Mehr Lotfpour ,Chngiz Dehghnin,? ,Mssoud Emmy ,Ahmd Bhmni,? ,Mehdi Mlekn ,Ahmd Sdti ,Mild Tghizdeh,Mohmmdrez Shokouhimehr

a School of Metallurgy and Materials Engineering,College of Engineering,University of Tehran,P.O.Box 11155-4563,Tehran,Iran

bDepartment of Mater.Sci.Eng.,Research Institute of Advanced Materials,Seoul National University,Seoul 08826,Republic of Korea

Abstract In the present study,a new biodegradable Mg-Zn-Cu magnesium alloy was introduced for biological applications.The microstructural analysis showed the formation of MgZnCu intermetallics for the Mg-2Zn-0.1Cu alloy and also the Mg(Zn,Cu)2 compounds for the Mg-2Zn alloys with higher Cu contents.Moreover,the hot extrusion was applied for the grain refinement and changing the distribution of intermetallics.In vitro immersion tests,electrochemical and corroded surface analyses represented the enhancement of corrosion resistance with 0.1wt.% Cu addition.Furthermore,the extruded alloys demonstrated more corrosion resistance behavior than that of the cast alloys.By considering the improved tensile properties of Mg-2Zn-0.1Cu alloy,this alloy was regarded as the potential candidate for use as the biodegradable magnesium implant.

Keywords: Mg alloys;Grain refinement;Second phases;Hot extrusion;Biodegradation behavior.

1.Introduction

Recently,magnesium and its alloys have been introduced as one of the potential candidates for use in biodegradable implants,owing to their similar stiffness and elastic modulus in regard to the natural bone and also having non-toxic ion for the human body [1–3].However,the major problem of these alloys is their unpredictability and rapid corrosion rate in the environments with high amounts of Cl?,leading to the loss of their mechanical integrity before the completion of the treatment process [2,4].To overcome this problem,many studies have been conducted using alloying elements[5,6],performing a plastic deformation processing [2,7–9]or applying surface coatings [10,11].Adding alloying elements may alter the nature of the passive film,change in the current density of cathodic or anodic reactions may create some new intermetallics which affects the corrosion mechanism [6,12].However,for the biological applications,alloying elements in Mg alloys must be selected with careful consideration to avoid negative consequences on human bodies [13].Zinc is an essential trace element in the human body,which its low contents showed suitable effects on mechanical and corrosion behavior of Mg alloys [2,6,13–17].Excessive Zn took down the immune functions,increased the corrosion rate and also created the cytotoxicity [17–20].Song et al.[16]also stated that more than 2wt.% of Zn additions accelerated the corrosion rate,due to the enhanced galvanic effect of the Mg-Zn intermetallics.Moreover,Gu et al.[20]demonstrated that the Mg-1wt.% Zn alloy had no negative effects on fibroblasts(L-929 and NIH3T3),osteoblasts(MC3T3-E1),and blood vessel-related cells.Furthermore,copper is another cost-effective element that its micro-additions could properly refine the microstructure,enhance the mechanical properties and also promote the age-hardening response of the Mgalloys [21–24].From the perspective of the corrosion properties,general beliefs are based on the fact that the Cu element is one of the impurities in Mg alloys which reduce the corrosion resistance [17,26].However,Cu not only is an important trace element for many enzymes in the human body,but also its ions can enhance the antibacterial activity and destroy the bacterial cells [25–27].Recent studies showed that Cu micro-additions could improve the degradation behavior of Mg alloys,with better antibacterial,biocompatibility and also mechanical properties,which makes it as a potential element for the biological applications [25–29].

On the other hand,the distribution of intermetallics and grain refinement during the thermo-mechanical processes can significantly affect the mechanical properties and corrosion resistance of Mg alloys [8,9,30–37].The grain boundaries have a great role in the alteration of the corrosion propagation and formation of the passive layer [32–37].Therefore,the formation of more grain boundaries during hot extrusion changes the corrosion resistance of Mg alloys.Moreover,adding new alloying element can introduce the new intermetallics to the system,which have different Volta-potential difference with theα-Mg phase and promote the galvanic corrosion [34].Hence,the re-distribution of intermetallics from that of the cast alloy to the scattered situation after hot deformation has a remarkable impact on the corrosion behavior [34].

The current study aims to investigate the effect of grain refinement and intermetallics by Cu additions and hot extrusion on the corrosion behavior of Mg-2Zn alloy and also find the optimum amounts of Cu which is suitable for biological applications.

2.Experimental

The nominal chemical compositions of the studied alloys are listed in Table 1.To prepare them,an induction furnace and graphite crucible were utilized to melt the pure Mg,Mg-50wt.%Zn and Mg-10wt.%Cu master alloys under the protection of an inert atmosphere with a CO2+SF6mixture.Then,the prepared molten alloys were poured into a pre-heated steel mold that was used in our previous study [23].

Table 1 Summary of the cast Mg-Zn-Cu alloys in the present study.

Table 2 The ions concentration of the SBF solution in the present study and its comparison with the ion concentrations of the blood plasma [38].

After that,the Z2 alloy was homogenized at 315 °C for 48h,and then,solutionized at 315 °C for 5 h followed by water quenching [24,30].Moreover,the Cu-containing alloys were homogenized and solutionized at 435 °C for 48h and 5 h,respectively followed by water quenching [24,30].Later,the prepared billets for the hot extrusion process were pre-heated at 300 °C for 1h in an electrical resistance fur-nace and extruded at the ram speed of 1mm/s and the ratio of 12:1.For the microstructural analysis,the polished specimens were etched by the acetic picral solution(10ml acetic acid+4.2g picric acid+10ml H2O+70ml ethanol)[8]and scrutinized by the optical microscopy(OM)and scanning electron microscopy(SEM)performed in FEI QUANTA 450 SEM equipped with Bruker QUANTAX XFlash 6 EDS detector.

The cylindrical samples with 3mm thickness and 6mm diameter were cut to perform the corrosion examination by using immersion,electrochemical and surface investigations tests in the simulated body fluid(SBF)at 36.5±1 °C.The chemical composition of the plasma blood and SBF solution are listed in Table 2[38].For the immersion tests,the samples were soaked for 336h(14 days)and the corrosion rate was calculated based on the measured hydrogen volume.Previous studies [8,39]described the setup of the experiment and the method for measuring the hydrogen level.Furthermore,the electrochemical tests were carried out after 30min,24h and 120h of immersion times.The electrochemical impedance spectrometry(EIS)analysis was conducted in the frequency range of 100 kHz to 0.1Hz.The simulations of Nyquist plots were done via the EC-Lab V10.40 software.The polarization test was done in the potential range of OCP±250mV at a scan rate of 0.2mV/s.The corrosion rate was calculated from the immersion and polarization techniques by using the equation for Mg alloys described by Cao et al.[40].Finally,the corroded surfaces were examined by the same SEM microscope.The corrosion products after the immersion test were scrutinized using an ESCALAB 250 X-ray photoelectron spectroscopy with the power of 150W,pass energy of 50.0 eV and a step size of 0.1eV.All energy values were corrected according to the adventitious C 1s signal at 284.6 eV.The data were analyzed with the CasaXPS software.

3.Results and discussions

3.1.Microstructure

Fig.1 shows the grain structures of the cast and extruded alloys.Adding Cu and performing the hot extrusion which significantly refined the grain structure of the base alloy.The average grain size values of the alloys are listed in Table 3.The grain size of the cast alloys was measured from 1/2r of the macrostructures.The cast Z2 alloy had a grain size of 914.27μm,which gradually decreased with more Cu additions.The average grain size of the cast ZC201 and ZC23 alloys were 49% and 72% finer than that of the base alloy,about 459.52μm and 247.46μm respectively.On the other hand,the dynamic recrystallization during hot extrusion was led to a reduction in grain size of the Z2 to be 13.25μm,declining 98% in comparison to that of the same cast alloy.Adding 0.1 and 3wt.% Cu reduced the average grain size to 9.53μm and 5.79μm respectively.

Fig.1.Optical micrographs of the cast and extruded Z2 alloys with:(a),(g)0,(b),(h)0.1,(c),(i)0.3,(d),(j)0.5,(e),(k)1 and(f),(l)3wt.% Cu additions.

Fig.2 represents SEM micrographs for the distribution of intermetallics in the cast and extruded alloys.As can be seen,the bright particles are dispersed among theα-Mg phase and their fractions were increased with more Cu additions.Moreover,the cast ZC21 and ZC23 alloys tended to form the semicontinuous networks of intermetallics,due to more constitutional undercooling at the interfaces of solid/liquid[23].Afterthe extrusion,the intermetallics were forced to scatter along the extrusion direction[24].The area fraction values are listed in Table 3 which indicated an increase from 0.06% for the cast Z2 alloy to 10.46% for the cast ZC23 alloy.After the extrusion process,the fraction of intermetallics slightly reduced due to their dissolution during the homogenization,solution heat treatment and also extrusion process.

Table 3 The values related to the average grain size and area fraction of new intermetallics of the cast and extruded alloys.

Table 4 The immersion data of the cast and extruded alloys after immersion for 336h in the SBF solution(PH is the corrosion rate measured by hydrogen evolution).

Fig.2.SEM micrographs of the cast and extruded Z2 alloys with:(a),(g)0,(b),(h)0.1,(c),(i)0.3,(d),(j)0.5,(e),(k)1 and(f),(l)3wt.% Cu additions.

The EDS map analysis in Figs.2(f)and 2(l)illustrate the formation of Mg-Zn-Cu-containing intermetallics for the cast and extruded ZC23 alloys.The tiny bright particles in the cast Z2 alloy(Figs.2(a)and 2(g))are the Mg-Zn intermetallics [23,24,30].By adding 0.1wt.% Cu,the small amounts of copper incorporated into the MgZn lava phase and formed the MgZnCu intermetallics [23].With higher Cu additions,the tendency for the formation of lamellar MgCu2lava intermetallics enhanced,in which the Zn elements incorporated into their crystal structures to create the Mg(Zn,Cu)2intermetallics [23].Therefore,the small bulky MgZnCu compounds are mainly observed in the ZC201 alloy(see in Fig.1(b)),while the fraction of lamellar Mg(Zn,Cu)2intermetallics improved by adding more than 0.1wt.% Cu(see in Figs.2(c),(d),(e),and(f)).

3.2.Corrosion behavior

3.2.1.Immersion tests

Fig.3 illustrates the amounts of evolved hydrogen and corrosion rate for the cast and extruded alloys after 336h of immersion in the SBF solution at 36.5 °C.At the first sight,the samples of the cast and extruded ZC205,ZC21,and ZC23 alloys were completely corroded before 336h,while the cast and extruded Z2,ZC201,and ZC203 alloys remained intact until 336h.The second point is related to the positive impact of extrusion on the corrosion behavior of ZC203,ZC205,ZC21,and ZC23 alloys and its neutral influence on the corrosion resistance of the Z2 and ZC201 alloys.Figs.3(e)and 3(f)depict the variations of pH values of the immersed alloys via the immersion times.As can be seen,the pH values increased with more immersion times,while the ZC205,ZC21,and ZC23 alloys had a remarkable pH increment and the Z2,ZC201,and ZC203 alloys had mild enhancement.

Fig.3.Immersion study of the cast and extruded alloys in SBF solution after 336h,(a),(b)volume of the evolved hydrogen and(c),(d)corrosion rate.

Fig.4.Comparison of the corrosion rates of the cast and extruded alloys after(a)24h and(b)336h immersion in the SBF solution.(c)Photographic images of the cast and extruded Z2,ZC201,and ZC203 samples after immersion for 336h.

Two immersion tests were performed and their related results are listed in Table 4.Moreover,the corrosion rates after 24h and 336h of immersion times are depicted in Fig.4 for the comparison.At the early stage of immersion(24h),the corrosion rate of cast alloys slowly increased by adding 0.1 and 0.3wt.% Cu,respectively.However,with 0.5,1.0 and 3.0wt.% Cu additions,the corrosion rate significantly increased,showing a more deteriorative effect of high Cu contents.On the other hand,the extrusion caused a decrease in the corrosion rate,comparing to that of the cast alloys.This effect was distinguished in the alloy with high Cu contents.At the long immersion times,the ZC201 alloy had the lowest corrosion rate among the others.Although the extrusion did not have a notable effect on the corrosion rate of Z2 and ZC201 alloys,the corrosion rates of the ZC203,ZC21,and ZC23 alloys remarkably decreased after hot extrusion.Fig.4(c)shows the optical micrographs of the corroded surfaces after immersion for 336h before removing the corrosion products.It can be seen that the surface of the cast and extruded ZC203 alloys are attacked more severe than that of the others.However,the cast and extruded Z2 and ZC201 alloys illustrated suitable surfaces with low amounts of corrosion pits.

Table 5 Values measured from the Nyquist curves of the cast and extruded Z2,ZC201,and ZC203 alloys after different immersion times.

Table 6 The data measured from the polarization curves of the cast and extruded Z2,ZC201,and ZC203 alloys after different immersion times.

Fig.5.The Corrosion rate measured by hydrogen evolution via(a),(b)D?0.5 and(c),(d)phase fraction,(e),(f)predicted corrosion rate.

Fig.6.Bode plots of |Z| and phase vs.Frequency,obtained from the EIS experiments after different immersion times for the cast and extruded alloys.

Fig.7.(a)-(f)Nyquist plots obtained from the EIS experiments after different immersion times for the cast and extruded alloys.(g)Equivalent circuit of the EIS spectra of the cast and extruded samples.

To analyze the corrosion rates quantitatively and discover the relationship between corrosion and microstructural parameters,the corrosion rate measured by hydrogen evolution after 48h immersion was compared with the grain size and fraction of the second phase.Figs.5(a)and(b)illustrate the corrosion rate values versus D?0.5for the cast and extruded alloys.As one can see,the slope of the fitted curves increased after 0.3wt.% Cu addition,showing the change in the dominating parameter with corrosion resistance.Several controversial behaviors have been reported for the effect of grain size on the corrosion rate.Song et al.[41]found out that the corrosion rate increased by decreasing grain size for AZ91 alloy and pure Mg [42],due to the addition of more lattice defects.However,many researchers found that the corrosion rate decreased with grain size decrement.Birbilis et al.[43]believed that the corrosion rate of ECAPed pure Mg was decreased,due to the better matching of the exposed surface layer with substrate and mismatch compensation regarding the formed dislocations.Han et al.[44]stated that grain boundaries function as a corrosion barrier against fine pits which need specific orientation to penetrate inside the grains.Bahmani et al.[34]summarized these statements and tried to integrate them with quantitative analysis and presented a new formula in this regard.They believed that a decrease or an increase in corrosion rate via grain refinement were due to the range of grain refinements as well as the aggressiveness of solution media [34].For example,adding a grain boundary to a single crystal material enhanced its corrosion resistance but grain size reduction up to ultra-fine or nano-size changed a localized into the uniform corrosion.Here,grain boundaries acted as a barrier effect in corrosion penetration and made a more uniform surface layer.Thus,the corrosion rate decreased and the corrosion degradation became uniform.Golapudi et al.[45]believed that corrosion behavior relating to the grain size corresponded with the behavior of the solution as well;while some solutions had the passivating capability and others did not have.Comparing several studies on pure Mg with grain size,Bahmani et al.[34]stated that escalating or de-escalating behavior was related to the level of aggressiveness of solution and range of grain size reduction.However,the second phase results in micro-galvanic corrosion and affects the corrosion rate seriously.As one can see,the corrosion rate of as-cast alloys containing more than 0.3wt.%Cu are not influenced considerably by grain refinement induced by Cu addition showing the dominating effect of second phases.To ascertain the galvanic effect,corrosion rates were plotted versus the fraction of the second phases(fi)and are shown in Figs.5(c)and(d).It can be observed that over a specific phase fraction around 0.01(wt.%),the slope significantly changed,showing different dominating mechanisms before and after that.The following equation is used to predict the corrosion rates considering both grain size and second phases [34].

Where CR0,b and c are three balancing constants and fi,fa,D and?Eiare the fraction of intermetallics,fraction of matrix(1-Σfi),grain size,and volta-potential difference relative to the matrix.As can be seen from(Figs.5(e)and 5(f)),all parts of the results can be predicted using this equation.

In addition,the corrosion resistance of Z2,ZC201,and ZC203 alloys,was analyzed by the EIS and polarization tests after 30min,24h,and 120h immersion times in the SBF solution.Figs.6 and 7 show the results of the EIS test(phase angle,Bode,and Nyquist plots).From the Bode and phase graphs,two peaks at the frequency ranges of ～10–10,000Hz and ～0.1–1Hz are observed after immersing for 24h,indicating the formation of two layers between metal and solution.However,for the 30min and 120h immersion times,only one layer was seen and the inner layer(observed at low frequencies)became almost weak or disappeared(except for the extrude Z2 alloy).On the other hand,the Nyquist curves had three semi-circles;(I)the high-frequency capacitive loop,corresponding to the charge transfer of a double electric layer.The size of the capacitive circuit determines the charge transfer resistance,(II)the medium-frequency circle,relating to corrosion products,and(III)the low-frequency circuit,associating with the inductance circle,which depicts substrate resistance that is degrading [6,35].According to Fig.7,the radius of the high-frequency circles enhances with more immersion times.Although the ZC201 alloy has the highest radius after immersing for 30min,the Z2 alloy has the largest high-frequency loop size after 24h and 120h immersion times(both cast and extruded alloys).Similar to the results of Bode and phase plots,the alloys formed two distinct circles after 24h immersion time(especially the extruded alloys).However,the medium-frequency layer is almost weak after 30min and 120h immersion times(except for the extruded Z2 alloy).The corresponding fitted plots are shown by solid lines in Fig.7.Moreover,Fig.7(g)shows the equivalent circuit model for data fitting of the EIS plots.

Fig.8.Polarization curves of the cast and extruded alloys after different immersion times.

In Fig.7(g),Rsis the solution resistance,Rcpis the resistance of the corrosion product layer,and Rctis the charge transfer resistance.CPE1 shows the associated constant phase element in parallel with Rcpand CPE2 is related to constant phase element parallel with Rct.Generally,CPEs are attributed to the mass transport processes in the solid phase and indicate a diffusion resistance.CPEs are well-defined for aberration from ideal dielectric properties of surface inhomogeneity orcurrent leakage in the surface [46].L is an inductive element in parallel with Rctand CPE2 which is attributed to the formation,adsorption,and desorption of the corrosion products on the surface of the samples [47].The right side loop of the equivalent circuit is related to the corrosion phenomena which occurs at the low frequencies and suggests that the corrosion occurs right on the surface,among cracks and discrete sites.This inductive loop mostly depends on the adsorption of chloride ions on the electrode surface [48,49].The applied model has been used widely by other researchers such as [50–52],which supports our analysis.

The extracted data in Table 5 illustrates the gradual enhancement of the Rcpvalues for the extruded alloys with more immersion time,due to the formation of more stable protection layer.However,the cast alloys faced with fluctuation of the Rcpvalues,owing to the formation of the unstable passive layer.At the early stages of immersion,cast ZC201 alloy illustrated the best protection among the others.However,after passing more immersion times,the cast and extruded Z2 alloys depicted the best corrosion resistance.After a significant reduction of Rcpfollowing 24h immersion,the Rcpreformed after 120h immersion,illustrating the improvement of the corrosion protection.

Fig.8 illustrates the potentiodynamic polarization curves for the alloys,which were carried out from the cathodic regions.icorrand Ecorrvalues are calculated by the Tafel extrapolation and listed in Table 6.

Comparing the potentiodynamic polarization plots in various immersion times,depicted the fact that an increase in immersion time changed the Ecorrvalues of the alloys.The Ecorrof the cast and extruded ZC203 alloys was initially enhanced to positive value after 24h and then decreased to more negative values after 120h of immersion.This outlines the formation of unstable film for the cast and extruded ZC203 alloys.Furthermore,the Ecorrof the cast and extruded Z2 and ZC201 alloys shifted to more positive values with more immersion time,showing an improvement in the film formation for these alloys.Furthermore,after 30min of immersion,the cast ZC201 alloy displayed the lowest corrosion rate(Pi),which was consistent with our previous studies [23,30].However,among the extruded alloys,Z2 alloy had the lowest icorramong the others [24].With more immersion times,the cast and extruded Z2 alloys showed good corrosion resistance in comparison to that of the ZC201 and ZC203 alloys,while the extruded alloys exhibited more efficient passivity than that of the cast alloys.Moreover,the cathodic regions of the polarization curves are associated with the hydrogen evolution reaction(HER)[53].The cathodic currents at?1.8 VSCE[53](which is adequately below the Mg OCP)for all of the alloys are calculated and listed in Table 6.The HER currents are almost enhanced by increasing the immersion time.

Fig.9.SEM micrographs of the corroded surfaces of the cast and extruded(a),(e)Z2,(b),(f)ZC201,(c),(g)ZC203,and(d),(h)ZC21 alloys after immersion for 24h.

The immersion,polarization,and EIS results demonstrated that the ZC201 alloy had unstable passive layer at the early stages of immersion and its protecting layer was destroyed after 24h of immersion and reformed again after 120h of immersion.The immersion test after 336h depicted that the corrosion rate of the ZC201 was lower than that of the Z2 alloy.The formation and deterioration of the corrosion protection layer with increasing the immersion duration were evidence for the change in the corrosion behavior.On the other hand,the extrusion enhanced the corrosion properties,due to the formation of a uniform passive layer [43,44].Moreover,at the early immersion times,weaker corrosion resistance for the extruded alloys was related to the presence of more grain boundaries and stored energy which promoted more degradation.

To study the corrosion mechanism,the samples were soaked for 24h and the surface was analyzed.Figs.9 and 10 illustrate the SEM micrographs of the corroded surfaces for the cast and extruded Z2,ZC201,ZC203,and ZC21 alloys,before and after removing the corrosion products.Moreover,the EDS point analysis of some regions in Figs.9 and 10 are listed in Table 7.

As can be seen in Fig.9,the cast Z2 and ZC201 alloys represent better surface features than that of the cast ZC203 and ZC21 alloys.The surfaces in Figs.9(a)and 9(b)contained a cracked-mud shape cover,in which the cracks were formed,due to dehydration of the corrosion layer during drying.Moreover,the cast ZC203 alloy was almost covered with the needle-like compounds(Fig.9(c1)).The EDS point anal-ysis from two regions of Fig.9(c1)in Table 7 depicted the changes in the atomic percent of Mg,O,P,Ca,Cl,Zn and Cu elements.It can be seen that the point A has lower amounts of Ca,P and O elements and higher amounts of Mg,Cl,Cu,and Zn elements than that of the point B.The needle-like compounds can be related to the MgCl2products which showed the high corrosion rate for cast ZC203 alloy.The presence of Zn and Cu in the region of the MgCl2compound exhibited the strong cathodic role of Cu-containing intermetallics to provoke the degradation of Mg [54].Due to the high amounts of Cl?in the solution,these ions may diffuse within the film and react with the Mg(OH)2compounds to produce the soluble MgCl2products [6,55].With more immersion time,the MgCl2compounds are dissolved into the Mg2+and Cl?ions,and increase the amounts of OH?and hence pH values near the surface.By progressing the above reaction,the Ca2+and H2PO4?ions in the solution may react with the OH?near the surface and form the insoluble phosphate there [55].The insoluble phosphates were seen near the MgCl2jungles,based on the EDS point analysis,which can be formed with the above mechanism.Furthermore,the volcano-like product on the surface of ZC203 alloy may be related to the presence of cathodic intermetallics beneath them [56].This hole can be formed,due to the creation of hydrogen gas during the immersion test underneath it [56].The cast ZC21 alloy also faced severe corrosion deterioration and its surface became rough(Fig.9(d)).On the other hand,the extrusion process significantly enhanced the surface features of the mentioned alloys.The cracked surfaces covered by the white compounds with different morphologies are discerned for the extruded Z2,ZC201,and ZC203 alloys(Fig.9).Based on the magnified SEM micrographs in Figs.9(e1),(f1),and(g1)and their related EDS point analysis in Table 7,the white products on these alloys mainly consist of the Ca,P,O and Mg elements,relating to the formation of Ca-P compounds.However,the surface of extruded ZC203 alloy included lower amounts of Ca and P than that of the Z2 and ZC201 alloys.Formation of insoluble Ca-P compounds is a piece of evidence for a degradation rate reduction of the extruded Z2,ZC201,and ZC203 alloys.However,on the surface of the extruded ZC21 alloy,the needle-like MgCl2compounds were scrutinized,illustrating the same corrosion condition of this alloy as that of the cast ZC203 alloy.This event also confirms the positive influence of extrusion on the corrosion behavior of ZC21 alloy.

Table 7 The data related to the EDS point analysis in Figs.9 and 10.

Fig.10.SEM micrographs of the corroded surfaces after removing the corrosion products for the cast and extruded(a),(e)Z2,(b),(f)ZC201,(c),(g)ZC203,and(d),(h)ZC21 alloys after immersion for 24h.

After removing the corrosion products(Fig.10),the extruded alloys had lower and shallower corrosion pits,comparing to that of the cast alloys,illustrating more uniform and less severe corrosion behavior for the extruded alloys than that of the cast alloys.The Z2 and ZC201 alloys also had fewer deteriorated regions,in comparison to that of the ZC203 and ZC21 alloys.Furthermore,there was evidence for the formation of more holes near the grain boundaries and intermetallics(Figs.10(c1)and(d1)).The magnified SEM in Fig.10(d1)demonstrated the presence of Mg(Zn,Cu)2intermetallics at the grain boundaries,according to its EDS point analysis(point F).Moreover,the magnified SEM micrograph in Fig.10(f1)represented the grain structure of the extruded ZC21 alloy with the dispersed intermetallics around the grains.The corrosion pits within the grain boundaries delineated the role of grain boundaries for the initiating the corrosion at the early stage of immersion.Moreover,inside some grains,the preferential corrosion attacks were perceived,which may represent the effect of grain misorientations on the deterioration of the grains [57].

Fig.11.Cross-sectional OM and SEM micrographs with their related EDS mapping analysis of the cast(a)Z2,(b)ZC201,and(c)ZC21 alloys after immersion for 30min.

Fig.12.Cross-sectional OM and SEM micrographs with their related EDS mapping analysis of the extruded(a)Z2,(b)ZC201,and(c)ZC21 alloys after immersion for 30min.

Fig.13.(a)-(f)SEM micrographs of the corroded surfaces of the cast and extruded Z2,ZC201,and ZC203 alloys after 336h immersion,(g)XPS analysis of the corroded surfaces of the cast Z2 and ZC201 alloys after 336h immersion.

The cross-sectional OM and backscattered SEM micrographs of the cast and extruded alloys with their mapping EDS analysis are shown in Figs.11 and 12,respectively.The OM images of the cast alloys exhibited the presence of mild corrosion attacks for the Z2 and ZC201 alloys and severe degradation for the cast ZC21 alloy.Moreover,the barrier effect of the continuous network of intermetallics for halting the corrosion progression was seen for the OM and SEM of the cast ZC21 alloy.This effect was also seen in the previous study [58],in which theβ-Mg17Al12intermetallics properly reduced the corrosion rate by acting as a barrier at the early stages of immersion.However,in the present study,this effect was reduced due to the significant cathodic impact of Mg(Zn,Cu)2intermetallics after longer immersion times.Moreover,the relationships between the corrosion degradation,grain boundaries,and intermetallics can be observed in the OM and SEM micrographs of the extruded alloys in Fig.12.In some regions,the grain boundaries halted the corrosion propagation and acted as a corrosion barrier.The SEM of the cast and extruded alloys in Figs.11 and 12 illustrated that the ZC201 and ZC21 alloys had the lowest and highest thickness of the corrosion products,respectively.Moreover,the corrosion products of the cast and extruded Z2 alloys were porous,while the Cu-bearing alloys had more compact corrosion products.Nevertheless,as can be seen in Figs.11(c1)and(c2),the corrosion layer of the cast ZC21 alloy was significantly messy,due to its high corrosion rate.Furthermore,the corrosion products of the alloys mostly had Mg,O,and Cl elements.The presence of Cl element may be related to the contamination of the surface and also formation of MgCl2compounds,showing the remarkable corrosion attacks on the passive layers.As can be seen in Fig.12(c2),the needle-like MgCl2products were detected on the top of the extruded ZC21 alloy.Moreover,the distributions of the Cucontaining intermetallics within the corrosion product layer in Figs.11(c2)and 12(c2)are evidence for their strong cathodic effect.Another interesting point was the presence of Ca and P elements in Figs.11(b2),12(a2),(b2),and(c2),illustrating the formation of Ca-P compounds for the extruded Z2,cast and extruded ZC201,and extruded ZC21 alloys.

Furthermore,the corroded surfaces of the cast and extruded Z2,ZC201,and ZC203 alloys after immersion for 336h were analyzed by the SEM micrographs and shown in Fig.13.Moreover,the XPS analysis for the corroded surfaces of the cast Z2 and ZC201 alloys are manifested in Fig.13(g).The XPS results were also checked by previous studies [16,59].The surface appearances revealed the mild corrosion attacks of Z2 and ZC201 alloys and the severe degradation of the ZC203 alloy.The XPS patterns of Z2 and ZC201 alloys depicted that both of the alloys had Mg,O,P,and Ca elements,showing the formation of Mg(OH)2and Ca-P compounds.The intensity of the formation of Mg(OH)2and Ca-P compounds was higher for the ZC201 and Z2 alloys,respectively.The presence of Mg(OH)2and Ca-P products on the surfaces of Z2 and ZC201 alloys confirmed their proper corrosion behavior,in comparison with the other alloys.

Fig.14.Schematic representation of the corrosion mechanism happening in the cast Z2,ZC201,and ZC21 and extruded ZC201 alloys after immersion in the SBF solution.

Based on the immersion,electrochemical and surface analysis tests,the corrosion behavior of the cast and extruded Z2 and ZC201 alloys were the best among the other alloys.The addition of 0.1wt.% Cu enhanced the corrosion resistance of the cast and extruded Z2 alloys,due to the grain refinement and also the formation of low volume fraction of MgZnCu intermetallics.Higher Cu additions created more Mg(Zn,Cu)2intermetallics which generated more Volta-potential difference with theα-Mg phase and hence more corrosion deterioration.Moreover,the formation of Ca-P compounds significantly improved the resistance of the passive layer and reduced the corrosion rate.Fig.14 schematically summarizes the abovementioned statements about the corrosion mechanism happening in the cast and extruded alloys in the SBF solutions.

Fig.15 represents the stress-strain curves of the cast and extruded Z2,ZC201,and ZC203 alloys and also the comparison of their tensile properties and corrosion rate with other cast and deformed binary,ternary,and quaternary Mg alloys.The experimental conditions used to measure thecorrosion rates are listed in Table 8.The tensile properties and corrosion rate of the cast and extruded Z2,ZC201,and ZC203 alloys were comparable with the cast Mg-Zn-Ca,cast Mg-Zn-Ca-Al,extruded Mg-Zn-Ca-Ag,rolled Mg-4Li-1Ca,and rolled Mg-Al alloys.However,the RE-containing alloys demonstrated a better combination of tensile and corrosion properties than that of the existing alloys.Nevertheless,the low cost of Zn and Cu,the necessity of Cu element for the body,and also the superior biocompatibility of Mg-Cu alloys[25–29]recommended that the cast and extruded ZC201 alloys may be a good candidate for using as biodegradable magnesium implant.Moreover,their properties may be enhanced with further treatments,i.e.solution treatment or severe plastic deformations.

Table 8 The experimental conditions used to measure the corrosion rates.

Fig.15.(a)stress-strain curves of the cast and extruded Z2,ZC201,and ZC203 alloys,(b),(c),and(d)the comparison of tensile properties and corrosion rates of present alloys with the some other reported Mg alloys.

4.Conclusion

In this research,microstructure and in vitro corrosion resistance of the cast and extruded Mg-2Zn-xCu(x=0,0.3,0.5,1.0,3.0 wt.%)alloys were investigated.The related conclusions are drawn as follows:

1-The addition of the Cu element influenced on microstructure of the Mg-Zn alloys.The grain refinement and also enhancement of the intermetallic volume fraction were achieved by Cu additions.The MgZnCu intermetallics were formed by adding 0.1wt.% Cu and excess Cu additions produced the Mg(Zn,Cu)2compounds.Applying hot extrusion led to the grain refinement by dynamic recrystallization and also changing the distribution of intermetallics.

2-The immersions after 14 days revealed that the degradation behavior of the cast and extruded Mg-2Zn alloys improved by adding 0.1wt.% Cu,but more additions deteriorated the corrosion resistance.The reason was found to be related to the grain refinement,formation of a better passive layer,and also the presence of low volume fraction of MgZnCu intermetallics that created lower Volta potential differences with theα-Mg phase.Furthermore,the extrusion enhanced the corrosion resistance of the Mg-2Zn-(0.3,0.5,1,and 3)Cu alloys,but it did not have a significant effect on the degradation behavior of Mg-2Zn and Mg-2Zn-0.1Cu alloys.

3-The electrochemical tests at different immersion times indicated that the Cu-containing alloys,formed the unstable passive layers at the short immersion times and the reformed with more immersion times.Moreover,the extruded alloys illustrated the formation of a more uniform passive layer,in comparison to that of the cast alloys.

4-The corroded surface analysis demonstrated the formation of Mg(OH)2and Ca-P compounds for the cast and extruded Mg-2Zn and Mg-2Zn-0.1Cu alloys.However,more Cu additions led to the formation of MgCl2compounds and severe degradation.

5-Regarding the enhancement of tensile properties and degradation behavior with 0.1wt.% Cu addition,the extruded Mg-2Zn-0.1Cu alloy can be considered as a potential candidate for using as a biodegradable implant.

Declaration of Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.