Wenbo Du,Ke Liu,Ke Ma,Zhaohui Wang,Shubo Li

College of Materials Science and Engineering,Beijing University of Technology,Beijing 100124,PR China

1.Introduction

Magnesium and its alloys are attractive as a promising biodegradable material for orthopedic implants and cardiovascular interventional devices[1-3].The specific advantages of magnesium and its alloys for their use in medical applications come from the following aspects.First,Mg as a fundamental element of human body is involved in many metabolic functions as it naturally exists in bone tissue[4].Second,the elastic modulus of magnesium is～40-45GPa,which is very close to that of the human bone(10-20GPa).Thus,the magnesium alloy bone implant would not lead to"stress-shield"phenomena,distinguishing itself from the traditional metallic implants such as stainless steel and titanium alloys[5].Meanwhile,magnesium alloy can not only promote osteogenesis through its degraded Mg ions to stimulate new bone formation but also inhibit bacterial infection,thereby improving the success rate of surgery[6,7].Third,lots of in vivo investigations have confirmed that magnesium alloy implants have good biocompatibility[8-13].However,one of the major hurdles for practical clinical application for magnesium alloys is their uncontrolled corrosion rate.It is well known that magnesium based implants degraded very rapidly in a physiological environment,which leads to a failure of the implants before the recovery of the body via new tissue formation adequately[14,15].It is a great challenge to modify the degradation of the magnesium-based implants in the biological environment.

At present,lots of approaches have been adopted to control the corrosion rate,such as thermal mechanical processing,surface coating and alloying[16-20].Especially,alloying with other elements is one of the most effective methods to improve the corrosion resistance as well as mechanical properties of magnesium alloys.Commercial magnesium alloys containing Al and rare earth(RE)elements have shown good corrosion resistance and relatively high tensile strength at room temperature.However,they are not suitable for biomedical application because Al and RE elements are harmful to health of human beings.Al will cause nerve toxicity and restrain the growth of human body while the RE elements may increase the sick of the cell proliferation and thrombosis after implantation[21,22].Also,some RE elements such as Nd and Y in the WE43 alloy have been reported to cause a disturbance at implantation sites[23].Thus,it is very important to produce a new type of magnesium alloys,which contains beneficial elements to health of human being in order to escape of the toxicity to human body.

Many efforts have been made to improve biocompatibility and biodegradability of magnesium alloys.It is well known that the zinc is one of the essential elements for many physiological functions of human body,and it has been investigated for developing biodegradable materials[24-27].However,the high degradation of Mg-Zn system alloys,which is mainly ascribed to the galvanic effect of MgxZnysecondary phases,will result in disability in biomedical application as implants[25].Addition of manganese can improve the corrosion resistance of magnesium alloys by removing iron as well as other heavy-metal elements into the relatively harmless intermetallic compounds[28].Previous research indicated that the Mg-Zn-Mn alloy was a promising material for clinical use,and other non-harmful elements have been added in order to tailor its property.The Ca as an indispensable element of bone is able to enhance both the corrosion resistance and mechanical properties of magnesium alloy[29,30].Meanwhile,the Sn is also tolerable and essential to human body as well as improving the corrosion resistance[31].

However,the effects of Ca and Sn addition on Mg-Zn-Mn alloy have rarely been investigated so far.In this paper,Ca and Sn were added into the Mg-4Zn-0.2Mn(wt.%)alloy,respectively,in order to produce a new type of magnesium alloys for biomedical applications.The corrosion behaviors and electrochemical responses of the as-cast Mg-4Zn-0.2Mn(wt.%,defined as alloy A),Mg-4Zn-0.2Mn-0.2Ca(wt.%,defined as alloy B)and Mg-4Zn-0.2Mn-0.2Sn(wt.%,defined as alloy C)alloys were investigated in detail.

2.Materials and methods

2.1.Materials and preparation

Pure magnesium ingot(＞99.95wt.%),pure zinc strip(＞98.5wt.%),pure tin particles(＞99.98wt.%),Mg-5wt.%Mn and Mg-9.4wt.%Ca master alloys were used to prepare magnesium alloys.The melting was conducted in an electronic resistance furnace under a mixed atmosphere of N2and SF6in a graphite crucible.The melting alloy liquid was stirred for 45min at 760°C to ensure a homogeneous composition.Then the molten mixture was casted into a steel mold at～720 °C.

2.2.Characterization,microstructure observation and mechanical properties testing

Chemical compositions of the alloys were analyzed by using X-ray fluorescence(XRF,Magix-PW2403),as listed in Table 1.The specimens for microstructure characterization and in vitro corrosion measurements were cut from the alloys ingots with a geometric size of 10mm×10mm×10mm.Microstructure was observed under optical microscope(OM,AXIO IMAGER A2m)and scanning electron microscope(SEM,HITACHI S3400N)equipped with energy dispersivespectrum(EDS).The grain size was measured according to the linear intercept method described in the ASTM standard E112-G6.The phase constituents of the investigated alloys were analyzed by using X-ray diffraction(XRD,D/MAX-3C)with Cu Kαradiation.Diffraction patterns were generated within the scanning range 10-90°with a step size of 0.02°.All the specimens for the microscope observation were polished and then etched with a solution of 5mL nitric acid+95mL ethyl alcohol.Mechanical properties were tested on a uniaxial tensile testing machine(DNS-20)under a strain rate of 1mm/s at room temperature.The specimens for the tensile test were made into dog-bone shape with 5mm in gauge diameter and 25mm in gauge length.

Table 1 Chemical compositions of as-cast alloys in wt.%.

2.3.Immersion test

The specimens with a size of 10mm×10mm×10mm for immersion test were polished with 200-2000 grit SiC papers,cleaned in acetone and ethanol solution,and then dried at room temperature.The specimens were immersed into Hank’s physiological solution(8.00g/L NaCl,0.40g/L KCl,0.14g/L CaCl2,0.35g/L NaHCO3,0.1g/L MgCl2·6H2O,0.06g/L MgSO4·7H2O,0.06g/L KH2PO4and 0.06g/L Na2HPO4·12H2O[32]).The ratio of surface area to solution volume was 1 cm2/150mL.The pH value of the Hank’s solution was adjusted to 7.40 by using the NaOH solution or the HCl solution before experiments,and the temperature was kept at 37±1 °C during experiments.After immersion periods,the specimens were removed from Hank’s solution,cleaned in acetone and ethanol solution,and then dried at room temperature.The surface morphology of specimens after corrosion was observed by using SEM.The corrosion products were identified by EDS and then removed into a chromic acid solution of 200g/L Cr2O3+10g/L AgNO3.

According to the ASTM-G31-72 standard,the immersion corrosion rate was calculated by using Eq.(1)[33]:

whereVis the corrosion rate (mm/year);coefficientK=8.76×104;W0andWtare the weight before immersion and after cleaning the corrosion products,respectively;Ais the surface area exposed to the solution(cm2);Tis the exposure time(h)andDis the density of the material(g/cm3).

2.4.Electrochemical test

Fig.1.X-ray diffraction patterns of as-cast Mg-4Zn-0.2Mn,Mg-4Zn-0.2Mn-0.2Ca and Mg-4Zn-0.2Mn-0.2Sn alloys.

Electrochemical measurements were performed on an electrochemical workstation(PARSTAT 2273)with a threeelectrode system containing a platinum wire as a counter electrode,a saturated calomel electrode(SCE)as a reference electrode and the sample with an exposed area of 1 cm2as a working electrode.The test was conducted at 37°C in a glass cell containing 150mL Hank’s solution with a pH value of～7.4.The open circuit potential(Eocp)of each sample was tested at first.Then the potentiodynamic polarization test was conducted at a constant scanning rate of 0.5mV/s within-1.85 VSCE～-1.2 VSCE.In order to obtain the representative results,each experiment was repeated three times.The current densities(icorr)and corrosion potentials(Ecorr)were obtained through graphical Tafel analysis[34,35].The relationship between icorr(mA/cm2)and the electrochemical corrosion rate(mm/year)is described as Eq.(2)[35]:

3.Results

3.1.Microstructure and mechanical properties of as-cast alloys

Fig.1 shows the XRD patterns of the as-cast alloys A(Mg-4Zn-0.2Mn),B(Mg-4Zn-0.2Mn-0.2Ca)and C(Mg-4Zn-0.2Mn-0.2Sn).It was indicated that the majority of X-ray reflections were identified as the complete spectrum of magnesium solid solution in the three alloys.The secondary phase constitutes were related with the alloying elements closely.The secondary phases were Mg4Zn7and Mg7Zn3because of the addition of the zinc.However,the peak corresponding to the Mg-Zn-Ca phase was observed as the Ca element was added into the Mg-4Zn-0.2Mn alloy.It was reported that the type of the secondary phase was related with the Zn/Ca atomic ratio in Mg-Zn-Ca based alloys,i.e.the atomic ratio of Zn/Ca more than 1.25 generally leading to the phase consisting ofα-Mg solid solution and Mg-Zn-Ca phase[36-38].Meanwhile,the addition of 0.2wt.%Sn(alloy C)would not result in some new precipitations,and the secondary phases in the alloy C were the same with the alloy A,i.e.Mg4Zn7and Mg7Zn3phase.It was because that the maximum equilibrium solubility of Sn in magnesium matrix is 14.5wt.%at eutectic temperature,indicating a complete dissolution of Sn into the magnesium matrix in alloy C[39].

The optical micrographs of as-cast alloys A,B and C are displayed in Fig.2.All the samples exhibited a typical dendritic microstructure,which were composed ofα-Mg dendrites and intermetallic compounds at interdendritic regions.In alloy A,two types of secondary phases were found,as shown in Fig.2a.One was the Mg7Zn3phase with island-shape morphology distributing at dendrite boundaries,the other was Mg4Zn7phase with globular-shape morphology presenting at magnesium matrix.However,as for alloy B the addition of Ca led to the secondary phases transforming from both the Mg7Zn3phase and the Mg4Zn7phase into the Mg-Zn-Ca phase.The newly generated Mg-Zn-Ca phase with thin strip-shape morphology distributed along grain boundaries.For alloy C,the addition of Sn into the Mg-4Zn-0.2Mn alloy resulted in the segregation of solute elements obviously,but the type and distribution of the secondary phases were the same with the alloy A.

Fig.2.The optical micrographs of as-cast Mg-Zn-Mn alloys:(a)(d)Mg-4Zn-0.2Mn,(b)(e)Mg-4Zn-0.2Mn-0.2Ca,(c)-(f)Mg-4Zn-0.2Mn-0.2Sn.

Table 2 Mechanical properties of the as-cast alloys.

Fig.3 shows the SEM images of microstructure as well as EDS results of secondary phases in the three alloys.The EDS result of the island-shape phase “A”in the alloy A was Mg-25.25 at.%Zn,and the atomic ratio of Mg/Zn was～2.96,which was close to the composition of the Mg7Zn3phase.The EDS result of the strip-shape phase “B”in the alloy B was Mg-20.17 at.%Zn-8.21 at.%Ca,and the atomic ratio of Zn/Ca was～2.46.In addition,the globular-shape particle“C”was observed in the alloy C,as shown in Fig.3c,and its EDS result was Mg-26.44 at.%Zn-0.15 at.%Sn.The atomic ratio of Mg/Zn in the globular-shape particle“C”was～2.77.Although its EDS result did not agree well with the composition of the Mg4Zn7phase,the morphology of the globularshape particle“C”confirmed that it was the Mg4Zn7phase.As reported by Sirkin et al.[40]and Mingolo et al.[41,42],the Sn has no solubility in Mg4Zn7and Mg7Zn3phases.That is to say,the high content of Mg and trace content of Sn belonged to the magnesium matrix.

Mechanical properties of the three as-cast alloys A,B and C are summarized in Table 2.It can be concluded that the alloy A and C displayed normal mechanical properties.The YS,UTS and elongation values of the alloy A were 56MPa,177MPa and 12%,respectively.The tensile properties of the alloy C were similar to those of the alloy A,with values of 57MPa,176MPa and 12%,respectively.However,the addition of Ca led to obvious improvement in mechanical properties,especially for the YS.It can be seen that with the addition of 0.2%Ca in the as-cast Mg-4Zn-0.2Mn alloy,the YS obviously increased from 56MPa to 71MPa,while the UTS and the elongation indicated slightly increase,which could be attributed to the grain refinement after the addition of Ca.

3.2.Corrosion behaviors

Fig.4 shows SEM micrographs of as-cast alloys A,B and C after immersion in Hank’s solution for 10 days(Fig.4a,c and e)and 20 days(Fig.4b,d,f),respectively.It indicates that the surfaces of all the studied alloys were covered by corrosion film layer and white particles corrosion products after 10 days immersion.The corrosion layers of alloys A and C have been cracked and the surfaces were covered by layered corrosion products accompanying with many pits in different depths and sizes,as shown in Fig.4a and e.The existence of cracks and corrosion pits allowed easier contact between the solution and the matrix,accelerating the corrosion rate of matrix.However,there was no significant corrosion pits on the surface of the alloy B which formed a compact film layer.Meanwhile,some corroded areas with white substrate were found on the film layer of alloy B due to the local corrosion.The white corroded areas would eventually become corrosion pits similar to those of alloys A and C.After immersion for 20 days,some very thick layer of corrosion products generated on the surface of the alloys,as shown in Fig.4b,d and f.The corrosion film layer was more compact on the surface of alloy B,indicating better corrosion resistance.It behaved as a barrier,protecting the matrix from solution significantly,and therefore decreasing the corrosion reaction.However,a part of the corrosion layer has fallen off and a large area of matrix was exposed to the solution after immersion for 20 days in alloy C,as shown in Fig.4f.

Fig.3.SEM micrographs of as-cast(a)Mg-4Zn-0.2Mn,(b)Mg-4Zn-0.2Mn-0.2Ca and(c)Mg-4Zn-0.2Mn-0.2Sn alloy.

The composition of the corrosion products were analyzed by selecting a part of corroded areas of alloy A(the white frame in Fig.4a).Fig.5 displays the SEM images with elements distributions map.It was found that the corrosion layer contained calcium,phosphorus,oxygen,magnesium and carbon elements.The Ca and P elements distributed uniformly in the corrosion layer,suggesting a formation of hydroxyapatite(HA)on the surface of alloy A.The distribution of the Mg element was similar to that of O element,indicating the formation of the Mg(OH)2compound.

In order to further analyze the composition of corrosion products after immersion for 10 days,the XRD was carried out on products of alloys A,B and C,respectively,as shown in Fig.6.It was found that the peaks corresponding to Mg(OH)2and HA(Ca10(PO4)3(OH)2)compounds were present besides peaks corresponding to theα-Mg in all the three alloys.Generally,the diffraction of the peaks corresponding to the HA was higher in both the alloy A and the alloy C than that in the alloy B,and it stood for the considerable amount of HA formed on the surface of both alloy A and alloy C.Conversely,the intensity of peaks corresponding toα-Mg was the highest in all three alloys,which indicated the corrosion reaction in alloy B was relatively slighter compared with those in both alloys A and C.

Fig.5.SEM micrographs with the element map distributions obtained from alloy A(the area marked by a white frame in Fig.4a)in Hanks’solution after 10-days immersion.

Fig.6.X-ray diffraction patterns attained from the corrosion products of Mg-4Zn-0.2Mn,Mg-4Zn-0.2Mn-0.2Ca and Mg-4Zn-0.2Mn-0.2Sn alloys after 10-days immersion in Hank’s solution.

Fig.7 shows the surface morphology of the three alloys with removal of the corrosion products after immersion in Hank’s solution for 10 days.A number of deep pits around secondary phases(including Mg7Zn3phase at grain boundaries and Mg4Zn7phase at matrix)were found in alloy A,as shown by white arrows in Fig.7a and b.Meanwhile,few corrosion deep pits were characterized in alloy B,as shown in Fig.7c.Further characterization on alloy B indicated that some corrosion crevices(as shown by white arrows in Fig.7d)distributed along the grain boundary,where the location of Mg-Zn-Ca phase was responsible for the corrosion morphology.Generally,alloy B displayed a uniform corrosion.However,Fig.7e and f showed the surface morphology of alloy C,which contained numbers of pits over the whole surfaces,indicating the presence of the most severely localized corrosion compared with alloys A and B.

Fig.7.Scanning electron micrographs of corroded appearance after 10-days immersion and corrosion product removal for(a)(b)Mg-4Zn-0.2Mn,(c)(d)Mg-4Zn-0.2Mn-0.2Ca,(e)(f)Mg-4Zn-0.2Mn-0.2Sn.

To further investigate the surface morphology after longterm immersion in Hank’s solution,we conducted a 40 days experiment.The SEM images of the three alloys after removal of corrosion products are shown in Fig.8.On the whole,the localized corrosion displayed a tendency to be more severe as the immersion time extended.Therefore,a great quantity of deep holes were found on the substrate of the three alloys because of the severe localized corrosion.The diameter sizes of corrosion holes of alloys A,B and C were about in range of 65-385μm,32-112μm and 88-466μm,respectively.The presence of holes was detrimental to the overall corrosion resistance,and the destruction of magnesium alloys would proceed by means of pitting corrosion from one or more of them[43].The sizes of corrosion holes in alloy B were the most uniform,and their values were smaller than those of alloys A and C.Besides,the corrosion holes on the surface of alloy C were deeper than those of both alloy A and alloy B,as shown in Fig.8b,d and f.It confirmed that the corrosion rates sequence was alloy C＞alloy A＞alloy B.

Fig.8.Scanning electron micrographs of corroded appearance after 40-days immersion and corrosion product removal for(a),(b)Mg-4Zn-0.2Mn,(c),(d)Mg-4Zn-0.2Mn-0.2Ca,(e),(f)Mg-4Zn-0.2Mn-0.2Sn.

Fig.9 shows an average mass loss of the present three alloys in Hanks’solution for duration of 10,20 and 40 days,respectively.The corrosion rates from high to low were alloy C＞alloy A＞alloy B with the same immersion time(10 days,20 days and 40 days).Besides,the corrosion rate of the three studied alloys first increased and then decreased as the immersion time increased.This phenomenon was more obvious in alloy B,for which the corrosion rates were～0.27mm/year,0.68mm/year,0.31mm/year for immersion of 10,20,40 days,respectively,showing lower weight loss(0.31mm/year)compared with alloy A and alloy C after exposure to Hank’s solution for 40 days.However,the weight loss of alloy C wasabout two times higher than that of alloy A,which indicated a serious corrosion attack on alloy C.

Table 3 Electrochemical parameters of specimens in Hank’s solution obtained from the potentiodynamic polarization tests.

Fig.9.Corrosion rate obtained by weight loss of alloys in Hank’s solution for duration of 10,20 and 40 days.

3.3.Electrochemical corrosion measurements

Fig.10 shows the polarization curves of the three alloys in Hank’s solution.In general,the cathodic polarization curves represented the evolution of cathodic hydrogen as a result of water reduction,while the anodic polarization curves represented the dissolution of magnesium[44].There were no significant differences between the corrosion potentials(Ecorr)of the three alloys.Meanwhile,a passivation stage was found in the anodic polarization curves of all the alloys.The breakdown potential(Eb)was usually characterized by a sudden drop on the polarization curve,which indicated a localized corrosion tendency.A more positiveEbmeans less likely localized corrosion[45].TheEbat which pitting corrosion initiated was detected at-1.458V,-1.362V and-1.472V for alloys A,B and C,respectively.Therefore,the localized corrosion was more difficult to occur on alloy B.

Furthermore,the characteristics of the corrosion of all the alloys measured by polarization test were presented in Table 3.The corrosion current density(icorr)values were 3.074 μA/cm2,2.786 μA/cm2,4.517 μA/cm2for alloys A,B and C,respectively.The corrosion current density of alloy B was less than those of both alloy A and alloy C in cathodic region,indicating that the cathodic hydrogen evolution rate of alloy B was the slowest in all the alloys.The corrosion rate(Pi)of alloy B was 0.064mm/year,lower than those of alloy A(0.070mm/year)and alloy C(0.103mm/year),indicating that the Ca element has a considerable effect on reducing the corrosion rate of the Mg-4Zn-0.2Mn alloy.Thus,the polarization curve results indicated that the corrosion resistance of the three alloys reduced in the order of alloy B＞alloy A＞alloy C,and alloy B showed the best corrosion resistance while alloy C was the worst one.This conclusion was in good agreement with the immersion results.

4.Discussion

During immersion test in Hank’s solution,the multiple layers mainly consisted of O,C,P,Mg,and Ca elements.First,the Mg transformed into the stable Mg2+ion(Eq.(4)).At the same time,the cathodic reaction occurred because of the galvanic corrosion between matrix and secondary phase companying with the hydrogen evolution[46].The cathodic reaction led to a type of thin heterogeneous porous layer which was mainly composed of Mg(OH)2on the surface of magnesium alloys(Eq.(6)).It hindered contact of the solution with the substrate,leading to a decrease in corrosion rate[47].Meanwhile,the chloride(Cl-)in Hank’s solution penetrated the film easily,and reacted with Mg(OH)2compound(Eq.(7)).Thus,the Mg(OH)2compound transformed into more soluble MgCl2compound which was easier to be dissolved into Mg2+and Cl-[48].The dissolution compound resulted in a reduction of Mg(OH)2compound around protect areas,resulting in further dissolution of substrate.The corrosion phenomena could be described as the following reactions:

The main factor influencing the alloys corrosion resistance is the secondary phase,which usually acts as micro-cathodes,resulting in an acceleration of matrix corrosion around the secondary phase.Song et al.[25]have reported that the effects of MgxZnysecondary phases on corrosion resistance of Mg-xZn alloys.The magnesium matrix around MgxZnysecondary phases suffered from severe attack and displayed abundant of corrosion pits due to galvanic corrosion.Zhang et al.[47]investigated Mg-xZn-1Ca alloys and found that the selective corrosion occurred along the boundary of the magnesium substrate and the secondary phases due to different electrochemical behaviors ofα-Mg and precipitates phases.Bakhsheshi-Rad et al.[48]investigated Mg-0.5Ca-xZn alloys and found that the corrosion resistance of alloys increased due to a formation of Mg-Zn-Ca phases and Mg2Ca phases when the Zn/Ca ratio was less than 1.23.Zander and Zumdick[49]found that the Mg-Zn-Ca phase inhibited the local corrosion through acting as a temporary local corrosion barrier against the corrosion attack of matrix,but it was also a cathodic site for matrix phase as the immersion time extended.

Fig.10.Potentiodynamic polarization curves of Mg-4Zn-0.2Mn alloys in Hank’s solution.

Table 4 Comparison of the corrosion rates of Mg-base alloys.

Therefore,the Mg-4Zn-0.2Mn-0.2Ca alloy(alloy B)containing Mg-Zn-Ca phase displayed an excellent corrosionresistance behavior in the present investigation.Meanwhile,the Mg-4Zn-0.2Mn alloy(alloys A)and Mg-4Zn-0.2Mn-0.2Sn(alloy C)mainly including the Mg-Zn phases showed a worse corrosion-resistance behavior because of the galvanic corrosion.Some previously reported studies were summarized and compared in Table 4.From this table,it can be concluded that the corrosion rate(0.31mm/year)of alloy B was lower than other investigated alloys[47,50-54].The better corrosion-resistance properties of the as-cast Mg-4Zn-0.2Mn-0.2Ca alloy(alloy B)were reflected in the following aspects:(a)the corrosion layer was thinner,(b)the corrosion was uniform and(c)the pits were not obviously found.Meanwhile,the performance of corrosion-resistance properties in both Mg-4Zn-0.2Mn alloy(alloy A)and Mg-4Zn-0.2Mn-0.2Sn alloy(alloy C)were converse comparing with alloy B,as shown in Fig.11.The results of EDS showed that the content of elements in different positions of the alloy C is different,which is related to the corrosion process of the alloy.The corrosion mechanism of the alloy will be discussed latter.

Fig.11.Cross-sectional scanning electron micrographs of the corroded microstructure for(a)Mg-4Zn-0.2Mn,(b)Mg-4Zn-0.2Mn-0.2Ca,(c)Mg-4Zn-0.2Mn-0.2Sn in Hank’s solution after 10-days immersion at 37 °C.

Fig.12.Scanning electron micrographs of the corroded surface for(a)Mg-4Zn-0.2Mn-0.2Ca and(b)Mg-4Zn-0.2Mn-0.2Sn in Hank’s solution after 10-days immersion at 37°C.

In order to study the corrosion mechanism,the formation of surface layers and corrosion products was investigated for all studied alloys.The surface of the Mg-4Zn-0.2Mn-0.2Ca alloy(alloy B)was covered with a compact and insoluble corrosion film of the Mg(OH)2as well as small particles,as shown in Fig.12a.The corrosion morphology of the Mg-4Zn-0.2Mn-0.2Sn alloy(alloy C)was showed in Fig.12b.It was suggested that the corrosion product film exhibited a lot of cracks and white corrosion particles,which was similar to that of the Mg-4Zn-0.2Mn alloy(alloy A).As the corrosion reaction progressed,the corrosion pits extended outward along cracks and led to more serious corrosion after corrosion films being destroyed and shelled.Meanwhile,the pH value of the solution increased as the generation of.Subsequently,the phosphate ionsandin the solution reacted with,transforming into a formation of hydroxyapatite(HA)which nucleated and grew on the surface of the untreated specimen[55].

It was noteworthy that a trace addition of Ca resulted in a stable and uniform corrosion product layer on the surfaces of the specimen.It was suggested that the electrochemical potentials of the present phases are as follows:MgxZny＞Mg-Zn-Ca＞α-Mg.The addition of Ca led to the precipitation of an amount of Mg-Zn-Ca phase by substituting the MgxZnyphase.The stability of the corrosion product layer on the surface of magnesium alloys was strongly affected by the surface area ratio of magnesium matrix and secondary phases,which acted as anode and cathode,respectively[56].In other words,the corrosion rate of Mg-Zn-Ca alloys was related with the Mg-Zn-Ca intermetallic phase closely[48].Thus,the Mg(OH)2corrosion layer was very compact on the surface of the Mg-4Zn-0.2Mn-0.2Ca alloy(alloy B),suggesting a relative better corrosion resistance.The Mg-Zn-Ca intermetallic phase was mainly responsible for better corrosion resistance especially at the beginning stage.

5.Conclusions

(1)The microstructure analysis indicated that the addition of 0.2 wt.%Ca could refine grain size of the Mg-4Zn-0.2Mn alloy and led to precipitation of a large amount of Mg-Zn-Ca intermetallic phase along grain boundary.However,the addition of Sn had little effect on the secondary phase.

(2)The secondary phases had a significant effect on advancing the corrosion resistance of magnesium alloys.A large amounts of pits were observed on the surface of the Mg-4Zn-0.2Mn alloy and the Mg-4Zn-0.2Mn-0.2Sn alloy due to galvanic corrosion,on which the MgxZnysecondary phases acted as cathode and magnesium matrix as anode,accelerating the corrosion rate of alloys.The Mg-4Zn-0.2Mn-0.2Ca alloy(alloy B)exhibited a relative better corrosion resistance due to formation of the Mg-Zn-Ca intermetallic phase,which acted as a temporary local corrosion barrier against the corrosion attack of the matrix at the beginning stage of corrosion.

(3)Galvanic corrosion was also occurred in Mg-4Zn-0.2Mn-0.2Ca alloy at the late stages of corrosion.The Mg-Zn-Ca intermetallic phase acted as the cathode and magnesium matrix as anode at the interface between Mg-Zn-Ca and magnesium matrix.As a result,much magnesium matrix around Mg-Zn-Ca was corroded and the integrity of the corrosion protective film was destroyed.

(4)The addition of Ca into Mg-4Zn-0.2Mn alloy increased the corrosion resistance due to formation of Mg-Zn-Ca intermetallic phase.The protection performance of the Mg(OH)2was more compact and stable on the surface of Mg-4Zn-0.2Mn-0.2Ca alloy compared to the other two alloys.The Mg-4Zn-0.2Mn-0.2Ca alloy with good corrosion resistance is a suitable candidate for the development of biodegradable implant materials.

Acknowledgments

The project was supported by National Key Research and Development Program of China(2016YFB0301101,2016YFB0301001)and Beijing Natural Science Foundation(2172013).

[1]F.Witte,N.Hort,C.Vogt,S.Cohen,K.U.Kainer,R.Willumeit,F.Feyerabend,Curr.Opin.Solid State Mater.Sci.12(2008)63-72.

[2]H.Hornberger,S.Virtanen,A.R.Boccaccini,Acta Biomater.8(2012)2442-2455.

[3]F.Witte,Acta Biomater.23(2015)S28-S40.

[4]G.Song,Corros.Sci.49(2007)1696-1701.

[5]M.P.Staiger,A.M.Pietak,J.Huadmai,G.Dias,Biomaterials 27(2006)1728-1734.

[6]K.Jahn,H.Saito,H.Taipaleenmaki,A.Gasser,N.Hort,F.Feyerabend,H.Schluter,J.M.Rueger,W.Lehmann,R.Willumeit-Romer,E.Hesse,Acta Biomater.36(2016)350-360.

[7]Y.Zhang,J.Xu,Y.C.Ruan,M.K.Yu,M.O’Laughlin,H.Wise,D.Chen,L.Tian,D.Shi,J.Wang,S.Chen,J.Q.Feng,D.H.Chow,X.Xie,L.Zheng,L.Huang,S.Huang,K.Leung,N.Lu,L.Zhao,H.Li,D.Zhao,X.Guo,K.Chan,F.Witte,H.C.Chan,Y.Zheng,L.Qin,Nat.Med.22(2016)1160-1169.

[8]Y.Chen,Z.Xu,C.Smith,J.Sankar,Acta Biomater.10(2014)4561-4573.

[9]D.Zhao,S.Huang,F.Lu,B.Wang,L.Yang,L.Qin,K.Yang,Y.Li,W.Li,W.Wang,S.Tian,X.Zhang,W.Gao,Z.Wang,Y.Zhang,X.Xie,J.Wang,J.Li,Biomaterials 81(2016)84-92.

[10]J.W.Lee,H.S.Han,K.J.Han,J.Park,H.Jeon,M.R.Ok,H.K.Seok,J.P.Ahn,K.E.Lee,D.H.Lee,S.J.Yang,S.Y.Cho,P.R.Cha,H.Kwon,T.H.Nam,J.H.Han,H.J.Rho,K.S.Lee,Y.C.Kim,D,Proc.Natl.Acad.Sci.U S A 113(3)(2016)716-721.

[11]J.Wang,J.Xu,W.Liu,Y.Li,L.Qin,Sci.Rep.6(2016)26341.

[12]D.Tie,R.Guan,H.Liu,A.Cipriano,Y.Liu,Q.Wang,Y.Huang,N.Hort,Acta Biomater.29(2016)455-467.

[13]D.W.Zhao,F.Witte,F.Q.Lu,J.L.Wang,J.L.Li,L.Q,Biomaterials 112(2017)287-302.

[14]W.Zhou,T.Shen,N.N.Aung,Corros.Sci.52(2010)1035-1041.

[15]C.L.Liu,Y.J.Wang,R.C.Zeng,X.M.Zhang,W.J.Huang,P.K.Chu,Corros.Sci.52(2010)3341-3347.

[16]C.Liu,Y.Xin,G.Tang,P.K.Chu,Mater.Sci.Eng.A 456(2007)350-357.

[17]C.Lorenz,J.G.Brunner,P.Kollmannsberger,L.Jaafar,B.Fabry,S.Virtanen,Acta Biomater.5(2009)2783-2789.

[18]W.He,E.Zhang,K.Yang,Mater.Sci.Eng.C 30(2010)167-174.

[19]M.Yang,Y.Zhu,X.Liang,F.Pan,Mater.Sci.Eng.A 528(2011)1721-1726.

[20]X.Liu,J.Sun,K.Qiu,Y.Yang,Z.Pu,L.Li,Y.Zheng,J.Alloys Compd.664(2016)444-452.

[21]E.Zhang,L.Yang,Mater.Sci.Eng.A 497(2008)111-118.

[22]X.Gu,Y.Zheng,Y.Cheng,S.Zhong,T.Xi,Biomaterials 30(2009)484-498.

[23]F.Witte,V.Kaese,H.Haferkamp,E.Switzer,A.Meyer-Lindenberg,C.J.Wirth,H.Windhagen,Biomaterials 26(2005)3557-3563.

[24]S.Zhang,X.Zhang,C.Zhao,J.Li,Y.Song,C.Xie,H.Tao,Y.Zhang,Y.He,Y.Jiang,Y.Bian,Acta Biomater.6(2010)626-640.

[25]Y.Song,E.Han,D.Shan,C.D.Yim,B.S.You,Corros.Sci.65(2012)322-330.

[26]S.Cai,T.Lei,N.Li,F.Feng,Mater.Sci.Eng.C 32(2012)2570-2577.

[27]H.Li,Q.Peng,X.Li,K.Li,Z.Han,D.Fang,Mater.Des.58(2014)43-51.

[28]M.M.Avedesian,H.Baker,Workshops on Abstract State Machines,1999.

[29]N.Erdmann,N.Angrisani,J.Reifenrath,A.Lucas,F.Thorey,D.Bormann,A.Meyer-Lindenberg,Acta Biomater.7(2011)1421-1428.

[30]Z.Li,X.Gu,S.Lou,Y.Zheng,Biomaterials 29(2008)1329-1344.

[31]C.Zhao,F.Pan,S.Zhao,H.Pan,K.Song,A.Tang,Microstructure,corrosion behavior and cytotoxicity of biodegradable Mg-Sn implant alloys prepared by sub-rapid solidification,Mater.Sci.Eng.:C 54(2015)245-251.

[32]X.Zhang,G.Yuan,L.Mao,J.Niu,P.Fu,W.Ding,J.Mech.Behav.Biomed.7(2012)77-86.

[33]Standard Practice for Laboratory Immersion Corrosion Testing of Metals1,Annual Book of ASTM Standards,USA(2004).

[34]Z.Shi,M.Liu,A.Atrens,Corros.Sci.52(2010)579-588.

[35]Z.Shi,A.Atrens,Corros.Sci.53(2011)226-246.

[36]Y.Sun,B.Zhang,Y.Wang,L.Geng,X.Jiao,Mater.Des.34(2012)58-64.

[37]H.R.Bakhsheshi-Rad,M.H.Idris,M.R.Abdul-Kadir,A.Ourdjini,M.Medraj,M.Daroonparvar,E.Hamzah,Mater.Des.53(2014)283-292.

[38]Y.Xin,T.Hu,P.K.Chu,Acta Biomater.7(2011)1452-1459.

[39]S.Wei,T.Zhu,M.Hodgson,W.Gao,Mater.Sci.Eng.A 585(2013)139-148.

[40]H.Sirkin,N.Mingolo,E.Nassif,B.Arcondo,J.NonCryst.Solids 93(1987)323-330.

[41]N.Mingolo,E.Nassif,B.Arcondo,H.Sirkin,J.NonCryst.Solids 113(1989)161-166.

[42]M.Mingolo,B.Arcondo,E.Nassif,H.Sirkin,Z.Naturforsch.A:Phys.Sci.41(1986)1357-1360.

[43]R.Hahn,J.G.Brunner,J.Kunze,P.Schmuki,S.Virtanen,Electrochem.Commun.10(2008)288-292.

[44]H.R.B.Rad,M.H.Idris,M.R.A.Kadir,S.Farahany,Mater.Des.33(2012)88-97.

[45]J.Chang,P.Fu,X.Guo,L.Peng,W.Ding,Corros.Sci.49(2007)2612-2627.

[46]Y.Song,E.Han,D.Shan,C.D.Yim,B.S.You,Corros.Sci.60(2012)238-245.

[47]B.Zhang,Y.Hou,X.Wang,Y.Wang,L.Geng,Mater.Sci.Eng.C 31(2011)1667-1673.

[48]H.R.Bakhsheshi-Rad,M.R.Abdul-Kadir,M.H.Idris,S.Farahany,Corros.Sci.64(2012)184-197.

[49]D.Zander,N.A.Zumdick,Corros.Sci.93(2015)222-233.

[50]F.Witte,J.Fischer,J.Nellesen,H.Crostack,V.Kaese,A.Pisch,F.Beckmann,H.Windhagen,Biomaterials 27(2006)1013-1018.

[51]X.N.Gu,X.H.Xie,N.Li,Y.F.Zheng,L.Qin,Acta Biomater.8(2012)2360-2374.

[52]N.I.Zainal Abidin,B.Rolfe,H.Owen,J.Malisano,D.Martin,J.Hofstetter,P.J.Uggowitzer,A.Atrens,Corros.Sci.75(2013)354-366.

[53]D.Hong,P.Saha,D.Chou,B.Lee,B.E.Collins,Z.Tan,Z.Dong,P.N.Kumta,Acta Biomater.9(2013)8534-8547.

[54]Y.Wang,Z.Zhu,Y.He,Y.Jiang,J.Zhang,J.Niu,L.Mao,G.Yuan,Int.J.Mol.Med.29(2012)178-184.

[55]H.R.Bakhsheshi-Rad,M.H.Idris,M.R.Abdul-Kadir,A.Ourdjini,M.Medraj,M.Daroonparvar,E.Hamzah,Design 53(2014)283-292.

[56]N.T.Kirkland,N.Birbilis,J.Walker,T.Wood fi eld,G.J.Dias,M.P.Staiger,J.Biomed.Mater.Res.B Appl.Biomater.95B(2010)91-100.