Effects of ytterbium addition and heat treatment on the mechanical properties and biocorrosion behaviors of Mg-Zn-Zr alloy

Lu Li, Tao Wang, Yu Wang, Cun-cai Zhang, Hao Lv, Hua Lin, Wen-bin Yu,Chu-jie Huang

1School of Materials and Energy, Southwest University, Chongqing, China

2 Department of Clinical Laboratory, Chongqing Ninth People’s Hospital, Chongqing 400700, PR China

Abstract Mechanical properties and biocorrosion behaviors in simulated body flui (SBF) of newly developed Mg-5.8 Zn-0.5 Zr-x Yb (ZK60-x Yb, x=0, 1.0, 2.0wt%) magnesium alloys in the solution-treated (T4) and artificially-age (T6) conditions were investigated.The results of mechanical properties show that with Yb addition, the microhardness and the ultimate tensile strength (UTS) of the tested alloys are significantl increased despite a slight decrease in tensile elongation in both T4 and T6 conditions.Especially, after the T6 treatment, the microhardness and the UTS of the samples were further improved, which was mainly attributed to the precipitation strengthening.The biocorrosion behaviors of the tested alloys were studied using electrochemical examinations and immersion tests.The results indicate that the biocorrosion resistance of the tested alloys is significantl improved by Yb addition in both T4 and T6 conditions.Although the corrosion resistance was slightly deteriorated after T6 treatment, the aged ZK60-2.0 Yb alloy still exhibited a favorable corrosion behavior, which was mainly ascribed to the corrosion barrier effect of a more compact and uniform protective fil induced by the dispersed nano-scale precipitates.Electrochemical measurements also confirme these observations.Given the favorable comprehensive performance in mechanical and biocorrosion behaviors, the T6 treated ZK60-2.0 Yb alloy may be considered as a promising candidate for biomedical applications.

Keywords: Mg-Yb-Zn-Zr alloy; Mechanical properties; Biocorrosion behaviors; Heat treatment; Biomedical applications.?Correspondence author.

1.Introduction

Magnesium (Mg) alloys are considered as an ideal material for biodegradable implants due to their favorable biological properties, degradation ability in physiological environment and stimulatory effect on the new bone formation[1-3].Unfortunately, the prevailing of Mg and Mg-based alloys as degradable biomaterials has been severely restricted by the limitations of the relatively high degradation rates in the human bio-environment and the ensuing loss of mechanical integrity [4-7].To this end, it is of significanc to tailor magnesium alloys with suitable biocorrosion rates and desirable mechanical properties for biomedical applications.

Alloying is justifie as a routine and practicable measure to address the aforementioned challenge.Various elements such as Ca [8,9], Sn [10,11], Zn [12], Si [13], Mn [14,15], Zr [16],Sr [17], and rare earth elements (REE) [18-20] have been incorporated and investigated.Among recent attempts, the ZK60 alloy (Mg-Zn-Zr series) is drawing increasing interest due to the excellent mechanical properties with yield strength(YS)>200MPa,ultimate tensile strength(UTS)>260MPa and elongation (EL)>10%, which is fully meet the requirements for Mg alloys to be successfully used as temporary implant materials [21].More importantly, the primary strengthening element Zn is an essential nutritive element [22] and a trace amount of grain-refinin agent Zr doping (less than 0.8wt%)has also been considered well tolerated for the human body[23], rendering reasonable biocompatibility and biosafety of ZK60 alloy.However, in spite of a developed corrosion resistance achieved in comparison with pure Mg [24], the corrosion rate of ZK60 alloy is still too fast for implant use [25-28], which is mainly ascribed to a relatively high Zn content.Literature once reported a decrease in corrosion resistance when the Zn content higher than 2.5wt% [29].Therefore, it is prerequisite to enhance the biocorrosion resistance to make full potential of ZK60 alloy with desirable mechanical performance and biocompatibility.

Ytterbium (Yb) is chemically classifie by the ionic radii as the group of heavy REEs from Tb to Lu.Yb can solute in Mg and it was proven to improve the corrosion resistance of Mg-Zn alloys in NaCl solution [30], Mg-Zn-Ca alloys in Ringer’s solution [31] and distilled water [32].In regard to the biocompatibility of Yb-containing alloys, Wang [33] confirme that compared with the Yb-free control, improved biocompatibility of MgZnCaYb metallic glasses (MGs) was achieved as determined by live-dead, MTT, and cytoskeleton staining assays, respectively [16].Li [34] reported that the CaMgZnSrYb high-entropy bulk metallic glass (HE-BMG)could stimulate the cells’ growth and proliferation of cultured osteoblasts and promote new bone formation after 2 weeks of implantation in vivo.Babilas [31] argued that the cytotoxicity tests by using CaMgZnYb extracts did not show any negative influenc in a wide range of 5, 10, 50 and 100%.Recently,Yb was introduced into the ZK60 alloy to further enhance the mechanical properties.In our previous studies, the effects of different Yb alloying levels (0.5 to 2.0wt% Yb addition)on the mechanical properties of the ZK60 alloy were systematically investigated [35-39].It was found that Yb acted a significan role in regulating the strength and ductility of the ZK60 base alloy.However, direct investigations of Yb addition on the biocorrosion behaviors of ZK60 alloy have been scarcely revealed.

To date, there is no agreement in the literature regarding the optimum content of Yb in Mg alloys from a corrosion perspective, although the solubility limit of Yb in Mg is 4.8% in mass at 400°C [40].As an extension of our previous work,the ZK60-xYb (x=0, 1.0 and 2.0wt%) alloys with different thermal treatments of solution and aging were prepared, and the biocorrosion behaviors combined with mechanical properties were systematically investigated in this study.The primary focus of this study is to achieve a comprehensive understanding of the Yb alloying content and the ensuing heat treatments on the biocorrosion behaviors and mechanical properties of ZK60 alloy.To the best of the authors’ knowledge,this is the firs time a detailed investigation of the biocorrosion behaviors of Mg-Yb-Zn-Zr (denoted as ZYbK) series alloys has been theoretically documented.The present work,focusing on apprehending the biocorrosion behaviors of the newly developed ZYbK alloys, is likely to provide valuable insight into the potential biomedical applications of the highperformance Yb-containing magnesium alloys as biodegradable implant materials.

2.Experimental methods

2.1.Materials preparation

Mg-Yb-Zn-Zr alloy with nominal composition (wt%) of Mg-5.8 Zn-0.5 Zr-xYb(x=0,1.0 and 2.0,hereafter denoted as ZK60, ZYbK610 and ZYbK620, respectively) were melted in a crucible furnace using pure Mg, pure Zn, Mg-30% Zr,Mg-15% Yb, and protected by a mixed atmosphere of SF6(10vol.%) and CO2(bal.).The as-cast ingots were soluted at 400°C for 24h (T4), and then aged at 200°C for 16h (T6).The same heat-treatment process parameters for different Yballoyed samples were set deliberately to evaluate the corrosion response with the same heat input.

2.2.Microstructure characterization

The specimens for microstructure characterization and in vitro corrosion measurements were machined with a geometric size of 10mm×10mm×3mm.The metallographic samples were successively grounded with a series of SiC abrasive paper (from 400 to 2000 grid), then etched using a solution consisting of 5g picric acid, 10ml acetic acid, 10ml distilled water and 100ml ethanol.The microstructures were observed by Olympus optical microscopy (OM), scanning electron microscopy (SEM, Sirion 200), and transmission electron microscopy (FEI Talos F200X operated at 200kV).The average gain size was evaluated by using the linear intercept procedure according to ASTM E112.

2.3.Mechanical properties

Mechanical properties were evaluated by tensile and Vickers micro-hardness tests.Tensile samples were machined along the longitudinal direction with 30mm gage length and 6mm cross-section diameter according to ASTM B557 M-02a.Tensile tests were conducted at an initial strain rate of 10?3s?1at ambient temperature by using a CMT5504 tensile test machine, and the fracture morphology was examined by SEM (Sirion 200).In each condition, 3 specimens for tensile tests were prepared and evaluated.Measurements of micro-hardness were performed on a Vickers hardness tester(MH-5L) using an indenter load of 9.8N and a loading time of 20s, and each hardness value was the average of 10 individual measurements.

2.4.Electrochemical test

Electrochemical measurements were conducted on an electrochemical workstation (CHI660e).To simulate the environments of the bone, the simulated body flui (SBF solution), which is composed of NaCl 8.035g/L, CaCl20.292g/L,NaHCO30.355g/L, Na2SO40.072g/L, KCl 0.225g/L,MgCl2·6H2O 0.311g/L, K2HPO4·3H2O 0.231g/L, and deionized water was prepared [41].The samples were sealed with epoxy with an exposed area of 1 cm2.According to ASTMG3-89 standard, the electrochemical tests were performed at 37±0.5°C in a beaker containing 200ml SBF solution with an initial pH value of ～7.4.A standard three-electrode cell was used, in which a platinum plate served as the counter electrode, a saturated calomel electrode acted as the reference electrode, and the tested sample was the working electrode.The specimen was immersed in SBF for 3600s to reach a relatively stable open-circuit potential, and then the polarization curves were acquired at a scanning rate of 1mV/s.All potential values were referred to SCE values.The electrochemical impedance spectra (EIS) was evaluated at a disturbing potential of 10mV over a frequency range of 0.01 Hz-100kHz and the data were analyzed by using ZsimpWin software.The polarization resistance,Rp, was calculated from the electrochemical parameters, according to the following equation:

Whereβaandβcare the anodic and cathodic Tafel slopes,respectively;Icorris the corrosion current density.

2.5.Immersion test

According to ASTM G31-72, immersion testing was performed in the SBF solution at 37±0.5°C for 10 days.The samples were machined into a cylinder ofφ10mm×10mm and hung in SBF solution.Moreover, the solution pH values were tested every 24 h, and each record was the average of 5 individual measurements.After 10 days’ immersion, all samples were cleaned with distilled water and dried in warm fl wing air.The samples were ultrasonic cleaned in a solution of 200g/L Cr2O3and 10g/L AgNO3for 10min, and then in ethanol for 5min to remove corrosion products.The phase constituent analysis of the corrosion product was analyzed by X-ray diffraction (XRD, Rigaku D / Max 2500).The average corrosion rate was calculated using the equation described in ASTM G31-72.

WhereVis the corrosion rate (mm / year); the constant of coefficienK=8.76×104,Wis the weight loss of corrosion during soaking,Ais the surface area exposed to the solution(cm2),Tis the exposure time (h) andDis the density of the material (g/cm3).

3.Results and discussion

3.1.Microstructure

Fig.1 depicts the optical micrographs of as-soluted and as-aged ZK60 alloys with different Yb contents of 0, 1.0,and 2.0wt%.It was obvious that Yb posed a positive effect on refinin microstructure.With increasing Yb addition from 0 to 2.0wt%, the average grain sizes in the T4 and T6 conditions decreased from ～48.7 to ～32.4μm and ～59.2 to ～36.1μm, respectively, which was mainly stemmed from the considerable grain growth restriction induced by Yb during solidificatio [42].Moreover, it was also found that the growth of grain size was negligible after T6 heat treatment,which revealed an improvement of thermal stability by Yb addition.A small amount of undissolved eutectic phase remained at triangular grain boundaries of the ZYbk620-T4,which was mainly attributed to the low diffusion coefficien of Yb in the Mg matrix.In addition, as shown in Fig.1b,d, and f, precipitation appears both at grain boundaries and in grain interior after T6 treatment, with increasing density in companion with Yb addition.According to our preceding studies [37], the precipitates primarily consisted of Mg-Zn binary and Mg-Zn-Yb ternary phases.

Fig.1.Microstructure of the soluted and aged ZK60 (a, b), ZYbK610 (c, d)and ZYbK620 (e, f) alloys.

3.2.Mechanical properties

3.2.1.Hardness

The Vickers hardnesses of specimens with different Yb additions were measured after T4 (colored in black) and T6(colored in red) heat treatments, as shown in Fig.2.It was shown that the hardness values increased with rising Yb content in both T4 and T6 conditions.Although the variation tendency was similar, the mechanisms for the enhancements were different.For the T4 condition,a slightly increased hardness was observed, which was mainly attributed to the effect of solid solution strengthening resulted from the dissolution of nonequilibrium eutectic phases into the matrix.In regard to the T6 condition, however, a more obvious enhancement of microhardness was observed, which was due to the effect of precipitation strengthening.With increasing Yb addition, a much fine and more dispersed precipitation pattern dominated in the specimens with T6 treatment (Fig.3), leading to an increase of the precipitation strengthening effect.Moreover, it was worth noting that the average grain size was decreased with increasing Yb addition (Fig.1).The refine microstructure also exerted a positive effect on the increase of the microhardness.

Fig.2.The Vickers hardnesses of ZK60, ZYbK610 and ZYbK620 alloys in the T4 and T6 conditions.

Fig.3.SEM micrographs of (a) ZK60, (b) ZYbK610 and (c) ZYbK620 alloys in the T6 condition.The inset in the upper right corner shows the magnifie region as indicated with a white box.

Table 1Mechanical properties of ZK60, ZYbK610 and ZYbK620 alloys in T4 and T6 conditions.

3.2.2.Tensile properties

The tensile properties of the tested samples with different Yb additions and heat treatments are shown in Table 1.It was obvious that with increasing Yb addition both in T4 and in T6 conditions, the tensile yield strength (TYS), ultimate tensile strength (UTS) increased accompanied by the decrease of the elongation (EL).In the T4 condition, the increase in UTS was mainly attributed to the effects of the solid solution strengthening combined with the grain refinemen strengthening [43].As shown in Fig.1, the average grain size apparently decreases with Yb alloying.In addition, a few undissolved eutectic phases remained in triangular grain boundaries of ZYbK-T4 alloys, which was detrimental to the elongation.As for the T6 condition, a remarkable increase in UTS was observed in all the alloys but the elongation decreased slightly.During aging treatment, a large number of fin particles precipitated, acting as the additional barriers to hinder the movement of dislocations and grain boundaries with straining, resulting in the improvement of the UTS.In the meantime, the precipitates may also serve as the crack source during uncoordinated deformation, which was harmful to the enhancement of ductility.The maximum UTS and the minimum elongation values were obtained in the ZYbK620-T6 alloy, which was primarily ascribed to the occurrence of a much higher density of nano-scaled precipitates compared with that in ZK60-T6 and ZYbK610-T6 alloys (Fig.3).Therefore, the strength enhancement in the T6 treated alloys should be attributed to the precipitation strengthening effect.

Fig.4.SEM micrographs of fracture surface (a) ZK60-T4, (b) ZYbK610-T4 and (c) ZYbK620-T6.

3.2.3.Fracture characteristics

Fig.4 shows the fracture surfaces of the ZK60-T4,ZYbK610-T4 and ZYbK620-T6 alloys.It was found that for all the samples, the cracking surfaces showed a mixed characteristic of dimples combined with cleavage planes, which suggested that both brittle and ductile fracture were taken place simultaneously.As for the ZK60-T4 sample,many dimples were covered on the fracture surface, indicating that the ductile fracture was the main fracture mechanism (Fig.4a).When 1.0wt% Yb alloyed, the dimple morphology still dominated on the fracture surface, but the increase of cleavage planes implied that the brittle fracture became active(Fig.4b).After aging treatment, the fracture surface of the ZYbK620-T6 sample was relatively fla and a large number of particles could be observed (Fig.4c).Herein, it inferred that the inter-granular fracture coupled with plastic tearing became the leading fracture mechanism.The fracture characteristics were in accordance with the results of mechanical properties.

Fig.5.Polarization curves of alloys with different Yb additions in (a) T4 and (b) T6 conditions after immersion in the SBF solution.

Table 2Fitting results of polarization curves.

3.3.Electrochemical polarization

3.3.1.Potentiodynamic polarization

The potentiodynamic polarization curves of the tested alloys with T4 and T6 treatments in the SBF solution are shown in Fig.5.It is shown that the curves of the Yb-containing samples in both T4 and T6 conditions shift left and up relative to the Yb-free counterparts, suggesting a smaller corrosion current density and more noble corrosion potential achieved,which is indicative of the enhancement of corrosion resistance by Yb addition.Table 2 lists the corrosion potential (Ecorr),corrosion current density (Icorr), and linear polarization (Rp)obtained from the potentiodynamic polarization curves by the Tafel extrapolation method.TheEcorr,IcorrandRpvalues indicated that although the corrosion resistance was improved by Yb doping in the T4 condition, the effect of a further increase in Yb content from 1.0% to 2.0% was negligible as the polarization curves of ZYbK610-T4 and ZYbK620-T4 overlapped to some extent.In contrast, the more Yb was alloyed in the matrix, the better corrosion resistance was achieved in the T6 condition.This may arise from the different corrosion mechanisms underlying different microstructure conditions after solution and aging treatments.In the T4 condition, most of the Yb-containing phases were dissolved into the matrix,resulting in the formation of much cleaner grain boundaries whereas after T6 treatment, profuse precipitates appeared and the morphology, density, and composition were fairly different with increasing Yb content from 0 to 2.0%.Moreover, the samples with T6 treatment achieved a more negativeEcorr, higherIcorrand lowerRpin comparison with those of T4 treated counterparts, which illustrated that the corrosion resistance was slightly deteriorated after aging treatment.It was interesting to fin that the values ofEcorr,Icorr, andRpof the ZYbK620-T6 sample were ?1.264 VSCE, 136 μA/cm2, and 270Ωcm2respectively, exhibiting better corrosion resistance than that of other two T6 treated samples.This could be attributed to the formation of more compact and uniform protective fil induced by the dispersed nano-scale precipitates in ZYbK620 aging sample (Fig.3), which acted as an effective corrosion barrier to prevent the penetration of Cl?in SBF solution.

3.3.2.Electrochemical impedance spectroscopy measurements

The electrochemical impedance spectroscopy (EIS) plots of the tested alloys with T4 and T6 treatments are depicted in Fig.6a and b, respectively.As shown in Fig.6a, the impedance spectra of the ZK60-T4 and ZYbK620-T4 alloys consist of a capacitive loop in the high-frequency range and an inductive loop in the low-frequency range, while the ZYbK610-T4 alloy exhibits two capacitive loops in both high and medium frequency ranges.It was suggested that the electrochemical corrosion mechanisms might be different in the ZYbK610 alloy relative to ZK60 and ZYbK620 alloys in the T4 condition.As well known, the high-frequency behavior is closely related to the charge transfer resistance, which can be evaluated qualitatively by the diameter of the highfrequency capacitive loop and a larger diameter usually indicates a higher corrosion resistance [44].The medium frequency capacitive loop implies the formation of the surface oxide fil and the low-frequency inductance loop may indicate the initiation of pitting corrosion [45].Accordingly,the ZYbK610-T4 alloy exhibited a better corrosion resistance and there should be no occurrence of clear localization corrosion.In addition, the degradation rate of the T4-treated samples increased evidently in the following order of ZYbK610,ZYbK620, and ZK60.

On the other hand, as for the T6 samples, all of the impedance spectra consisted of a high-frequency capacitive loop and a low-frequency inductive loop, although the inductive loop of the ZYbK620 sample was depressed to some extent, which indicated that the corrosion mechanisms of these three samples are similar.By comparing the diameters of the high-frequency capacitive loops, the corrosion resistance of the T6 treated samples ranked as ZK60

Fig.6.EIS patterns of the tested alloys immersed in SBF: (a, b) Nyquist plots and (c, d) Bode plots.

Additionally, the Bode impedance diagrams of the T4 and T6 samples are shown in Fig.6c and d,respectively.The samples with T4 treatment depict a higher impedance than that of T6-treated samples, which suggests that the T4 treated alloys possess better corrosion resistance because the difficult of charge transfer is proportional to the impedance in the Bode plot.The ZYbK620 alloy exhibited a higher impedance in the T6 condition whereas the ZYbK610-T4 alloy showed the largest impedance among the tested alloys, which was the feature of passive behavior.It was obvious that the corrosion laws presented in the EIS tests in both T4 and T6 conditions were well consistent with the polarization results.

3.4.Immersion test

Fig.7.Corrosion rates of ZK60, ZYbK610 and ZYbK620 alloys with T4 and T6 treatments after immersion in SBF at 37°C for 10 days.(The values are the equivalent corrosion rates in mm/y.).

3.4.1.Corrosion rate calculation

The corrosion rate is widely used as a general index to evaluate the overall corrosion of the tested samples.Fig.7 gives the corrosion rates obtained from the immersion tests in the SBF solution for 10 days (obtained from Eq.(1)).The corrosion rates of Yb free alloys were higher than those of ZYbK alloys in both T4 and T6 conditions, which suggested that the addition of Yb could enhance the corrosion resistance of the magnesium matrix.Moreover, heat treatments also affected the corrosion rate but the degree of influenc was different.It was apparent that the change of corrosion rate in both T4 and T6 conditions for ZYbK alloys was much lower than that for ZK60 alloys.It means that the corrosion behavior of matrix may be stabilized by Yb addition after heat treatments.The corrosion evaluations were well agreed with the electrochemical tests.

Fig.8.pH-value evolutions of the samples immersed in SBF at 37°C for 10 days: (a) T4 and (b) T6.

3.4.2.pH value monitoring

Fig.8 depicts the pH-value variations of the SBF solution immersing the ZK60, ZYbK610 and ZYbK620 alloys in the T4 (a) and T6 (b) conditions versus incubation times.In spite of the duration differences in reaching stabilization, the pH-value evolutions exhibited a similar variation trend, i.e.,a continuous increase followed by a plateau.The increase at the initial immersion period was mainly attributed to the alkalization reaction of the sample in SBF, resulting in redundant OH?ions in the solution and an evident increase in the pH value.With a prolonged exposure time, the pH value reached a relatively stable level due to the ions equilibrium achieved in solution reactions.Moreover, for both two conditions (T4 and T6), the pH values of ZYbK samples were lower than those of Yb-free alloys during immersion, indicating that the addition of Yb could enhance the corrosion resistance of the ZK60 base alloy since a higher pH value represented a higher corrosion rate of the specimen [46].In addition, heat treatment also played a significan role in the corrosion resistance of the tested samples.It was shown that the pH values of the T6 treated alloys were higher than those of the samples with T4 treatment at the same immersion duration prior to the stabilization.The ZYbK610-T4 and ZYbK620-T6 alloys exhibited lower pH values and slower evolution rates.These observations correlate well with the results of the electrochemical measurements.

3.4.3.X-ray diffraction examination

The XRD patterns of corrosion products after immersion in the SBF solution at 37°C for 10 days are shown in Fig.9.It was shown that the corrosion products both in T4 and T6 conditions were a mixture of MgO, Mg(OH)2and Mg.It should be illustrated that no diffraction peaks corresponding to Yb containing phases could be detected due to the limited Yb content of less than 5wt% and the Mg intensity should be attributed to the influenc of the substrate.In the T4 condition, as shown in Fig.9a, the Mg intensities of the corrosion products for ZYbK alloys were much higher than those of ZK60 alloy, which suggested that the corrosion products of Yb-free counterpart were much more than those of ZYbK alloys.Moreover, the ZYbK610 sample exhibited the highest intensity in the Mg phase, indicating limited corrosion products covered.In the T6 condition, as depicted in Fig.9b, Mg intensities of the corrosion products were much lower than those in the T4 condition, implying that the corrosion might be more severely occurred in the T6 treated samples.In addition, the intensive presence and higher intensity of Mg(OH)2diffraction peaks in ZYbK620 revealed the corrosion fil was thicker than other two T6 treated alloys [47].As well acknowledged, the magnesium hydroxide was considered as an insoluble protective fil to retard the degradation [48], which may be a reasonable explanation for the favorable corrosion resistance of ZYbK620-T6 sample.

3.4.4.Corrosion morphologies

Fig.10 shows the surface characteristics of the corroded regions of the T4 treated samples with and removal of the corrosion products after immersion in the SBF solution at 37°C for 10 days.It was clear that the corrosion severity of Yb-free alloy (ZK60) was higher than that of ZYbK alloys.As shown in Fig.10a, the almost entire surface was severely corroded combined with a large number of deep pits and corrosion cracks.In contrast, no evident localized corrosion was observed on the surfaces of Yb alloyed samples.The ZYbK610-T4 alloy exhibited a lower corrosion depth in companion with fewer corrosion products, indicating a more protective fil should work during immersion.As shown in Fig.10d, a far more fla surface with general corroded areas were exposed after the corrosion products removed, which further verifie the speculation.The improved corrosion resistance in the ZYbK610-T4 alloy was attributed to a better dissolution of the second phases in the Mg-matrix,eliminating the formation of the galvanic couple.Furthermore, the Yb induced corrosion suppression in the T4 condition may indicate that Yb can stabilize the Mg-matrix and enhance the corrosion resistance, presenting a similar effect of zirconium in the magnesium substrate [47].As for the ZYbK620-T4 alloy, relatively more severe corrosion compared with the ZYbK610 alloy may be due to the enrichment of undissolved eutectic phases in triangular grain boundaries (Fig.1e), triggering the occurrence and the acceleration of micro-galvanic corrosion.These results were consistent with the finding in the electrochemical tests, pH value monitoring, and XRD determinations.

Fig.9.XRD patterns of corrosion products after immersion in SBF at 37°C for 10 days: (a) T4 and (b) T6.

Fig.10.Scanning electron micrographs of the corroded surfaces for the T4 treated specimens with and removal of the corrosion products after immersion in the SBF solution at 37°C for 10 days.

Fig.11.Scanning electron micrographs of the corroded surfaces for the T6 treated specimens with and removal of the corrosion products after immersion in the SBF solution at 37°C for 10 days.

Fig.11 shows the surface features of the corroded regions of the T6 treated samples with and removal of the corrosion products after immersion in the SBF solution at 37°C for 10 days.The degree of corrosion was apparently increased in the T6 treated samples compared with that in the T4 condition.The Yb-free sample presented a much thicker corrosion fil with the broader and deeper spread of corrosion pits and cracks, indicating the occurrence of the localized corrosion.In contrary,Yb-containing alloys with T6 treatment both showed a fairly lower degree of general corrosion although a few corrosion pits were still found in the ZYbK610 sample.The ZYbK620-T6 sample exhibited a much thicker and more homogeneous productive layer without evident corrosion attack being investigated, which played a significan role in reducing the current exchange between the corroded interfaces and alleviating the corrosion sensitivity [49], leading to favorable corrosion resistance in the T6 condition.The formation of a more uniform and compact protective layer in the ZYbK620-T6 sample should be stemmed from the appearance of an increased area fraction of dispersed nano-scale second phases during the T6 treatment.As shown in Fig.12, the TEM examination reveals that three kinds of typical phases are observed in the ZYbK620-T6 alloy.The dominant fin dot-like Mg-Zn-Yb (Yb-poor, point 1, Figs.12a and c) and rod-shaped Mg-Zn (point 2, Figs.12a and c) phases both appear within the grains primarily; and a small number of coarse blocky Mg-Zn-Yb phase (Yb-rich) mainly precipitate at grain boundaries (point 3, Figs.12b and c).Although the intermetallic compounds in magnesium alloys possess a relatively higher potential as the cathode, the aforementioned second phases with different sizes and distribution acted different roles in corrosion.The coarse blocky Mg-Zn-Yb phase discontinuously located at grain boundaries gave rise to the localized corrosion to some extent, which was due to the big difference in potential between the Mg-matrix and the intermetallic compound with a relatively large size [50,51].It was worth noting that the pitting corrosion was not pronounced in the ZYbK620-T6 corroded morphology, as the number of that coarse phase was very limited.On the other hand, for the intragranular dispersed Mg-Zn-Yb and Mg-Zn phases,the internal galvanic corrosion would occur uniformly over the matrix as the much-refine size and more homogeneous distribution, leading to the formation of a passive fil covering them completely and thus restricting their activity for further corroding[52,53].Consequently,an anti-corrosion barrier was established to protect the magnesium matrix.It was worth illustrating that the overwhelming nano-scaled second phases with such a high density were primarily attributed to the increasing Yb solution in the base alloy.With Yb concentration increasing from 0 to 2wt%, more Zn element was redistributed as the form of much fine Mg-Zn-Yb ternary phases [38,39], which consumed the Zn element and thus resulted in a decreased size and density in original Mg-Zn binary phases.Therefore, the galvanic corrosion arisen from the original coarse Mg-Zn precipitates was suppressed in companion with the improvement of corrosion resistance induced by the dominance of dispersed fin Mg-Zn and Mg-Zn-Yb nano-scaled precipitates.These finding agree with the observations in the electrochemical tests, pH value monitoring,and XRD examinations.And the specifi identificatio of the ternary Mg-Zn-Yb phases with different sizes requires further investigation.

Fig.12.TEM images of ZYbK620-T6 alloy (a and b) and EDS results of the typical precipitates (c).

3.4.5.Influenc of grain size and second phases on corrosion morphologies

Numerous studies have investigated the effects of grain size and second phases on corrosion [54,55].It was recognized that the magnesium alloys with smaller and more refine second phases offer a marginally lower corrosion rate and better passivation, achieving improved corrosion resistance [50,56,57].According to the present study, the second phase plays a more significan role in corrosion due to a comparable grain size the T4 and T6 treated alloys with the same Yb content presented.Furthermore, the ZYbK610 alloy in the T4 condition with almost clear grain boundaries and moderate grain sizes presents the best corrosion resistance, whereas, for the T6 treated alloys, the ZYbK620 with most dense nanoprecipitates exhibits lower corrosion rate.The slightly deteriorated corrosion resistance of ZYbK620-T4 alloy should be ascribed to the undissolved coarse eutectic phase at triangular grain boundaries, triggering the occurrence and the acceleration of micro-galvanic corrosion.Therefore, the second phase should dominate the corrosion behavior in this work.Finely and uniformly distributed second phases could effectively improve the corrosion resistance.In contrast, the large and coarse second phases were detrimental to the corrosion resistance due to the micro-galvanic corrosion [58,59].

4.Conclusion

(1) Ytterbium demonstrates a favorable effect on improving the mechanical properties of the ZK60 alloy.With Yb addition, the microstructure is refine obviously and the microhardness coupled with the UTS are increased remarkably despite a slight decrease in tensile elongation.The enhancements can be ascribed to the effects of the grain refinement solid solution strengthening, and precipitation strengthening.

(2) The addition of Ytterbium plays a beneficia effect on the corrosion resistance of the ZK60 alloy.The ZYbK alloys exhibit lower corrosion rates than those of Ybfree alloys in both T4 and T6 conditions.A higher corrosion potential and EIS impedance coupled with a lower corrosion current density and pH value are observed in the Yb-containing alloys.

(3) Heat treatment plays an important role in corrosion behavior.The corrosion resistance of the T4 treated alloys is better than that of alloys with T6 treatment.The ZYbK620-T6 alloy exhibits a favorable corrosion resistance, which may be attributed to the corrosion barrier effect of a more compact and uniform protective fil induced by the precipitation of dispersed nano-scale particles.

(4) Compared with the grain size, second phases dominate the corrosion behavior in the present investigation.Although the ZYbK610-T4 alloy shows a better corrosion resistance in the SBF solution, the ZYbK620-T6 alloy should be considered as a promising candidate for biomedical applications due to a favorable comprehensive performance in mechanical and biocorrosion behaviors.

Funding

This work was supported by the National Natural Science Foundation of China (grant number 51975484 and 51605392)and Fundamental Research Funds for the Central Universities(grant number XDJK2020B001).

Declaration of Competing Interest

The authors declare that there is no conflic of interest.