Available online at www.sciencedirect.com

ScienceDirect

Journal of Magnesium and Alloys 2 (2014) 181e189www.elsevier.com/journals/journal-of-magnesium-and-alloys/2213-9567

Full length article

Microstructure, mechanical property and in vitro biocorrosion behavior ofsingle-phase biodegradable Mge1.5Zne0.6Zr alloy

Tao Li, Yong He, Hailong Zhang, Xitao Wang*

State Key Laboratory for Advanced Metals and Materials, University of Science and Technology Beijing, Beijing 100083, China

Received 5 May 2014; accepted 27 May 2014

Available online 30 June 2014

Abstract

The microstructure, mechanical property, and in vitro biocorrosion behavior of as-cast single-phase biodegradable Mge1.5Zne0.6Zr alloywere investigated and compared with a commercial as-cast AZ91D alloy. The results show that the Mge1.5Zne0.6Zr alloy had a single-phasesolid solution structure, with an average grain size of 34.7 ± 13.1 mm. The alloy exhibited ultimate tensile strength of 168 ± 2.0 MPa, yieldstrength of 83 ± 0.6 MPa, and elongation of 9.1 ± 0.6%. Immersion tests and electrochemical measurements reveal that the alloy displayed lowerbiocorrosion rate and more uniform corrosion mode than AZ91D in Hank's solution. The elimination of intensive galvanic corrosion reactionsand the formation of a much more compact and uniform corrosion film mainly account for the better biocorrosion properties of theMge1.5Zne0.6Zr alloy than AZ91D.Copyright 2014, National Engineering Research Center for Magnesium Alloys of China, Chongqing University. Production and hosting byElsevier B.V.

Keywords: Magnesium alloys; Biomaterials; Single-phase; Microstructure; Corrosion; Biodegradable

Open access under CC BY-NC-ND license.

1. Introduction

Recently, biodegradable magnesium based alloys areexperiencing a renaissance as potential choice for implantmaterials in bone and cardiovascular surgery due to theiroutstanding biocompatibility, biodegradability, and verysimilar mechanical properties compared with bone [1e5].Even though much attention has been paid to optimize theperformance of these materials [6e11], comparably high

* Corresponding author. Tel.: þ86 10 82375280; fax: þ86 10 62333447.

E-mail address: xtwang@ustb.edu.cn (X. Wang).

Peer review under responsibility of National Engineering Research Center for

Magnesium Alloys of China, Chongqing University.

Production and hosting by Elsevier

http://dx.doi.org/10.1016/j.jma.2014.05.006.

2213-9567/Copyright 2014, National Engineering Research Center for Magnesium Alloys of China, Cho

corrosion rate is still the biggest obstacle on the way to clinicalapplications [12]. On the other hand, at present the demand forbiodegradable materials with better performance is so urgent,while studies are still not sufficient [13].

In early investigations, a large amount of alloying compo-nents were added to magnesium alloys such as AZ91 [14e16],ZK60 [17e19], WE43 [15,20,21], LAE442 [21e23], andMge20Zne1Ca [24], with an intention of enhancing themechanical properties. These alloys generally contain manyprecipitates. In the multiphase magnesium alloys, however,potentials of second phases are higher than that of a-Mgmatrix [25]. So intensive galvanic corrosion reactions willoccur beside the second phases and many corrosion pits mayform, sometimes leading to the falling off of unattacked sec-ond phases from the alloy body. This will cause stress con-centration and the coming rapid reduction of mechanicalintegrity of implants during degradation. Huan et al. [26]demonstrate that ZK30 possesses a better corrosion

ngqing University. Production and hosting by Elsevier B.V. Open access under CC BY-NC-ND license.

182 T. Li et al. / Journal of Magnesium and Alloys 2 (2014) 181e189

resistance than ZK60 and they accredit that to the lower Zncontent in ZK30, since there is less second-phase particles andthus less micro-galvanic cells in ZK30. Rad et al. [27] alsoconclude that dissolution rate of MgexCa alloy increasessignificantly with increasing Mg2Ca content, and corrosiondamage in specimens with lower Ca content is more moderateand uniform than with higher Ca content. Based on the aboveconsideration, a single-phase structure may inhibit the highcorrosion rate of magnesium alloys to some extent.

Another consideration for implant biomaterials is thebiocompatibility. Aluminum element is mostly added tomagnesium alloys such as AZ91, AZ31, and AE21 to achieveboth solid solution strengthening and precipitation strength-ening. However, aluminum is a risk factor for causing Alz-heimer's disease [28] and muscle fiber damage [29]. Therefore,Al-free magnesium alloys are recommended for biodegradablematerials. In addition, investigations show that addition of rareearth elements can enhance the corrosion resistance of mag-nesium alloys [30e32]. Nevertheless, controversies are stillexist on the biological effects of those rare earth elements [2],and this process can dramatically increase the production costof alloys.

As Al-free magnesium alloys, high-strength MgeZneZrseries are widely studied. Researchers generally add a highcontent of Zn with 3%e6%, e.g., ZK30 [26], Mge5Zne0.3Zr[33], and ZK60 [17e19], which results in many second-phaseprecipitates and thus inevitable galvanic corrosions. Accordingto the above consideration, we have developed a single-phaseMge1.5Zne0.6Zr alloy to avoid intensive galvanic corrosionreactions [34]. The Mge1.5Zne0.6Zr alloy exhibits perfectcorrosion properties in 5% NaCl solution [34]. As a potentialdegradable biomaterial, it is strongly needed to characterizethe biocorrosion behavior of the Mge1.5Zne0.6Zr alloy insimulated body fluid. In the present study, the microstructure,mechanical property and especially in vitro biocorrosionbehavior in Hank's solution of the as-cast single-phasebiodegradable Mge1.5Zne0.6Zr alloy are investigated andcompared with a commercial AZ91D alloy.

2. Experimental

2.1. Sample preparation

Thealloywith the nominal compositionofMge1.5Zne0.6Zrwas prepared by the gravity casting process in a protective at-mosphere. The details can be referred to our former research[34]. The actual chemical compositions were determined byinductively coupled plasma atomic emission spectroscopy (ICP-AES, Varian 715-ES, USA), as given in Table 1. To verify the

Table 1

Chemical compositions of the as-cast Mge1.5Zne0.6Zr and AZ91D alloys

(wt.%).

Alloy Al Zn Zr Fe Ni Cu Mg

Mge1.5Zne0.6Zr e 1.25 0.55 �0.01 0.012 �0.01 Balance

AZ91D 8.99 0.61 e 0.003 0.001 0.003 Balance

effectiveness of eliminating the galvanic corrosion of the single-phase magnesium alloy, a representative double-phase as-castAZ91Dalloy (BeijingGuangling JinghuaScience&TechnologyCo., China) was used as a reference. The AZ91D alloy is chosenbecause it is one of the most commonly used magnesium alloysand is believed to have good corrosion resistance [35]. Also,AZ91D has been widely studied as a potential biodegradablematerial [14e16,36e38].

2.2. Microstructure characterization and mechanicalproperties evaluation

The microstructure of as-cast Mge1.5Zne0.6Zr andAZ91D alloys were observed using an optical microscope(OM) after prepared by a standard metallographic procedureand then followed by chemical etching in a solution of 12 gpicric acid, 80 ml acetic acid, 350 ml ethanol, and 80 ml water.The phase structure was detected by X-ray diffraction (XRD,Rigaku TTRIII, Japan), with a scan range from 10 to 100 and ascan speed of 8�/min. The XRD measurements were operatedat 40 kV and 150 mA, using Cu Ka radiation.

Tensile tests were carried out on an MTS810 universaltesting machine at a constant crosshead speed of 1.0 mm/minat room temperature. For tensile tests, specimens were cut byan electric-sparking wire-cutting machine from the as-castalloys and machined to a gauge size of F6 mm � 30 mm.Three specimens were tested for each alloy.

2.3. In vitro immersion tests

Immersion tests were conducted at 37 ± 0.5 �C in a Hank'ssolution, with a pH value of 7.4 adjusted with NaOH and HClsolutions. The Hank's solution was composed of 8.0 g/L NaCl,0.4 g/L KCl, 0.14 g/L CaCl2, 0.35 g/L NaHCO3, 1.0 g/Lglucose (C6H6O6), 0.1 g/L MgCl2$6H2O, 0.06 g/LMgSO4$7H2O, 0.06 g/L Na2HPO4$12H2O, and 0.06 g/LKH2PO4 in distilled water [7]. For in vitro immersion tests,specimens with a gauge size of 10 mm � 10 mm � 3 mm werecut from the as-cast alloys. A hole of 1 mm in diameter wasdrilled near one edge of each sample to accommodate a nylonstring to suspend it in the solution during immersion test. Inorder for immersion test, the specimens were polished using2000 grit SiC papers. The solution volume to specimen surfacearea ratio was 30 ml/cm2 according to ASTM G31-72 [39].The test lasted for 168 h and the immersion solution wasrenewed every 24 h. After immersion test, the corrosionproducts were removed from the specimens in a chromic acidsolution of 200 g/L CrO3 and 10 g/L AgNO3 according toASTM G1-90 [40]. The generation of hydrogen gas and thevariation of pH value in the Hank's solution were monitoredduring a continuous immersion test lasting for 168 h. Threespecimens were tested for each alloy at the same time. Thesurface morphologies of specimens before and after removingcorrosion products were examined by a field emission scan-ning electron microscope (FE-SEM, Zeiss Supra 55, Ger-many) equipped with an energy dispersive spectroscopy (EDS,Thermo 2247A, USA).

Table 2

183T. Li et al. / Journal of Magnesium and Alloys 2 (2014) 181e189

2.4. Electrochemical measurements

Mechanical properties of the as-cast Mge1.5Zne0.6Zr and AZ91D alloys.

Alloy UTS (MPa) YS (MPa) Elongation (%)

Mge1.5Zne0.6Zr 168 ± 2.0 83 ± 0.6 9.1 ± 0.6

AZ91D 151 ± 8.2 97 ± 5.9 2.7 ± 0.2

Electrochemical impedance spectroscopy (EIS) andpotentiodynamic polarization (PDP) curves were measured forthe alloys in the Hank's solution at 37 �C using an electro-chemical workstation (Chenhua CHI660D, China). For allmeasurements, a three-electrode electrochemical cell wasused, with a saturated calomel electrode (SCE) as the refer-ence electrode and a platinum electrode as the counter elec-trode. Prior to each measurement, the specimens weremechanically polished, followed by ultrasonic cleaning inabsolute alcohol for 5 min and drying in flowing air at roomtemperature. The EIS measurement started after open circuitpotential was stabilized for 30 min. The measurements wereperformed at open circuit potential with an AC amplitude of10 mV over a frequency range from 1 MHz to 100 mHz. ThePDP curves were subsequently measured with a scan rate of1 mV/s. The polarization started from a cathodic potential of�200 mV relative to the open circuit potential and stopped atan anodic potential where the anodic current increaseddramatically.

3. Results and discussion

3.1. Microstructure and mechanical properties

Typical microstructures and XRD patterns of the as-castAZ91D and Mge1.5Zne0.6Zr alloys are shown in Fig. 1.For AZ91D, dendritic and island-like b-Mg17Al12 phases [38]coexisted with a-Mg matrix (Fig. 1(a)). From the XRD patternof AZ91D in Fig. 1(c), the a-Mg and b-Mg17Al12 phases were

Fig. 1. OM observations and XRD patterns of the (a

identified. While for Mge1.5Zne0.6Zr, only one phase wasobserved and no precipitate was found (Fig. 1(b)). The singlephase was manifested as an equiaxed grain structure, with theaverage grain size of 34.7 ± 13.1 mm calculated by an Image-Pro Plus 6.0 software. Some twin boundaries were also iden-tified in the Mge1.5Zne0.6Zr alloy, owing to the mechanicalloading during sample preparation. Fig. 1(d) demonstrates theXRD pattern of the as-cast Mge1.5Zne0.6Zr alloy. Only Mgphase could be identified, as is consistent with the OMobservation in Fig. 1(b). The results confirm that the as-castMge1.5Zne0.6Zr belongs to a single-phase solid solutionalloy.

Table 2 summarizes the ultimate tensile strength (UTS),yield strength (YS), and elongation of the as-castMge1.5Zne0.6Zr and AZ91D alloys. It shows that theMge1.5Zne0.6Zr alloy has higher UTS and elongation valuesthan the AZ91D alloy. Since Zr serves as an effective grainrefining agent, the as-cast Mge1.5Zne0.6Zr alloy has a finegrain size of merely 34.7 ± 13.1 mm. The improved UTS andelongation values should be mainly ascribed to grain refine-ment strengthening. However, the Mge1.5Zne0.6Zr alloy hasa lower YS value than the AZ91D alloy, probably owing to theabsence of second phases. Second phases are likely to bring

, c) AZ91D and (b, d) Mge1.5Zne0.6Zr alloys.

184 T. Li et al. / Journal of Magnesium and Alloys 2 (2014) 181e189

about additional barriers to dislocation movement and grainboundary sliding at initial stage of deformation, resulting inthe enhancement of YS. Later, due to uncoordinated defor-mation with matrix, second phases can act as crack initiationsources and in turn decrease the UTS and elongation. For boneimplants, a high elongation value is desirable to prevent brittlefracture under complex stress conditions after implantingoperation.

3.2. Immersion tests

Fig. 3. Variation of pH value of Hank's solution as a function of immersion

time for the AZ91D and Mge1.5Zne0.6Zr alloys.

The generation of hydrogen gas of the Mge1.5Zne0.6Zrand AZ91D alloys over immersion time in Hank's solution isillustrated in Fig. 2. The generated hydrogen gas of theMge1.5Zne0.6Zr alloy remains a low level throughout thewhole immersion test period. For both alloys, hydrogen evo-lution started at a high rate and then climbed slowly. Nearlyfifty percent of the hydrogen was generated during the first24 h for the Mge1.5Zne0.6Zr alloy, whereas that was notmuch high for the AZ91D alloy. It is well known that, formagnesium alloys, the hydrogen evolution trend is directlyrelated to the corrosion rate. Thus, the results indicate that theMge1.5Zne0.6Zr alloy has a lower corrosion rate in Hank'ssolution than the AZ91D alloy. The corrosion rate is high atthe initial stage. As the corrosion process goes on, somecorrosion films are formed on the alloy surface. The corrosionfilms provide a protection for the alloy. So the corrosion rategets lower with immersion time. The protection is moreeffective for the Mge1.5Zne0.6Zr. It suggests that thecorrosion films formed on the Mge1.5Zne0.6Zr could bemore compact.

Fig. 3 shows the variation of pH value of Hank's solutionas a function of immersion time for the Mge1.5Zne0.6Zrand AZ91D alloys. As can be seen, the Mge1.5Zne0.6Zrhad lower pH values than the AZ91D throughout the im-mersion test period. The pH values for both alloys initially

Fig. 2. Generation of hydrogen gas of the AZ91D and Mge1.5Zne0.6Zr al-

loys as a function of immersion time in Hank's solution.

increased sharply with increasing immersion time, thenincreased very slowly, and finally reached a nearly steadystate. Similar to the hydrogen evolution trends, this revealsthat the Mge1.5Zne0.6Zr has a better corrosion resistance

Fig. 4. Surface morphologies of corrosion products on the (a) AZ91D and (b)

Mge1.5Zne0.6Zr alloys after immersion test in Hank's solution at 37 �C for

168 h.

Fig. 5. XRD patterns of the immersed (a) AZ91D and (b) Mge1.5Zne0.6Zralloys with corrosion products on the surfaces.

185T. Li et al. / Journal of Magnesium and Alloys 2 (2014) 181e189

than the AZ91D. It is reported that the increase in pH valueresults in haemolysis and is not appropriate for growth andproliferation of cells [27]. From this point of view, theMge1.5Zne0.6Zr shows a better biocompatibility.

Fig. 4 displays the typical surface morphologies of corro-sion products on the as-cast AZ91D and Mge1.5Zne0.6Zralloys after immersion test in Hank's solution at 37 �C for168 h. A large number of cracks were formed in the corrosionfilms. The cracks should be mainly generated by the dehy-drating process prior to the SEM characterization. They aremore easily generated in thicker coatings. Deep and widecracks were observed in the corrosion films grown on AZ91D,as shown in Fig. 4(a), while the cracks were shallow andnarrow for Mge1.5Zne0.6Zr (Fig. 4(b)). The result impliesthat the corrosion film formed on AZ91D is much thicker thanthat on Mge1.5Zne0.6Zr. This indicates that the corrosionrate of AZ91D is higher than that of Mge1.5Zne0.6Zr.Moreover, the corrosion film for Mge1.5Zne0.6Zr was muchmore compact and uniform than for AZ91D. The presence of aloose film makes corrosion solution contact with alloy matrixeasier and thus accelerates the corrosion of AZ91D. Forcomparison, the compact film on Mge1.5Zne0.6Zr is moreprotective. For magnesium alloys, the hydrogen evolution isdirectly related to corrosion process. The larger the corrosionrate, the larger the hydrogen evolution rate is. Intensehydrogen evolution will go against deposition of corrosionproducts so that the corrosion film is not compact, which inturn the corrosion is promoted. The EDS analyses of corrosionfilms of these two alloys in the white square areas are listed ontop of Fig. 4(a) and (b), respectively. The results reveal that thecorrosion films are rich in Ca and P, suggesting Ca/P com-pounds in the films. Furthermore, the content of Mg mainlycoming from alloy matrix was much higher forMge1.5Zne0.6Zr than for AZ91D. It means thatMge1.5Zne0.6Zr has a thinner corrosion film than AZ91D,which allows more Mg matrix information to be collected byEDS. The thinner corrosion film corresponds to the lowercorrosion rate of Mge1.5Zne0.6Zr, which is consistent withthe surface morphologies observation.

Fig. 5 shows the XRD patterns of the immersed AZ91D andMge1.5Zne0.6Zr alloys with corrosion products on the sur-faces. As can be seen from Fig. 5(a), for AZ91D alloy, thecorrosion products on the surface are mainly Mg(OH)2 withsome MgO and a slight amount of Ca10(PO4)6(OH)2 (HA).Since the b-Mg17Al12 phases are unattacked, the informationof b-Mg17Al12 phases is also collected. While forMge1.5Zne0.6Zr alloy (Fig. 5(b)), it is clear that the corro-sion products on the surface are mainly Mg(OH)2, MgO,(Ca,Mg)3(PO4)2, and Ca10(PO4)6(OH)2 (HA). The informationof Mg matrix was also collected by XRD due to the thin film.As well known, HA is an essential component of human bone.The deposited (Ca,Mg)3(PO4)2 and HA particles on theMge1.5Zne0.6Zr alloy substrate surface can support theattachment, differentiation, and proliferation of osteoblastsand mesenchymal cells. This fact is helpful in the healing ofbone tissue [41]. This indicates that the Mge1.5Zne0.6Zralloy has good biocompatibility.

Fig. 6 shows the surface morphologies of the AZ91D andMge1.5Zne0.6Zr alloyswith corrosion products removed afterimmersion test in Hank's solution at 37 �C for 168 h. The twoalloys displayed completely different corrosion morphologiesafter removing the corrosion products. For AZ91D (Fig. 6(a)),the b phases remained unattacked while the adjacent a-Mgmatrix seriously dissolved, resulting in many deep pits. Somelarge cavities were even formed due to the intensive galvaniccorrosion reactions. On the contrary, for Mge1.5Zne0.6Zr(Fig. 6(b)), the corrosion surface looked more uniform and thedepths of corroded areas were relatively shallow.

Fig. 6(c) and (d) show the magnified images of certain areasof the AZ91D and Mge1.5Zne0.6Zr alloys with corrosionproducts removed, respectively. As demonstrated in Fig. 6(c),the a-Mg matrix dissolved from the alloy surface to the center,leaving the b phase skeleton. The corrosion of AZ91D waslocalized and spread quickly to the depth, which had a sig-nificant effect on the mechanical properties. Comparatively,the corrosion of Mge1.5Zne0.6Zr was more uniform and

Fig. 6. Surface morphologies of the (a, c) AZ91D and (b, d) Mge1.5Zne0.6Zr alloys with corrosion products removed after immersion test in Hank's solution at

37 �C for 168 h.

186 T. Li et al. / Journal of Magnesium and Alloys 2 (2014) 181e189

mainly spread to the width (Fig. 6(d)). The localized corrosioncan cause stress concentration easily and rapid reduction ofmechanical integrity of implant materials during degradation.Ideally, the uniform corrosion is desired for implant materialsbecause the corrosion rate can be controlled to get bettersynchronization between implant degradation and new boneformation.

As shown in Table 3, the biocorrosion rate concluded bymassloss after immersion test in Hank's solution at 37 �C for 168 hwas characterized to be 0.3045 mm/y for Mge1.5Zne0.6Zr,41.9% of 0.7266 mm/y for AZ91D. The better corrosion resis-tance of Mge1.5Zne0.6Zr can be attributed to its single-phasestructure (Fig. 1) and the compact and uniform thin films on it(Fig. 4). For the Mge1.5Zne0.6Zr, the potential is distributedso uniformly that there is no extra impetus other than theaggressive solution. While for the AZ91D, intensive galvaniccorrosion reactions occur around the b phases and manycorrosion pits are formed, which extremely deteriorate thecorrosion behavior. In addition, the higher corrosion rate resultsin an intense hydrogen evolution that goes against the depositionof corrosion products. Thus the formed corrosion film is not socompact, and in turn the corrosion is promoted. By contrast, theMge1.5Zne0.6Zr does not possess the harmful galvanic

Table 3

Corrosion rates of Mge1.5Zne0.6Zr, AZ91D, ZK40, and ZK60 alloys.

Alloy State Solution Technique

Mge1.5Zne0.6Zr As-cast Hank's Mass loss

AZ91D As-cast Hank's Mass loss

ZK40 As-cast DMEM þ 10% FBS Mass loss

ZK60 As-extruded Hank's Potentiodyn

MAO-coated ZK60 As-extruded Hank's Potentiodyn

corrosion effect and holds a compact and uniform corrosionfilm.

3.3. Electrochemical measurements

The PDP curves of the as-cast AZ91D andMge1.5Zne0.6Zr alloys immersed in Hank's solution at 37 �Care shown in Fig. 7. For magnesium alloys, the cathodic po-larization curve generally represents the cathodic hydrogenevolution through water reduction, and the anodic curve rep-resents the dissolution of magnesium. As shown in Fig. 7, thecathodic curves display that Mge1.5Zne0.6Zr held a lowerhydrogen evolution current density than AZ91D. That is tosay, the overpotential of the cathodic hydrogen evolution re-action is higher on Mge1.5Zne0.6Zr than on AZ91D. Thisreveals that the cathodic reaction is kinetically harder onMge1.5Zne0.6Zr than on AZ91D, probably due to theelimination of galvanic effect in Mge1.5Zne0.6Zr. As can beseen from the anodic curves, both alloys showed an apparentcurrent plateau, indicating the existence of a protective film onthe alloy surface. However, the breakdown potential (Ebd) washigher for Mge1.5Zne0.6Zr, which manifests that the pro-tective film formed on Mge1.5Zne0.6Zr is more compact and

Duration Corrosion rate (mm/y) Refs.

7 days 0.3045 This work

7 days 0.7266 This work

7 days 0.39 [46]

amic polarization e 0.532 [47]

amic polarization e 0.003 [47]

Fig. 7. PDP curves of the as-cast AZ91D and Mge1.5Zne0.6Zr alloys

immersed in Hank's solution at 37 �C.

187T. Li et al. / Journal of Magnesium and Alloys 2 (2014) 181e189

protective. This finding is consistent with the surfacemorphology observations of corrosion products on the as-castAZ91D and Mge1.5Zne0.6Zr alloys after immersion test, asshown in Fig. 4.

Fig. 8 shows the Nyquist curves of the as-cast AZ91D andMge1.5Zne0.6Zr alloys immersed in Hank's solution at 37 �Cas well as an equivalent circuit model. The EIS results

Fig. 8. (a) Nyquist curves of the as-cast AZ91D and Mge1.5Zne0.6Zr alloys

immersed in Hank's solution at 37 �C and (b) the equivalent circuit model.

(Fig. 8(a)) reveal that both alloys displayed a single capacitiveloop except for the difference in diameter. The single capac-itive loop is also obtained for AZ31, WE43 [42], andMgeZneX (X ¼ Ca, Mn, Si) alloys [43]. This suggests thatthe corrosion mechanism of the two alloys is the same but thecorrosion rate is different. The larger diameter ofMge1.5Zne0.6Zr means better corrosion resistance, as is inagreement with the immersion test results (Table 3). Theimproved biocorrosion resistance should be related to thesingle-phase solid solution structure and fine microstructure ofthe Mge1.5Zne0.6Zr alloy. To analyze the EIS data, thesingle capacitive loop was modeled to a parallel combinationof a Rcf (corrosion film resistance) and a CPEcf (constant phaseelement of corrosion film) in series with a Rso (solutionresistance), as is shown in Fig. 8(b). The capacitor in equiv-alent electric circuit could be replaced by the constant phaseelement to improve simulation of the impedance thinking ofthe nonuniformity of surface and diffusion factors [11]. Thevalues of Rso, CPEcf, and Rcf were fitted through a ZViewsoftware and are listed in Table 4. Theoretically, the larger thecorrosion film resistance Rcf value is, the stronger the ability ofcorrosion film protects magnesium matrix. The much highervalue of Rcf for Mge1.5Zne0.6Zr implies that the corrosionfilm formed on it is more protective. The reason is that thecorrosion film formed on Mge1.5Zne0.6Zr is more uniformand compact than on AZ91D, as shown in Fig. 4.

MgeZneZr series magnesium alloys belong to a class ofhigh-strength magnesium alloy and they are suitable for loadbearing in clinical applications. Huan et al. [26] demonstratethat ZK30 possesses a better corrosion resistance than ZK60and they accredit that to the lower Zn content in ZK30, sinceZK30 has less second-phase particles and the correspondingmicro-galvanic cells. Song and Atrens [44] suggest that theaddition of Zn to magnesium alloys should be limited to 1%e3% if acceptable corrosion resistance is expected. Evenwithin the range of 1%e3% Zn, investigations reveal that anincrease in Zn content from 2% to 3% moves corrosion po-tential to a more negative value and the corrosion resistance isthus reduced [45]. Since the Mge1.5Zne0.6Zr alloy onlycontains 1.5% Zn and demonstrates a single-phase structure,theoretically, it should have a better corrosion resistance thanZK30 and ZK60. The comparison in Table 3 confirms that.The as-cast Mge1.5Zne0.6Zr alloy shows a lower corrosionrate than as-cast ZK40 [46] and as-extruded ZK60 [47] al-loys. Through alloy designing, researchers can harvest a newalloy with excellent properties. Besides, they can also modifythe existing alloys to get better performances. Surfacemodification is an effective way to controlling the degrada-tion of magnesium alloys, especially at the initial stage. Lin

Table 4

Rso, CPEcf, and Rcf values for the Mge1.5Zne0.6Zr and AZ91D alloys

immersed in Hank's solution at 37 �C.

Alloy Rso (U cm2) CPE-T CPE-P Rcf (U cm2)

Mge1.5Zne0.6Zr 5.28 2.91 � 10�5 0.665 11,376

AZ91D 9.02 2.32 � 10�5 0.757 5201

188 T. Li et al. / Journal of Magnesium and Alloys 2 (2014) 181e189

et al. [47,48] adopt a silicate electrolyte-based micro-arcoxidation (MAO) treatment on ZK60 alloy. The resultsindicate that the MAO coating effectively decreases the initialdegradation rate of the ZK60 alloy. To get better biocorrosionproperties, the MAO treatment could also be performed onthe as-cast Mge1.5Zne0.6Zr alloy. Furthermore, in order toconfirm the biocompatibility of the Mge1.5Zne0.6Zr alloy,in vitro and in vivo studies are needed and are left for futurestudy.

4. Conclusion

The microstructure, mechanical property, and in vitro bio-corrosion behavior of as-cast single-phase biodegradableMge1.5Zne0.6Zr alloy and commercial AZ91D alloy wereinvestigated. The Mge1.5Zne0.6Zr exhibited a single-phasecharacteristic and fine equiaxed grains of 34.7 ± 13.1 mm insize, while the AZ91D consisted of a-Mg matrix and b-Mg17Al12 second phases. The Mge1.5Zne0.6Zr had an ulti-mate tensile strength of 168 ± 2.0 MPa and an elongation of9.1 ± 0.6%, both higher than those of AZ91D. Due to theabsence of second phases, the Mge1.5Zne0.6Zr had loweryield strength of 83 ± 0.6 MPa than AZ91D. After immersedin Hank's solution at 37 �C for 168 h, a thinner and much morecompact and uniform corrosion film was formed on theMge1.5Zne0.6Zr. Most importantly, the Mge1.5Zne0.6Zrshowed uniform corrosion behavior during degradation pro-cess. Both the immersion tests and electrochemical measure-ments reveal that the Mge1.5Zne0.6Zr maintained betterbiocorrosion properties than AZ91D. The elimination ofintensive galvanic corrosion reactions and the formation ofcompact and uniform thin films are mainly responsible for thebetter biocorrosion properties of the Mge1.5Zne0.6Zr alloy.

Acknowledgments

This project was supported by the National Natural ScienceFoundation of China (No. 51174025) and SpecializedResearch Fund for the Doctoral Program of Higher Educationof China (No. 20110006130004).

References

[1] F. Witte, V. Kaese, H. Haferkamp, E. Switzer, A. Meyer-Lindenberg,

C.J. Wirth, H. Windhagen, Biomaterials 26 (2005) 3557e3563.[2] X. Gu, Y. Zheng, Y. Cheng, S. Zhong, T. Xi, Biomaterials 30 (2009)

484e498.

[3] N. Hort, Y. Huang, D. Fechner, M. St€ormer, C. Blawert, F. Witte, C. Vogt,

H. Drucker, R. Willumeit, K.U. Kainer, Acta Biomater. 6 (2010)

1714e1725.

[4] Y.F. Zheng, X.N. Gu, F. Witte, Mater. Sci. Eng. R 77 (2014) 1e34.

[5] L. Tan, X. Yu, P. Wan, K. Yang, J. Mater. Sci. Technol. 29 (2013)

503e513.

[6] L. Mao, G. Yuan, J. Niu, Y. Zong, W. Ding, Mater. Sci. Eng. C 33 (2013)

242e250.

[7] R.-C. Zeng, L. Sun, Y.-F. Zheng, H.-Z. Cui, E.-H. Han, Corros. Sci. 79

(2014) 69e82.

[8] W.R. Zhou, Y.F. Zheng, M.A. Leeflang, J. Zhou, Acta Biomater. 9 (2013)

8488e8498.

[9] A. Atrens, G.-L. Song, F. Cao, Z. Shi, P.K. Bowen, J. Magnesium Alloys

1 (2013) 177e200.

[10] K. Lietaert, L. Weber, J. Van Humbeeck, A. Mortensen, J. Luyten,

J. Schrooten, J. Magnesium Alloys 1 (2013) 303e311.

[11] B. Salami, A. Afshar, A. Mazaheri, J. Magnesium Alloys 2 (2014)

72e77.

[12] R. Zeng, W. Dietzel, F. Witte, N. Hort, C. Blawert, Adv. Eng. Mater. 10

(2008) B3eB14.[13] Y. Yun, Z. Dong, N. Lee, Y. Liu, D. Xue, X. Guo, J. Kuhlmann,

A. Doepke, H.B. Halsall, W. Heineman, S. Sundaramurthy, M.J. Schulz,

Z. Yin, V. Shanov, D. Hurd, P. Nagy, W. Li, C. Fox, Mater. Today 12

(2009) 22e32.

[14] W. Zhou, T. Shen, N.N. Aung, Corros. Sci. 52 (2010) 1035e1041.

[15] X.N. Gu, W.R. Zhou, Y.F. Zheng, Y. Cheng, S.C. Wei, S.P. Zhong,

T.F. Xi, L.J. Chen, Acta Biomater. 6 (2010) 4605e4613.

[16] D. Xue, Y. Yun, Z. Tan, Z. Dong, M.J. Schulz, J. Mater. Sci. Technol. 28

(2012) 261e267.

[17] X.N. Gu, N. Li, Y.F. Zheng, L. Ruan, Mater. Sci. Eng. B 176 (2011)

1778e1784.[18] Y.K. Pan, C.Z. Chen, D.G. Wang, T.G. Zhao, Colloids Surf. B 109 (2013)

1e9.

[19] Q. Zhang, X. Lin, Z. Qi, L. Tan, K. Yang, Z. Hu, Y. Wang, J. Mater. Sci.

Technol. 29 (2013) 539e544.[20] P. Gunde, A. Furrer, A.C. H€anzi, P. Schmutz, P.J. Uggowitzer, J. Biomed.

Mater. Res. A 92A (2010) 409e418.

[21] A. Krause, N. H€oh, D. Bormann, C. Krause, F.-W. Bach, H. Windhagen,

A. Meyer-Lindenberg, J. Mater. Sci. 45 (2009) 624e632.[22] F. Witte, J. Fischer, J. Nellesen, C. Vogt, J. Vogt, T. Donath,

F. Beckmann, Acta Biomater. 6 (2010) 1792e1799.

[23] W.-D. Mueller, M. Lucia Nascimento, M.F. Lorenzo de Mele, Acta

Biomater. 6 (2010) 1749e1755.

[24] Z. Xu, C. Smith, S. Chen, J. Sankar, Mater. Sci. Eng. B 176 (2011)

1660e1665.

[25] G. Song, A. Atrens, X. Wu, B. Zhang, Corros. Sci. 40 (1998)

1769e1791.

[26] Z.G. Huan, M.A. Leeflang, J. Zhou, L.E. Fratila-Apachitei, J. Duszczyk,

J. Mater. Sci. Mater. Med. 21 (2010) 2623e2635.

[27] H.R.B. Rad, M.H. Idris, M.R.A. Kadir, S. Farahany, Mater. Des. 33

(2012) 88e97.

[28] P.C. Ferreira, K.d.A. Piai, A.M.M. Takayanagui, S.I. Segura-Mu~noz, Rev.

Latino-Am. Enfermagem 16 (2008) 151e157.

[29] M. Shingde, J. Hughes, R. Boadle, E.J. Wills, R. Pamphlett, Med. J. Aust.

183 (2005) 145e146.

[30] X. Zhang, Y. Wu, Y. Xue, Z. Wang, L. Yang, Mater. Lett. 86 (2012)

42e45.[31] X.B. Zhang, G.Y. Yuan, Z.Z. Wang, Mater. Sci. Technol. 29 (2013)

111e116.

[32] A.C. H€anzi, I. Gerber, M. Schinhammer, J.F. L€offler, P.J. Uggowitzer,

Acta Biomater. 6 (2010) 1824e1833.[33] P. Gunde, A.C. H€anzi, A.S. Sologubenko, P.J. Uggowitzer, Mater. Sci.

Eng. A 528 (2011) 1047e1054.

[34] T. Li, H. Zhang, Y. He, X. Wang, Mater. Corros. http://dx.doi.org/10.

1002/maco.201307528.

[35] J.W. Chang, L.M. Peng, X.W. Guo, A. Atrens, P.H. Fu, W.J. Ding,

X.S. Wang, J. Appl. Electrochem. 38 (2008) 207e214.

[36] L. Choudhary, R.K.S. Raman, Eng. Fract. Mech. 103 (2013) 94e102.[37] J. Zhou, X. Zhang, Q. Li, Y. Liu, F. Chen, L. Li, J. Mater. Chem. B 1

(2013) 6213e6224.

[38] C. Del Gaudio, P. Bagal�a, M. Venturini, C. Grandi, P.P. Parnigotto,

A. Bianco, G. Montesperelli, J. Mater. Sci. Mater. Med. 23 (2012)

2553e2562.

[39] ASTM G31-72, Standard Practice for Laboratory Immersion Corrosion

Testing of Metals, American Society for Testing & Materials, West

Conshohocken, 2004.

[40] ASTM G1-90, Standard Practice for Preparing, Cleaning, and Evaluating

Corrosion Test Specimens, American Society for Testing & Materials,

West Conshohocken, 1999.

[41] Q.Z. Chen, C.H. Zhu, G.A. Thouas, Prog. Biomater. 1 (2012) 2.

189T. Li et al. / Journal of Magnesium and Alloys 2 (2014) 181e189

[42] X.Zhang,G.Yuan,L.Mao, J.Niu,W.Ding,Mater. Lett. 66 (2012)209e211.[43] F. Rosalbino, S. Negri, A. Saccone, E. Angelini, S. Delfino, J. Mater. Sci.

Mater. Med. 21 (2010) 1091e1098.

[44] G.L. Song, A. Atrens, Adv. Eng. Mater. 1 (1999) 11e33.

[45] E. Zhang, D. Yin, L. Xu, L. Yang, K. Yang, Mater. Sci. Eng. C 29 (2009)

987e993.

[46] D. Hong, P. Saha, D.-T. Chou, B. Lee, B.E. Collins, Z. Tan, Z. Dong,

P.N. Kumta, Acta Biomater. 9 (2013) 8534e8547.

[47] X. Lin, L. Tan, Q. Zhang, K. Yang, Z. Hu, J. Qiu, Y. Cai, Acta Biomater.

9 (2013) 8631e8642.

[48] X. Lin, L. Tan, Q. Wang, G. Zhang, B. Zhang, K. Yang, Mater. Sci. Eng.

C 33 (2013) 3881e3888.