigue limit of 40 MPa, which are higher than the physiological strength ( MPa) and bone fatigue strength (23 30 MPa) [59]. In the case of the die-cast AZ91D alloy, the strength threshold of 25.5 MPa appears at cycles, or nearly 1.5 months, which means that the fatigue strength or corrosion resistance still needs to be improved further. 5. Conclusions (1) The fatigue strength of the die-cast AZ91D in air was 50 MPa. The fracture mode of the die-cast AZ91D alloy was brittle fracture. The corrosion fatigue strength of the die-cast AZ91D alloy in SBF was much lower than that in air, being 20 MPa at 10 6 cycles. (2) The fatigue strength of the extruded WE43 alloy in air was 110 MPa. The initiation of the crack was attributed to micropores in the samples. The fracture mode of the extruded WE43 alloy was toughness fracture. The corrosion fatigue strength of the extruded WE43 alloy in SBF was 40 MPa at 10 7 cycles. (3) Under cyclic loading in SBF, both the die-cast AZ91D and extruded WE43 alloys showed increased corrosion rate. Their corrosion rates increased with increasing cyclic loading. (4) The fatigue life of the die-cast AZ91D and extruded WE43 alloys was mainly related to their alloying elements, microstructures, surrounding environment and cyclic loading. Under the combined conditions of cyclic loading and a corrosive environment, the crack initiated from the corrosion pit with the release of stress concentration. Acknowledgements This work was supported by the Research Funds for the Central Universities (PKUJC ), Foundation of Key Laboratory for Advanced Materials Processing Technology, Ministry of Education ( ), National Natural Science Foundation of China (No ), Beijing Municipal Natural Science Foundation ( ) and Program for New Century Excellent Talents in University (NCET ). Appendix A. Figures with essential colour discrimination Certain figures in this article, particularly Figs. 1, 2, 5, 6, 9, 10, are difficult to interpret in black and white. The full colour images can be found in the on-line version, at doi: /j.actbio References [1] Heublein B, Rohde R, Kaese V, Niemeyer M, Hartung W, Haverich A. Biocorrosion of magnesium alloys: a new principle in cardiovascular implant technology? Heart 2003;89: [2] Witte F, Kaese V, Switzer H, Meyer-Lindenberg A, Wirth CJ, Windhag H. In vivo corrosion of four magnesium alloys and the associated bone response. Biomaterials 2005;26: [3] Kannan MB, Raman RKS. Evaluating the stress corrosion cracking susceptibility of Mg Al Zn alloy in modified-simulated body fluid for orthopaedic implant application. Scripta Mater 2008;59: [4] 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: [5] Kannan MB, Raman RKS. In vitro degradation and mechanical integrity of calcium-containing magnesium alloys in modified-simulated body fluid. Biomaterials 2008;29: [6] Witte F, Ulrich H, Rudert M, Willbold E. Biodegradable magnesium scaffolds. Part I. Appropriate inflammatory response. J Biomed Mater Res 2007;A81: [7] Witte F, Ulrich H, Palm C, Willbold E. Biodegradable magnesium scaffolds. Part II. Peri-implant bone remodeling. J Biomed Mater Res 2007;A81: [8] Zhang XP, Zhao ZP, Wu FM, Wang YL, Wu J. Corrosion and wear resistance of AZ91D magnesium alloy with and without microarc oxidation coating in Hank s solution. J Mater Sci 2007;42: [9] Dutta Majumdar J, Bhattacharyya U, Biswas A, Manna I. Studies on thermal oxidation of Mg-alloy (AZ91) for improving corrosion and wear resistance. Surf Coat Technol 2008;202: [10] Mario C, Griffiths H, Goktekin O, Peeters N, Verbist J, Bosiers M, et al. Drugeluting bioabsorbable magnesium alloys. J Interven Cardiovasc 2004;17(6): [11] Erbel R, Mario CD, Bartunek J, Bonnier J, de Bruyne B, Eberli FR, et al. Temporary scaffolding of coronary arteries with bioabsorbable magnesium stents: a prospective non-randomised multicentre trial. Lancet 2007;369: [12] Waksman R, Pakala R, Kuchulakanti PK, Baffour R, Hellinga D, Seabron R, et al. Safety and efficacy of bioabsorbable magnesium alloy stents in porcine coronary arteries. Catheter Cardiovasc Interven 2006;68: [13] Loos A, Rohde R, Haverich A, Barlach S. In vitro and in vivo biocompatibility testing of absorbable metal stents. Macromol Symp 2007;253:103 8.

10 X.N. Gu et al. / Acta Biomaterialia 6 (2010) [14] James BA, Sire RA. Fatigue-life assessment and validation techniques for metallic vascular implants. Biomaterials 2010;31: [15] Azevedo CRF. Failure analysis of a commercially pure titanium plate for osteosynthesis. Eng Fail Anal 2003;10: [16] Magnissalis EA, Zinelis S, Karachalios Th, Hartofilakidis G. Failure analysis of two Ti-alloy total hip arthroplasty femoral stems fractured in vivo. J Biomed Mater Res 2003;66B(1): [17] US Congress. The Bjork Shiley heart valve: earn as you learn. In: Subcommittee on Oversight and Investigations House Committee on Energy and Commerce, [18] US Food and Drug Adminstration. Non-clinical tests and recommended labeling for intravascular stents and associated delivery systems: guidance for industry and FDA staff. US Department of Health and Human Services, Food and Drug Administration, Center for Devices and Radiological Health; 13 January [19] ASTM-F Standard test methods for in vitro pulsatile durability testing of vascular stents. Philadelphia, PA: American Society for Testing and Materials; [20] Eliezer A, Gutman EM, Abramov E, Unigovski Y. Corrosion fatigue of die-cast and extruded magnesium alloys. J Light Metals 2001;1: [21] Unigovski Y, Eliezer A, Abramov E, Snir Y, Gutman EM. Corrosion fatigue of extruded magnesium alloys. Mater Sci Eng A 2003;360: [22] Bhuiyan SM, Mutoh Y, Murai T, Iwakami S. Corrosion fatigue behavior of the extruded magnesium alloy AZ61 under three different corrosive environments. Int J Fatigue 2008;30: [23] ASTM-E Standard practice for conducting force controlled constant amplitude axial fatigue tests of metallic materials. Philadelphia, PA: American Society for Testing and Materials; [24] ASTM-E8-04. Standard test methods for tension testing of metallic materials, annual book of ASTM standards. Philadelphia, PA: American Society for Testing and Materials; [25] Kokubo T, Takadama H. How useful is SBF in predicting in vivo bone bioactivity? Biomaterials 2006;27: [26] Wolf B, Fleck C, Eifler D. Characterization of the fatigue behavior of the magnesium alloy AZ91D by means of mechanical hysteresis and temperature measurements. Int J Fatigue 2004;26: [27] Tokaji KamakuraM, Ishiizumi Y, Hasegawa N. Fatigue behaviour and fracture mechanism of a rolled AZ31 magnesium alloy. Int J Fatigue 2004;26: [28] Yu K, Li W-X, Wang R-C, Wang B, Li C. Effect of rolling and heat treatment on mechanical properties and microstructure of WE43 magnesium alloy. Trans Mater Heat Treat 2008;29(2):95 8. [29] Song G, Atrens A. Corrosion mechanisms of magnesium alloys. Adv Eng Mater 1999;1(1): [30] Witte F, Fischer J, Nellesen J, Crostack H-A, Kaese V, Pisch A, et al. In vitro and in vivo corrosion measurements of magnesium alloys. Biomaterials 2006;27: [31] Kainer KU. Magnesium alloys and technology. Weinheim: Wiley-VCH Verlag; [32] Zucchi F, Grassi V, Frignani, Monticelli C, Trabanelli G. Electrochemical behavior of a magnesium alloy containing rare earth elements. J Appl Electrochem 2006;36(2): [33] El-Rahman SSA. Neuropathology of aluminum toxicity in rats (glutamate and GABA impairment). Pharmacol Res 2003;47: [34] Jones G, Sambrook PN. Drug-induced disorders of bone metabolism: incidence, management and avoidance. Drug Safety 1994;10(6): [35] Flaten TP, Alfrey AC, Birchall JD, Savory J, Yokrl RA. Status and future concerns of clinical and environmental aluminum toxicology. J Toxicol Environ Health 1996;48: [36] Horstemeyer MF, Yang N, Gall K, McDowell D, Fan J, Gullett P. High cycle fatigue mechanisms in a cast AM60B magnesium alloy. Fatigue Fract Eng Mater Struct 2002;25: [37] Horstemeyer MF, Yang N, Gall K, McDowell DL, Fan J, Gullett PM. High cycle fatigue of a die cast AZ91E-T4 magnesium alloy. Acta Mater 2004;52: [38] Shih T-S, Liu W-S, Chen Y-J. Fatigue of as-extruded AZ61A magnesium alloy. Mater Sci Eng A 2002;325: [39] Ogarevic VV, Stephens RI. Fatigue of magnesium alloys. Annu Rev Mater Sci 1990;20: [40] Zúberová Z, Kunz L, Lamark TT, Estrin Y, Janeček M. Fatigue and tensile behavior of cast, hot-rolled, and severely plastically deformed AZ31 magnesium alloy. Metall Mater Trans A 2007;38A: [41] Xin Y, Huo K, Tao H, Tang G, Chu PK. Influence of aggressive ions on the degradation behavior of biomedical magnesium alloy in physiological environment. Acta Biomater 2008;4(6): [42] Gu X, Zheng Y, Cheng Y, Zhong S, Xi T. In vitro corrosion and biocompatibility of binary magnesium alloys. Biomaterials 2009;30: [43] Rettig R, Virtanen S. Time-dependent electrochemical characterization of the corrosion of a magnesium rare-earth alloy in simulated body fluids. J Biomed Mater Res A 2007;85(1): [44] Lévesque J, Hermawan G, Dubé D, Mantovani D. Design of a pseudophysiological test bench specific to the development of biodegradable metallic biomaterials. Acta Biomater 2008;4: [45] Eliezer A, Abramov E, Gutman EM. Mechanochemical effect on Mg-alloys. J Mater Sci Lett 1998;17: [46] Sajuri ZB, Miyashita Y, Mutoh Y. Effects of humidity and temperature on the fatigue behavior of an extruded AZ61 magnesium alloy. Fatigue Fract Eng Mater Struct 2005;28: [47] Fontana MG. Corrosion engineering. 3rd ed. New York: McGraw-Hill; [48] Hoeppner DW. Model for prediction of fatigue lives based upon a pitting corrosion fatigue process. Fatigue Mechanisms, Philadelphia, PA: ASTM STP; [49] Miyashita Y, Sajuri Z, Mutoh Y, Kamado S. Fundamental study on high cycle fatigue behavior of magnesium alloy in order to establish a standard testing method. In: The 3rd Asian symposium on magnesium alloys, Shenyang, Liaoning, China; [50] 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(5 6): [51] Kimura Y, Takubo S. Corrosion fatigue of bio-ceramic sapphire in isotonic sodium chloride solution. Int J Fatigue 2000;22: [52] Ninomi M, Kobayashi T, Toriyama O, Kawakami N, Ishida Y, Matsuyama Y. Fracture characteristics, microstructure, and tissue reaction of Ti 5Al 5Fe for orthopedic surgery. Metall Mater Trans A 1996;27A: [53] Okazaki Y, Rao S, Ito Y, Tateishi T. Corrosion resistance, mechanical properties, corrosion fatigue strength and cytocompatibility of new Ti alloys without Al and V. Biomaterials 1998;19: [54] Giordani EJ, Guimarães VA, Pinto TB, Ferreira I. Effect of precipitates on the corrosion-fatigue crack initiation of ISO stainless steel biomaterial. Int J Fatigue 2004;26: [55] Zhao L, Burguera EF, Xu HHK, Amin N, Ryou H, Arola DD. Fatigue and human umbilical cord stem cell seeding characteristics of calcium phosphate chitosan-biodegradable fiber scaffolds. Biomaterials 2010;31: [56] Kurtz SM, Villarraga ML, Zhao K, Edidin AA. Static and fatigue mechanical behavior of bone cement with elevated barium sulfate content for treatment of vertebral compression fractures. Biomaterials 2005;26: [57] Sauer WL, Weaver KD, Beals NB. Fatigue performance of ultra-high-molecularweight polyethylene: effect of gamma radiation sterilization. Biomaterials 1996;17: [58] Gilbert JL, Ney DS, Lautenschlager EP. Self-reinforced composite poly(methyl methacrylate): static and fatigue properties. Biomaterials 1995;16: [59] Taylor D. Fatigue of bone and bones: an analysis based on stressed volume. J Orthop Res 1998;16(2):163 9.