J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 8 ( 2 0 1 2 ) 1 – 7

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/jmbbm

Research paper

The microstructure and properties of cyclic extrusioncompression treated Mg–Zn–Y–Nd alloy for vascularstent application

Qiong Wu, Shijie Zhu, Liguo Wang, Qian Liu, Gaochao Yue, Jun Wang, Shaokang Guan∗

Materials Research Centre, School of Materials Science and Engineering, Zhengzhou University, Zhenzhou, 450002, PR China

A R T I C L E I N F O

Article history:

Received 14 October 2011

Received in revised form

13 December 2011

Accepted 21 December 2011

Published online 3 January 2012

Keywords:

Mg–Zn–Y–Nd alloy

Cyclic extrusion compression

Mechanical property

Biocorrosion behavior

Biomedical applications

A B S T R A C T

Magnesium alloys are promising candidate materials for cardiovascular stents due to their

good biocompatibility and degradation properties in the human body. However, in vivo tests

also show that improvement in their mechanical properties and corrosion resistance is

necessary before wide application. In this study, cyclic extrusion compression (CEC) was

used to enhance the mechanical properties and corrosion resistance of Mg–Zn–Y–Nd alloy.

The results show that the grain size was greatly refined to 1 µm after CEC treatment.

The second phase distributed along the grain boundaries with grid shape and nano-sized

particles uniformly distributed in grains. The elongation (δ), ultimate tensile strength (UTS)

and yield strength (YS) of the CEC treatment samples were 30.2%, 303 MPa and 185 MPa

respectively. The CEC treated samples showed homogeneous corrosion because of the grain

refinement and the homogeneous distribution of nano-sized second phase. The corrosion

current density of the alloy decreased from 2.8×10−4 A/cm2 to 6.6×10−5 A/cm2 after CEC

treatment. Therefore, improved mechanical properties, uniform corrosion and reduced

corrosion rate could be achieved by CEC.c⃝ 2012 Elsevier Ltd. All rights reserved.

1

d

1. Introduction

Metallic materials including stainless steel, titanium alloys,and cobalt-based alloys have been used as materials forvascular stents due to their high strength, ductibility andgood corrosion resistance (Mani et al., 2007). However,these metallic materials are not biodegradable in thehuman body and may cause long term complications.Magnesium alloys have been attracting growing attentionin biodegradable stent materials because of their goodbiocompatibility and degradation properties. In vivo studieshave shown that magnesium alloys could be degraded in

∗ Corresponding author. Tel.: +86 371 67780051; fax: +86 371 6778005E-mail address: skguan@zzu.edu.cn (S. Guan).

1751-6161/$ - see front matter c⃝ 2012 Elsevier Ltd. All rights reservedoi:10.1016/j.jmbbm.2011.12.011

.

the human body and did not cause any serious adverseeffects. However, these studies also indicated that anearly loss in mechanical properties occurred in magnesiumimplants along with rapid degradation in the humanbody. Therefore, it is necessary to improve the mechanicalproperties and corrosion resistance. Several methods havebeen developed to improve the mechanical properties, whichinvolve purification (Song, 2007), microalloying (Häänzi et al.,2009a,b), fluoride conversion coatings (Chiu et al., 2007),alloying with other elements, and anodizing (Song, 2007;Wu et al., 2007; Zhang et al., 2005). Previous studies(Gao et al., 2011) have shown that Mg–Zn–Ca alloy after

.

2 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 8 ( 2 0 1 2 ) 1 – 7

high pressure torsion (HPT, one kind of severe plasticdeformation) treatment exhibited uniform corrosion due tothe homogeneous distribution of the nano-sized secondphase. However, HPT processed samples are usually used forchildren’s bone lamella, but not for vascular stent applicationbecause of their small size. Cyclic extrusion compression(CEC), another kind of severe plastic deformation, wasproposed by Richet in 1979. In CEC, an initially cylindricalsample with diameter D0 is pressed through two similarcylindrical channels with the same diameter connected viaa reduced cross section neck d. Therefore, each CEC passconsists of an extrusion followed by compression of thesample in the second channel to retain its initial dimensions.The CEC process can obtain ultra-fine grain and nano-particles uniformly distributed in grains. Recently, CEC wassuccessfully used to produce a variety of metallic materialswith ultra-fine grain structures in AZ31 (Chen et al., 2008)and ZK60 alloys (Lin et al., 2010). Peng et al. (2011) showedthat the yield strength and ductility of GW102K alloy werestrongly increased to 247 MPa and 22% by the CEC process at450 ◦C, the tensile strength was increased to 310 MPa. In thiswork, Mg–Zn–Y–Nd alloy with ultra-fine grain structures andnano-sized particle homogeneous distribution was preparedby CEC. The microstructure, mechanical properties andcorrosion behavior of the CEC treated magnesium alloy wereinvestigated. In order to clearly understand the corrosionbehavior of the CEC treated Mg–Zn–Y–Nd alloy, as-cast andextruded samples were also studied as control groups.

2. Experimental

2.1. Specimen preparation

A Mg–Zn–Y–Nd alloy ingot was prepared with high purity Mg,high purity Zn, Mg–25Y (wt%, 99.99% in purity) and Mg–25Nd(wt%, 99.99% in purity) master alloys through induction in amild steel crucible at approximately 740 ◦C under CO2/SF6(volume fraction rate, 3000 : 1) atmosphere in an electronicresistance furnace. A cast bar (Ø60 × 75 mm) cut from theingot was extruded at 543 K with an extrusion ratio of 25and then cooled in air. The samples (Ø30 mm × 55 mm)

used for CEC were machined from the ingot. The specimens(Ø8 mm × 5 mm) of as-cast and extrusion were prepared forelectrochemical tests and immersion tests. The CEC treatedsamples were rectangular specimens of 10 mm in length,5 mm in width and 4 mm in thickness.

2.2. Cyclic extrusion compression (CEC)

A schematic illustration of the CEC facility is shown in Fig. 1(a)(Lin et al., 2009). The number of extrusion passes was definedas the number of specimen passes through the cross sectionneck die. In the present study, the sample was extruded by 2passes. D0 and d were 30 mm and 18 mm, respectively. Thedie was lubricated using supramoly and preheated to 523 Kbefore processing. The extrusion temperature was 543 K andthe ram speed was 4 mm/s. During the final extrusion pass,the opposite ram B was removed in order to release the rod.The specimen of CEC 2 passes is shown in Fig. 1(b).

Fig. 1 – (a) Schematic illustration of the CEC facility andprocedure; (b) the specimen of CEC 2 passes.

2.3. Microstructural analysis and mechanical propertytest

The microstructure of the specimens was observed by opticalmicroscopy (Olympus, H2-UMA). The fracture morphologywas observed by scanning electron microscopy (Philips,Quanta-2000). The crystal structure was identified by X-raydiffraction (Philips, 1700X) with Cu-Kα radiation. The 2θ wascollected from 20◦ to 80◦ at a scan rate of 4◦/min. Thegrain size was measured according to the linear interceptmethod described in the ASTM standard E112-G6. The tensiletests (15 mm × 3.5 mm × 2 mm) were carried out on anInstron5585 test machine at room temperature with a strainrate of 1.0 mm min−1.

2.4. Electrochemical tests

The potentiodynamic polarization experiments were con-ducted on an electrochemical workstation (RST5200). Thescan rate of the potentiodynamic polarization experimentswas 0.5 mV/s. All of the specimens were connected to cop-per wire and embedded in epoxy resin with only one sideof 0.5 cm2 exposed for the test. The potentiodynamic polar-ization experiments were carried out in simulated body fluid(SBF) in a water bath at 37 ◦C. The composition of the SBF islisted in Table 1, and it was prepared based on the work ofKokubo and Takadama (2006).

2.5. Weight loss measurement

The immersion test was carried out in SBF at 37 ◦C andlasted for 5 days. The specimens were prepared in thesame way as those for the electrochemical tests. Eachprepared sample was placed in a sterilized bottle containing20 ml of SBF solution according to ASTM G31-72. Theimmersion solution was refreshed every 24 h in order tokeep a relatively stable pH value. After the immersiontest, the corrosion products were removed in chromicacid solution (200 g/LCr2O3 + 10 g/LAgNO3). Then thesamples were rinsed with distilled water and ethanol and

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 8 ( 2 0 1 2 ) 1 – 7 3

Table 1 – Ion concentration of SBF.

Ion Na+ K+ Ca2+ Mg2+ Cl− HCO−

3 SO2−

4 HPO2−

4

Concentration (mmol/L) 142.0 5.0 2.5 1.5 147.8 4.2 0.5 1.0

Fig. 2 – Microstructures of Mg–Zn–Y–Nd alloy : (a) as-cast; (b) hot extruded alloy at 543 K with an extrusion ratio of 25:1; (c)CEC 2 passes at 543 K.

Fig. 3 – The distribution of second phase: (a) hot extruded; (b) CEC 2 passes; (c) the nanoparticles of CEC 2 passes alloydistributed in grains.

then dried in warm flowing air. The dried samples wereweighed and the corrosion rate was calculated as follows(Liu et al., 2007):

CR = ∆W/At,

where CR is the corrosion rate; W is the weight loss; A is theoriginal surface area exposed to the test solution; and t is theexposure time.

3. Results and discussion

3.1. Microstructure observation

Fig. 2 shows the optical microstructures of as-cast, extrusionand CEC treated Mg–Zn–Y–Nd alloy. As shown in Fig. 2(a),the as-cast alloy was mainly composed of two different typesof phase: the matrix with coarse grains and some smallparticles distributed inside the crystal. The microstructure ofthe extruded alloy (Fig. 2(b)) was heterogeneous, the typicalstructure of incompletely recrystallized alloys. Coarse grainsof 10 ∼ 15 µm in size were surrounded by fine recrystallizedgrains of 5 ∼ 8 µm in size. The second phase distributed alongthe grain boundaries with rod shapes and some particlesgathered at the junction of three grains, as shown in Fig. 3(a).After 2 passes of CEC treatment, the microstructure ofthe alloy became finer and more homogenous due to the

dynamic recrystallization occurring during the deformation(Fig. 2(c)). The grain size was refined to about 1 µm andthe second phase distributed at the grain boundaries witha large grid shape, part of which is shown in Fig. 3(b).Some particles were brittle and formed into a cluster. Atthe same time some nano-particles which came from thecluster phase uniformly precipitated in grains, as shown inFig. 3(c). There are two possible reasons for the evolutionof microstructure after CEC treatment of alloy. Firstly, theelements of Y and Nd were difficult to diffuse because theiractivations at the extrusion temperature were low and therecrystallized microstructure would not obviously grow dueto plastic deformation occurring below the recrystallizationtemperature. Secondly, the grains and the second phase werebrittle under the pressure of extrusion and compressionduring the CEC process. So a refined and homogenousmicrostructure was obtained by the CEC technique.

3.2. Mechanical properties

The obtained values of UTS, average YS and δ are summarizedin Table 2. It shows that the UTS and YS of the Mg–Zn–Y–Ndalloy were improved significantly while the elongation wasnot improved apparently after hot extrusion. However, theelongation of CEC treated samples was about 2 times higherthan that of as-cast alloy and was about 1 time higherthan that of extruded alloy. This is related to the grain

4 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 8 ( 2 0 1 2 ) 1 – 7

Table 2 – The tensile properties of the samples.

Sample UTS (MPa) YS (MPa) Elongation (%)

As-cast 209 105 10.6Hot extrusion 316 183 15.6CEC 2 passes 303 185 30.2

Fig. 4 – The relationship of yield stress and the averagegrain size corresponded to Hall–Petch.

size, second phase distribution, dislocation density, andgrain boundary structure of the alloy. Firstly, the strengthclearly corresponded to the Hall–Petch relationship after theCEC process. A plot of strength versus square root of thegrain size is shown in Fig. 4. The microstructure of theCEC treated specimens was uniform and fine; meanwhilenano-sized particles precipitated in grains which improvedthe mechanical properties due to the grain refinementstrengthening and precipitation strengthening. Secondly, thefracture in magnesium and magnesium alloys generallyoriginates at deformation twins. But the formation of twinswas prevented by the CEC process due to texture changeand grain refinement, and hence increased activation of non-basal slip systems (Somekawa et al., 2006; Koike et al., 2003)and grain boundary sliding (Koike, 2004). We can concludethat the CEC process more strongly improved the ductility ofthe alloy than hot extrusion; simultaneously, the strength stillmaintained an ideal value.

Fig. 5 shows the fracture surfaces of the Mg–Zn–Y–Nd alloy.The as-cast alloy showed typical cleavage facets and cleavagesteps, while after hot extrusion and CEC treatment the alloyshowed many dimples typical of a ductile fracture. However,the CEC treated samples exhibited obviously uniformdeformation characteristics. In the dimples, second phaseparticles were observed; these particles can be regardedas reinforcements in the CEC specimens, by strengtheningfrom dislocation-particle interactions, load transfer toa lower E-modulus matrix and dislocation generation(Peng et al., 2009).

3.3. Electrochemical measurementsFig. 6 shows the corrosion potential curves (a) and thepolarization curves (b) of the specimens after immersion inSBF. It can be seen that the corrosion potential of the CECtreated alloy shifted to a more positive value than that of

Fig. 5 – Fracture surfaces of the samples: (a) as-cast; (b) hotextruded; (c) CEC 2 passes.

the as-cast and extruded ones, and the corrosion currentdensity of the alloy decreased from 2.8 × 10−4 A/cm2 to6.6 × 10−5 A/cm2. However, the current density of extrudedalloy has no obvious changes compared with as-cast. Onereason is that the grain size of the alloy was strongly refinedduring the CEC process, and the small grain size creates moregrain boundaries that act as a corrosion barrier to increasethe corrosion resistance. It has been reported that grainrefinement can improve a material’s corrosion resistance,resulting in decreased general corrosion rate and alleviationof corrosion localization (Alvarez-Lopez et al., 2010). Theother reason is the different distribution of the secondphase. As shown in Fig. 3(a), the second phase distributesalong the grain boundaries with a rod or blocky shape,which accelerates the corrosion pitting apparently due to thegalvanic corrosion between the cathodic compound and theanodic matrix. In contrast, the second phase of CEC treatedalloy presents cluster state and nano-particles, and at thesame time a large grid formed at the grain boundaries duringthe CEC process, which prevents corrosion from extendingand therefore increases the corrosion resistance.

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 8 ( 2 0 1 2 ) 1 – 7 5

Fig. 6 – The electrochemical measurements of different samples in SBF solution at 37 ◦C: (a) the corrosion potential ofsamples with time; (b) potentiodynamic polarization curves.

Fig. 7 – SEM images of samples after immersion in SBF solution at 37 ◦C for 48 h: (a) as-cast, (b) hot extruded, (c) CEC 2passes.

Fig. 8 – The macrography of samples immersed in SBF solution at 37 ◦C for 48 h.

3.4. Immersion testFig. 7 shows the morphology of the Mg–Zn–Y–Nd alloy afterimmersion in SBF for 48 h. As shown in Fig. 7(a) and(b), the as-cast and extruded alloy suffered from apparentpitting corrosion, a Ca–P film formed on the surface ofthe sample and some peeling-off locations were observed.The precipitation of Ca–P salts in SBF was a spontaneousprocess after a certain period of immersion. The dissolutionof Mg resulted in increase of pH, which was favorable tothe precipitation of Ca–P salts (Wang et al., 2008). Thus,the rapid pitting corrosion at the grain boundaries couldlead to a quick increase of the pH value in the vicinity of

the pitting area, which results in the deposition of Ca–Psalts. So the corrosion products of the as-cast and extrudedalloy were dark and many microcracks were observed. Afterconventional extrusion, second phase particles precipitatedat the grain boundaries as rod shapes or small blocks, andinternal stress formed at the grain boundaries. In Fig. 7(b),pitting corrosion occurred at the peeling-off locations andcracks were detected. The presence of cracks makes thesolution contact with the matrix easier and accelerates thecorrosion of the matrix.

However, the CEC treated alloys showed the trend ofuniform corrosion. A uniform layer was generated on the

6 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 8 ( 2 0 1 2 ) 1 – 7

Fig. 9 – XRD patterns of the alloys.

surface of the alloy. The corrosion cracks were fewer thanthose of the as-cast and conventional extruded alloys; fewercracksmean that the film layer is more compact which resultsin better corrosion resistance. The surface morphologies ofthe samples which have been removed corrosion products bychromic acid solution are shown in Fig. 8. Corrosion pittingswere observed apparently in the as-cast and extruded alloys,while the CEC treatment alloy shows a flat surface. Twofactors illustrate the uniform corrosion of CEC treated alloy.Firstly, as mentioned above, the second phase exhibits a gridshape along the grain boundaries, the nano-particles comingfrom the clusters distribute in grains uniformly during CEC.It was proved that the second phase was Mg12Nd by XRDmeasurement (Fig. 9). Chang et al. (2007) reported that thecorrosion potential of the Mg12Nd phase is only a little morepositive than that of pure Mg, so its negative influence oncorrosion caused by galvanic corrosion is much less thanthat of other second phases in Mg, such as Mg17Al12 inAZ series Mg alloys. Secondly, the results of electrochemicalmeasurements showed that the refined microstructure couldincrease the corrosion potential and decrease the currentdensity. When the corrosion potential tends to be positive, thealloys have high thermodynamic energy and low activity. Thereaction of alloys with SBF will be slow and uniform for the

homogeneous and fine microstructures. Therefore, the CECtreated sample exhibited uniform corrosionmainly due to thedistribution of the second phase and the grain refinement.

Fig. 10(a) illustrates the variation of pH value withimmersion time in SBF for 24 h. The pH values of CEC treatedsamples were lower than those of as-cast and extruded alloys.The pitting corrosion of the as-cast and extruded alloysgenerated more Mg2+ ions which increased the pH values.In contrast, the compact film layer on the CEC treated alloysurface makes the solution’s contact with the matrix moredifficult and decreases the corrosion of the matrix. Fig. 10(b)shows the corrosion rates of the samples. It can be seen thatthe corrosion rate of the as-cast alloy was 2 times higher thanthat of the CEC treated counterpart. This indicates that theCEC processing can improve the corrosion properties strongly.

4. Conclusion

Cyclic extrusion compression (CEC) has been used to un-derstand the influence on the microstructure and corrosionbehavior of Mg–Zn–Y–Nd alloy compared with as-cast andextrusion alloys. The following conclusions can be drawnfrom this investigation.

1. After CEC treatment and extrusion the grain size ofMg–Zn–Y–Nd alloy was refined strongly due to dynamicrecrystallization. The CEC treatment microstructure werefine and uniform, the average grain size was about 1 µmand a large number of nano-sized second phase particlesprecipitated in the grains. However, the microstructures ofthe extruded alloy were inhomogeneous and the secondphase distributed along the grain boundaries with rod-likeor small block shapes.

2. The mechanical properties of the alloy were significantlyimproved after CEC treatment. A UTS of 303 MPa wasachieved after 2 passes of CEC treatment, and the elon-gation of the alloy reached 30.2%, which can be ascribed tograin refinement and precipitation strengthening.

3. The CEC treated sample exhibited uniform corrosion dueto the grain refinement, the grid second phase distributed

Fig. 10 – (a) The pH values and (b) the corrosion rate curves of different samples with immersion time in SBF solutionat 37 ◦C.

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 8 ( 2 0 1 2 ) 1 – 7 7

at the grain boundaries and nano-sized particle distribu-tion in the grains. In contrast, the as-cast and extruded al-loy suffered from pitting corrosion. So the CEC processingis a promising candidate process for vascular stents.

Acknowledgments

We are grateful for the technical support of X.F. Guo and J.X.Zhang from Henan Polytechnic University and the support byNational Key Technology R&D Program (No. 2011BAE22B04-1-3) and by the National Natural Science Foundation of China(No. 51171174).

R E F E R E N C E S

Alvarez-Lopez, M., Pereda, M.D., Valle, J.A., Fernandez-Lorenzo, M., Garcia-Alonso, M.C., Ruano, O.A., Escudero,M.L., 2010. Corrosion behaviour of AZ31 magnesium alloywith different grain sizes in simulated biological fluids. ActaBiomater. 6, 1763–1771.

Chang, J.W., Fu, P.H., Guo, X.W., Peng, L.M., Ding, W.J., 2007. Theeffects of heat treatment and Zirconium on the corrosionbehavior of Mg–3Nd–0.2Zn–0.4Zr (wt%) alloy. Corros. Sci. 49,2612–2627.

Chen, Y.J., Wang, Q.D., Roven, H.J., Karlsen, M., Yu, Y.D., Liu, M.P.,Hjelen, J., 2008. Microstructure evolution in magnesium alloyAZ31 during cyclic extrusion compression. J. Alloys Compd.462, 192–200.

Chiu, K.Y., Wong, M.H., Cheng, F.T., Man, H.C., 2007. Characteri-zation and corrosion studies of fluoride conversion coating ondegradable Mg implants. Surf. Coat. Technol. 202, 590–598.

Gao, J.H., Guan, S.K., Ren, Z.W., Sun, Y.F., Zhu, S.J., Wang, B.,2011. Homogeneous corrosion of high pressure torsion treatedMg–Zn–Ca alloy in simulated body fluid. Mater. Lett. 65,691–693.

Häänzi, A.C., Dalla Torre, F.H., Sologubenko, A.S., Gunde, P.,Schmid-Fetzer, R., Kuehlein, M., Lööffler, J.F., Uggowitze,P.J., 2009a. Design strategy for microalloyed ultra-ductilemagnesium alloys. Philos. Mag. Lett. 89, 377–390.

Häänzi, A.C., Sologubenko, A.S., Uggowitzer, P.J., 2009b. Designstrategy for new bioabsorbable Mg–Y–Zn alloys for medicalapplications. Int. J. Mater. Res. 100, 1127–1136.

Koike, J., 2004. Dislocation plasticity and complementarydeformation mechanisms in polycrystalline Mg alloys. Mater.Sci. Forum. 449–452, 665–668.

Koike, J., Kobayashi, T., Mukai, T., Watanabe, H., Suzuki, M.,Maruyama, K., Higashi, K., 2003. The activity of non-basal slipsystems and dynamic recovery at room temperature in fine-grained AZ31B magnesium alloys. Acta Mater. 51, 2055–2065.

Kokubo, T., Takadama, H., 2006. How useful is SBF in predicting invivo bone bioactivity? Biomaterials 27, 2907–2915.

Lin, J.B., Wang, Q.D., Chen, Y.J., Liu, M.P., Roven, H.J., 2010.Microstructure and texture characteristics of ZK60 Mg alloyprocessed by cyclic extrusion and compression. Trans.Nonferrous Met. Soc. 20, 2081–2085.

Lin, J.B., Wang, Q.D., Peng, L.M., Roven, H.J., 2009. Microstructureand high tensile ductility of ZK60 magnesium alloy processedby cyclic extrusion and compression. J. Alloys Compd. 476,441–445.

Liu, C.L., Xin, Y.C., Tang, G.Y., Chu, P.K., 2007. Influence of heattreatment on degradation behavior of bio-degradable die-castAZ63magnesium alloy in simulated body fluid. Mater. Sci. Eng.A. 456, 350–357.

Mani, G., Feldman, M.D., Patel, D., Agrawal, C.M., 2007. Coronarystents: a materials perspective. Biomaterials 28, 1689–1710.

Peng, T., Wang, Q.D., Lin, J.B., 2009. Microstructure andmechanicalproperties of Mg–10Gd–2Y–0.5Zr alloy recycled by cyclicextrusion compression. Mater. Sci. Eng. A. 516, 23–30.

Peng, T., Wang, Q.D., Lin, J.B., Liu, M.P., Roven, H.J., 2011.Microstructure and enhanced mechanical properties of anMg–10Gd–2Y–0.5Zr alloy processed by cyclic extrusion andcompression. Mater. Sci. Eng. A. 528, 1143–1148.

Somekawa, H., Singh, A., Mukai, T., 2006. Deformation structureafter fracture-toughness test of Mg–Al–Zn alloys processedby equal-channel-angular extrusion. Phil. Mag. Lett. 86,195–204.

Song, G.L., 2007. Control of biodegradation of biocompatiblemagnesium alloys. Corros. Sci. 49, 1696–1701.

Wang, Y., Wei, M., Gao, J.C., Hu, J.Z., Zhang, Y., 2008. Corrosionprocess of pure magnesium in simulated body fluid. Mater.Lett. 62, 2181–2184.

Wu, C.S., Zhang, Z., Cao, F.H., Zhang, L.J., Zhang, J.Q., Cao, C.N.,2007. Study on the anodizing of AZ31 magnesium alloys inalkaline borate solutions. Appl. Surf. Sci. 253, 3893–3898.

Zhang, Y.J., Yan, C.W., Wang, F.H., Li, W.F., 2005. Electrochemicalbehaviour of anodized Mg alloy AZ91D in chloride containingaqueous solution. Corros. Sci. 47, 2816–2831.