Viheslv Bzhenov, Andrey Koltygin, Alexnder Komissrov, Ann Li, Vsiliy Butin,Regin Khsenov, Alexey Anishhenko, Alexnder Seferyn, Juli Komissrov,Yuri Estrin

aFoundry Department, National University of Science and Technology “MISiS”, Leninsky pr.4, 119049 Moscow, Russian Federation

b Laboratory of Hybrid Nanostructured Materials, National University of Science and Technology “MISiS”, Leninsky pr.4, 119049 Moscow, Russian Federation

c Department of Metallurgy of Steel, New Production Technologies and Protection of Metals, National University of Science and Technology “MISiS”,Leninsky pr.4, 119049 Moscow, Russian Federation

d Department of Materials Science and Engineering, Monash University, Clayton VIC 3800, Australiа

e Department of Mechanical Engineering, The University of Western Australia, Crawley, WA 6009, Australia

Abstract In recent years magnesium alloys have been studied intensively with a view to their potential use in bioresorbable medical implants.In the present work the microstructure and the corrosion properties of a new bioresorbable Mg-4wt% Ga-4wt% Zn alloy and its variants with low Ca, Nd or Y additions were investigated.These alloys are of interest due to the effica y of gallium as an element inhibiting bone resorption,osteoporosis, Paget’s disease, and other illnesses.A severe plastic deformation technique of equal channel angular pressing (ECAP) was shown to provide the alloys with favorable mechanical properties.In addition, a desirable low rate of degradation in a simulated body flui(Hanks’ solution) was achieved.

Keywords: Bioresorbable materials; Magnesium alloys; Gallium; ECAP; Hanks’ solution; Simulated body fluid

1.Introduction

A new paradigm in bone implantology and cardiovascular surgery is the use of bioresorbable implants that would gradually dissolve as the healing process progresses.This approach helps minimizing implant-induced inflammatio of the surrounding tissue and eliminates the need for revision surgery for implant removal.In particular, the benefit of using biodegradable implants in pediatric orthopedics are evident as children’s bones grow after surgery [1].Magnesium alloys are promising candidates for bioresorbable implant materials because of their good biocompatibility,sufficientl high mechanical strength, and acceptable biodegradation rate [2,3].Also, the density and Young’s modulus of magnesium alloys are close to those of cortical bone [4,5].

The aim of this work was to investigate alloys of the Mg-Ga-Zn system that could be suitable for use in osteosynthesis.Gallium was chosen as an alloying element due to its beneficia properties.Gallium is known as one of the bone resorption inhibitors [6].It is also known to be effective in the treatment of disorders associated with accelerated bone loss,including osteoporosis [7], hypercalcemia [8-10], Paget’s disease[11-12],and multiple myeloma[13].In addition,gallium increases bone thickness, strength and mineral content [14].In a few studies the effect of Ga on osteoclasts resorption was investigated and it was shown that in a concentration of 100μM gallium has anti-osteoclastic effect reducing osteoclastic resorption, differentiation and formation without negatively affecting osteoblasts [6,15,16].Furthermore, gallium was found to have an antibacterial effect [17,18].The antibacterial activity of gallium was further confirme by in vitro and in vivo studies on Mg micro-alloyed with Ga [19].

A concern about potential cytotoxicity of gallium was allayed by an investigation [20] that showed no pronounced cytotoxicity at Ga concentration below 310ng/ml.Further information about therapeutic activity of gallium can be found in articles [5,7-11,21-24].There are commonly no issues with the biocompatibility of zinc, the third constituent of the alloy studied.As a matter of fact, the daily demand for Zn in humans (15mg) is higher than that for other alloying elements typically contained in magnesium alloys [4].

The elements Mg, Ga and Zn have the atomic radii of 0.160nm, 0.135nm and 0.137nm, respectively [25].Due to similar values of the atomic radius, Zn and Ga atoms have about the same solid solution strengthening effect on Mg[26,27].The mechanical properties of Mg-Ga and Mg-Zn alloys are rather low for as-cast and heat treated conditions[20,26,27].The mechanical characteristics of the alloys can be improved by severe plastic deformation [28], and that was the path chosen in the present work.

The influenc of small additions of Ca,Y,and Nd was also investigated.Calcium is one of the essential trace elements for human body.It is mainly stored in bones and its level is controlled by homeostasis of the skeleton [29].Rare-earth (RE)elements are also used in biodegradable magnesium alloys to provide improved strength and corrosion resistance characteristics.However, they can only be added in small quantities because of their tendency to get accumulated in the liver and bone, thus having a negative systemic effect.Some of RE elements were found to have anti-tumor properties and can potentially be used for treatment of cancer [29-31].There are some reports, however, that the use of yttrium may result in severe hepatotoxicity [32].In view of the reported anticancer activity of gallium [33], Ga-containing magnesium alloys may be an interesting alternative to RE-containing ones in the context of bone replacement in cancer patients.

2.Material and methods

Magnesium (99.98wt% Mg), zinc (99.995wt% Zn), gallium (99.99wt% Ga) and Mg - 30wt% Ca, Mg - 20wt%Nd, Mg - 20wt% Y master alloys were used as raw materials for alloy preparation.Melt was prepared using a resistance furnace with a steel crucible.Before melting,the crucible was coated with a boron nitride layer.For melt protection from ignition an Ar+ 2vol% SF6atmosphere was used.Cylindrical ingots with 35mm diameter and 140mm length were cast into a steel mold preheated to 150°C.For dissolution of a non-equilibrium eutectic phase and homogenization of the ingot composition, a solution heat treatment at 400°C for 75 h was used.

Fig.1.(a)Schematic representation of the ECAP die,and(b)the deformation route used.(c) Geometry and dimensions (in mm) of the non-standard tensile specimens.

The liquidus and solidus temperatures were determined using a differential scanning calorimetry (DSC) analyzer (Texas Instruments, Q600, USA).The samples were heated and cooled in nitrogen atmosphere at a rate of 10K/min.

Cylindrical specimens 20mm in diameter and 100mm in length for equal channel angular pressing (ECAP) were cut from ingots using spark cutting.The ECAP processing was carried out on a 250 t vertical hydraulic press (NanoMet,RHP 250, Russian Federation) equipped with an ECAP die,which is shown in Fig.1a.The die channel angle was 120°.Processing of specimens was done in three passes using a combined route CB, which involved specimen rotation about the pressing axis by 180° and 90° after the firs and second pass, respectively, Fig.1b.The pressing rate was 1mm/s.Before ECAP the specimens and the die were preheated to 310°C.

Specimens of the as-cast, heat treated, and ECAP processed alloy were ground and polished for metallographic observations.For scanning electron microscopy (SEM) analysis and energy dispersive X-ray spectroscopy analysis (EDS) a Scanning Electron Microscope (Tescan, Vega SBH3, Czech Republic) with the EDS system (Oxford Instruments, X-act,UK) were used.The chemical composition of the alloys was determined by EDS analysis on metallographic sections with an analysis area of 1 mm2.For each specimen three areas were analyzed.

In vitroimmersion corrosion testing was carried out on disk-shaped samples after ECAP processing.The disks, having 12mm diameter and 2mm height, were spark machined and ground using 320 grit abrasive SiC paper.Samples were immersed in 400ml Hanks’ solution (PanEco, Russian Federation) at 37°C for 192h.Hydrogen evolution from corroding samples was measured as a function of time and related to 1 cm2of the specimen surface area.In addition, the variation of the pH value of Hanks’ solution was recorded using a pH meter (Hanna Instruments, HI83141, USA).

Fig.2.(a) The composition of phases in binary Mg-Zn, Mg-Ga, and ternary Mg-Zn-Ga alloys in as-cast condition.(b) An enlarged part of (a) showing the maximum solubility line for Zn and Ga in magnesium solid solution (Mg).

The electrochemical corrosion investigations were also performed using a potentiostat/galvanostat corrosion system(Volta, IPC Pro-MF, Russian Federation) with frequency response analyzer (FRA) at the same temperature and corrosion medium.The samples 13×13×10mm3in size were spark machined and ground using 320 grit abrasive SiC paper.A three-electrode system was employed in which the alloy samples served as working electrodes with an exposure area of 1 cm2.Platinum and saturated calomel electrodes were used as the counter and reference electrodes, respectively.Potentiodynamic polarization experiments were performed from the cathodic region at ?2.3V to the anodic region at ?1V with a scan rate of 1mV/s.The corrosion current density (icorr) and corrosion potential (Ecorr) were determined from Tafel fitting

For mechanical testing, tensile specimens were spark machined from the ECAP processed materials.Specimens with non-standard geometry were used(Fig.1c).Tensile tests were performed on the universal testing machine (Instron, 5569,UK).Brinell hardness (HB) was determined using an universal hardness tester (Innovatest, Nemesis 9001, NL).The following test parameters were used: ball indenter 2.5mm in diameter; load: 61.3kN; holding time under load: 30s.

3.Results and discussion

The phase diagram of Mg-Zn-Ga system is unknown.In accordance with Mg-Ga and Mg-Zn binary phase diagrams,Mg5Ga2and Mg7Zn3phases can occur in binary alloys[34-36].Some binary and ternary alloys were melted and analyzed with regard to their microstructure and phase composition using SEM and EDS.The composition of phases is shown in Fig.2a.Magnesium solid solution and two eutectic phases with a composition Mg-1.9 at% Zn-24.6 at% Ga and Mg-20.9 at% Zn-6 at% Ga were observed in the microstructure of the alloys.The eutectic phases found were obviously based on binary Mg5Ga2and Mg7Zn3phases and can be denoted as (Mg,Zn)5Ga2and (Mg,Ga)7Zn3, both containing Mg, Ga, and Zn.The microstructure and mixed EDS map of Mg-7.4wt% Zn-20.0wt% Ga alloy are shown on Fig.3.It can be seen that both (Mg,Zn)5Ga2and (Mg,Ga)7Zn3phases have the same backscattered electron (BSE) contrast and look the same in the alloy’s microstructure.In Fig.2b an enlarged part with maximum solid solubility line is shown.It can be seen that high solubility of Zn and Ga in Mg is observed in Mg-Zn-Ga alloys.The solidus temperature of Mg-7.4wt% Zn-20.0wt.% Ga alloy obtained by thermal analysis is 307°C.It is known that the eutectic temperatures in Mg-Zn and Mg-Ga systems are 341°C and 420°C, respectively[34,35].From these data, it can be surmised that a ternary eutectic with (Mg), (Mg,Zn)5Ga2and (Mg,Ga)7Zn3phases does exist in the Mg-Zn-Ga system.

Table 1Chemical composition of the alloys prepared.

Mg-4wt%Zn-4wt%Ga was chosen as the base alloy(BA)composition.After appropriate heat treatment, this alloy can exist as a solid solution of Zn and Ga in Mg.This state is of interest as it offers a possibility of precipitation hardening.Also, alloys with this composition and small additions of Ca,Y, or Nd were cast.As will be shown below, these additions promote the formation of small eutectic particles that remain in the alloy after heat treatment potentially acting as recrystallization centers during subsequent deformation.The chemical composition of the alloys prepared is shown in Table 1.

The microstructures of the alloys in the as-cast condition are shown in Fig.4a-d.The microstructure of the base alloy in the as-cast condition comprises dendrites of the magnesium solid solution(Mg)and the eutectic phases(Mg,Zn)5Ga2(Mg,Ga)7Zn3.Ca, Y, and Nd promote the formation of new eutectic phases that are denoted in Fig.4 as Mg,Ga,Ca,Zn;Y,Ga,Mg,Zn and Ga,Nd,Mg,Zn.It is difficul to determine the composition of these phases because of their small size.The content of Ca, Y, and Nd in magnesium solid solution (Mg)is lower than the EDS detection limit (<0.01wt%).

Fig.3.(a) SEM micrograph and (b) mixed EDS map of Mg-7.4wt% Zn-20.0wt% Ga alloy in as-cast condition.

Fig.4.Microstructure of the alloys: (a), (e), (i) Base alloy (Mg-4wt% Ga-4wt% Zn); (b), (f), (j) Base alloy+0.2wt% Ca; (c), (g), (k) Base alloy+0.3wt%Y; (d), (h), (l) Base alloy+0.3wt% Nd in (a-d) as-cast state, (e-h) after heat treatment for 75h at 400°C, and (i-l) after three ECAP passes.

Differential thermal analysis results showed that the liquidus and solidus temperatures of the base alloy obtained for an as-cast sample is 608°C and 307°C, respectively.The solidificatio range is large and is expected to promote the formation of shrinkage porosity in the cast samples.Variations in liquidus and solidus temperatures due to addition of small amounts of Ca, Y, and Nd did not exceed 5°C.

Fig.5.The influenc of the heat treatment duration at 400°C on the Zn and Ga content in the dendrite cores and at the edges, and on the amount of eutectic in alloy microstructure for base alloy (Mg-4wt% Ga-4wt% Zn).

The content of the alloying elements in the Mg solid solution dendrites in the base alloy was measured by EDS during heat treatment at 400°C.Measurements were taken every fi e hours.The concentrations of Zn and Ga at dendrite edges and in the core of a dendrite are shown in Fig.5.It can be seen that in the as-cast condition the Zn and Ga content at a dendrite edge was close to the nominal content of the respective element in the alloy, while that in the middle of a dendrite was about 1wt%.After 20 h of heat treatment the element concentrations in the core of a dendrite were nearly the same as at the edges.The amount of eutectic phases in the base alloy during the heat treatment is also shown in Fig.5.After the firs fi e hours of heat treatment the amount of eutectic phases dropped sharply: from 5.1 to 0.3vol%.After 20 h the amount of eutectic phases was as low as 0.1vol%.A concomitant decrease from 65.8 to 56.9 HB in the hardness of the base alloy after 20h of heat treatment was observed.

The microstructures of the alloys after 75h of heat treatment at 400°C are shown in Fig.4e-h.A significan change of the microstructure is evident.In the base alloy the eutectic was almost dissolved in magnesium and only a small amount of eutectic phase was retained.The eutectic particles were spheroidized.Also,a large amount of spherical diffusion pores were observed in the microstructure of the base alloy after heat treatment, as shown in Fig.6a.The microstructure of the base alloy with a 0.3wt% Y addition is seen in Fig.6b.Remarkably, no diffusion pores were found in this alloy.Neither did diffusion pores occur in BA+0.2wt% Ca and BA+0.3wt% Nd.By contrast, like in the base alloy, eutectic phases were retained after heat treatment in the alloys with Ca, Y, and Nd additions.The volume fraction of retained eutectic phases in Y- and Nd-containing alloys was lower than in the alloy with a Ca addition.As mentioned above, in the base alloy the composition of the solid solution (Mg) after heat treatment was close to the nominal composition of the alloy.The same situation was found for the alloys with Ca, Y,and Nd additions.In the modifie alloys,these micro-alloying elements were not observed in (Mg) after the heat treatment.Most probably their content was below the EDS detection limit.Apparently, most of Ca, Y, and Nd atoms were contained in the eutectic phases and are believed to have stabilized them.

The microstructure of the base alloy and the base alloy with Ca, Y, and Nd additions after ECAP (three passes) is shown in Fig.4i-l.It is seen that microstructure after ECAP is similar for all alloys studied.It comprises (Mg) grains and precipitate particles about 1μm in size.These stress-induced precipitates were caused by ECAP deformation and reside inside the grains and on the grain boundaries.The precipitates situated at the grain boundaries inhibit grain growth during ECAP and stabilize the microstructure [27].In binary Mg-Ga alloys Mg5Ga2precipitates with a size of about 0.5μm were observed after deformation [20,27].In our case, Mg7Zn3precipitates were observed alongside Mg5Ga2in the base alloy due to the presence of Zn.The grain size was measured using the linear intercept method.The results are presented in Fig.7.From the micrograph for base alloy in the insert, it is seen that its microstructure is bi-modal: it comprises large grains surrounded by small ones.The average size of the large grains is about 90μm and that of the small grains is about 10μm.Kubasek et al.also obtained large (47μm) and small(6.5μm) grains in hot extruded Mg-3.5wt% Ga alloy [27].More uniform grain size distributions with an average grain size of approximately 10μm were observed after ECAP processing of the alloys micro-alloyed with Ca, Y, or Nd.A possible reason for that is the presence of small eutectic particles in those alloys.These eutectic particles retained in the material after heat treatment may act as nucleation centers during recrystallization.Indeed, these alloys were fully recrystallized after ECAP.

For potential applications in bone implants,the mechanical properties of a candidate alloy should be as close to those of cortical bone as possible.A mismatch between bone and implant results in stress shielding that leads to bone resorption[29,37].The mechanical properties of cortical bone are as follows.Yield strength (YS): 105-114MPa, ultimate tensile strength (UTS): 35-283MPa, tensile elongation: (EL) 1.1-2.1%,and Young’s modulus:10-30GPa [4,5].Broadly speaking, the mechanical strength characteristics of the alloys obtained by employing the above processing route match those of the cortical bone reasonably well,and additional benefi being a high tensile ductility of the alloys.The YS, UTS, and EL values of the alloys that underwent ECAP processing are shown in Fig.8.The YS of the base alloy and the alloy with 0.2wt% Ca addition are about the same (approx.165MPa).Addition of 0.3wt%Y or 0.3wt%Nd to the base alloy reduce YS to 140MPa.The magnitude of UTS is within the 280-300MPa range for all alloys studied, except for BA+0.2wt%Ca whose UTS is only 255MPa.All alloys exhibit a high EL in excess of 15%.The largest tensile elongation (over 25%)was found for BA+0.3wt.% Nd.These mechanical properties are comparable with those of Mg-6wt%Zn alloy hot extruded at 250°C [38]and are a little better than those of Mg-3.5wt%Ga alloy [27].

Fig.6.The microstructure of alloys after 75h of 400°C heat treatment: (a) base alloy (Mg-4wt% Ga-4wt% Zn); (b) base alloy+0.3wt% Y.

Fig.7.The grain size of the base alloy (Mg-4wt% Ga-4wt% Zn) and the base alloy with 0.2wt% Ca, 0.3wt% Y or 0.3wt% Nd additions after three ECAP passes.In the insert the bimodal microstructure of BA with large and small grains is shown.The grain boundaries are marked by yellow arrows.

Fig.8.The mechanical properties (YS - yield strength, UTS - ultimate tensile strength, El - elongation) of the base alloy (Mg-4wt% Ga-4wt% Zn)and the base alloy with 0.2wt% Ca, 0.3wt% Y or 0.3wt% Nd additions after ECAP processing (three passes).

While biodegradability is a core requirement on a material for bioresorbable implant applications, the excessively high biocorrosion rate of most magnesium alloys poses a serious problem.A particular handicap is the large amount of hydrogen that evolves due to a chemical reaction of Mg with water in a bodily fluid This may give rise to the formation of hydrogen cavities inhibiting bone growth [39] and to the occurrence of gas gangrene [40].In any case, the biodegradation time must be sufficientl large for bone fracture healing to occur before the disintegration of the implant.Commonly,for leg fractures the healing process is completed within three to four months,but in some cases it can take up to six months or longer [23,41-43].The presence of the corrosion products affects cell fusion/differentiation and high concentrations of corrosion products negatively impact the number of osteoclast progenitor cells and the mature osteoclast cell function [44].One of the problems with magnesium implants is associated with pitting corrosion that reduces the mechanical integrity of the implant and may cause premature failure before the completion of the healing process [29].The complexity of the magnesium biocorrosion phenomena has been discussed in numerous publications, see, e.g., recent reviews [4,45].Gallium and zinc are known to reduce the corrosion rate of magnesium [20,38].Strain and surface roughness commonly exert a negative effect on corrosion resistance [46-48].A remedy can be provided by coatings [29,49-51].Our results shown below demonstrate the effect of alloy composition on the rate of corrosion of samples after three ECAP passes in Hanks’solution, Fig.9.The smallest amount of evolved hydrogen was observed for the base alloy.After 192h of immersion it was 0.64ml/cm2.The 0.2-0.3wt% additions of Ca, Nd,and Y enhanced the evolution of hydrogen and raised its levels after 192h of incubation to 1.44, 1.20, and 0.87ml/cm2,respectively.

Fig.9.Hydrogen evolution during 192h of immersion of the base alloy(Mg-4wt% Ga-4wt% Zn) and the base alloy with 0.2wt% Ca, 0.3wt% Y or 0.3wt% Nd after ECAP processing (three passes) in Hanks’ solution at 37°C.

It is seen from Fig.10 that the corrosion rate define above diminishes with time for all four alloys.In medical practice,in the firs week after surgery the gas formed had to be removed through a subcutaneous needle [29].This time frame indicates that the data on the corrosion rate after 192h of immersion is relevant.The lowest corrosion rate after 192h(about 1.3·10?3ml/cm2h) was observed for the base alloy.This is three times higher than the corrosion rate tolerable to human body (0.4·10?3ml/cm2h) [52].Alloying with Nd and Y yielded even higher corrosion rates of, respectively,1.8·10?3and 1.7·10?3ml/cm2h after 192h of immersion.The highest corrosion rate,3.5·10?3ml/cm2h,was observed for the base alloy with Ca addition.

Fig.10.Corrosion rate during 192h of immersion of the base alloy (Mg-4wt% Ga-4wt% Zn) and its variants with 0.2wt% Ca, 0.3wt% Y or 0.3wt%Nd after ECAP processing (three passes) in Hanks’ solution at 37°C.

Fig.11.Variation of pH of the medium over 192h of immersion for the base alloy (Mg-4wt% Ga-4wt% Zn) and its variants with 0.2wt% Ca, 0.3wt% Y or 0.3wt% Nd after ECAP processing (three passes) in Hanks’ solution at 37°C.

The pH value of Hanks’ solution was independent of the alloy composition and varied during the immersion test from 7.1-7.3 to 6.1-6.3, as shown in Fig.11.At firs the pH level rose to 7.7-7.9, which can be related to a high corrosion rate(Fig.10).The pH magnitude peaked at about 60 h of immersion, followed by a gradual decrease replicating the decrease in the corrosion rate.At this stage, the exact reasons for the non-monotonic behavior of pH are not fully understood.A decrease of pH during corrosion testing of Mg-based alloys was also observed in [53-55].A possible reason for this decline in the pH level might be contamination by microorganisms acidifying the medium [54].Even if such an effect did occur in our case, it would not invalidate the corrosion test results.Indeed, Bornapour et al.studied the alloy degradation with or without daily renewal of simulated body flui (SBF).They reported that there was almost no difference in the biodegradation behavior between the two conditions, the pH having only a minor effect on alloy degradation [56].We thus maintain that contamination by microorganisms, even if cannot be ruled out, does not reduce the reliability of the corrosion test results.

Evaluation of corrosion resistance based on the kinetics of hydrogen evolution is, of course, only an indirect method of assessing the corrosion rate.The latter quantity is related to,but not synonymous with the hydrogen evolution rate.Still,the general trends in the corrosion behavior of various alloys can be determined on that basis.Unfortunately, the measured mass loss after immersion was too low for direct precise quantificatio of the corrosion rate.Because of that to calculating the mass loss-related corrosion rate by calibrating the hydrogen evolution amount using a correlation between the two quantities obtained earlier.Such a correlation was established for the alloys with the same composition as in the present study, but prepared via flu melting of commercial purity raw materials [57].Both gas hydrogen evolution and mass loss were measured.It was established that evolution of 1ml of H2corresponds to approximately 1mg mass loss of the alloys and this value was used for data conversion.A similar conversion coefficien was obtained via calculation and by comparison of mass loss and hydrogen evolution results elsewhere[20,58].On that basis the average corrosion rate in mm per year was calculated from the data in Fig.9 in accordance with the ASTM Standard Practice code [59].The average corrosion rate for base alloy was found to be 0.16mm/year.The addition of 0.2, 0,3 and 0.3wt% of Ca, Nd, and Y resulted in an increase of the corrosion rate to 0.37, 0.30 and 0.22mm/year, respectively.

Polarization curves of ECAP-processed samples was shown on Fig.12.Three curves for each alloy was obtained.In accordance with our results, the additions of Ca, Y and Nd to base alloy lead to positive shift of corrosion potential form?1.55V to near ?1.48V.The mentioned additions also lead to an increase in the corrosion current density.This quantity was used for corrosion rate calculation [60].The calculated corrosion rates of the alloys obtained via electrochemical and immersion corrosion tests are shown in Fig.13.The corrosion rates obtained via electrochemical test was 6-10 times higher than obtained by immersion test.For both tests the lowest corrosion rate was obtained for base alloy and highest corrosion rate was obtained for base alloy with 0.2wt% Ca addition.The long-time immersion corrosion test is more precise than rapid electrochemical corrosion measurements and the results of the immersion tests should be used to evaluate the biodegradation kinetics of the alloys in question.

Fig.12.Polarization curves of the base alloy (Mg-4wt% Ga-4wt% Zn) and its variants with 0.2wt%Ca,0.3wt%Y or 0.3wt%Nd after ECAP processing(three passes) in Hanks’ solution at 37°C.

Fig.13.Corrosion rates obtained by immersion and electrochemical corrosion testing of the base alloy (Mg-4wt% Ga-4wt% Zn) and its modification with 0.2wt% Ca, 0.3wt% Y or 0.3wt% Nd after ECAP processing (three passes) in Hanks’ solution at 37°C.

The surface of the samples was analyzed after immersion corrosion test in Hanks’ solution.A surface layer of the base alloy samples is shown in Fig.14.The composition of different surface areas obtained by EDS analysis for the base alloy and its modification with Ca, Y, and Nd is shown in Table 2.The bright part of the sample surface, designated as area A,with microstructure shown in Fig.14b, mostly contains magnesium and oxygen.This indicates the presence of a thin Mg(OH)2fil that is partially protective [20].In alloys with Ca, Nd, and Y additions the area A contains more oxygen(41.3, 46.3 and 36.9 at%, respectively) than its counterpart in the base alloy (26.9 at%); it also contains small amounts of P and Ca originating from Hanks’ solution.The mentioned change in the oxygen concentration is presumably associated with an increase of the thickness of the Mg(OH)2layer.The presence of P and Ca can be indicative of the formation of a thin passivation layer of Ca3Mg3(PO4)4of the kind found previously [61].The dark area, designated as area B (Fig.14c),which contains magnesium,oxygen,phosphorus,and calcium,is associated with a Ca3Mg3(PO4)4layer whose thickness is large relative to that in area A.The composition of this layer was nearly the same for all alloys.At a transverse section through area B (Fig.14d) the thickness of the Ca3Mg3(PO4)4layer is visualized.It is expected that RE elements should be localized within the layer formed by the corrosion products [29], but in this study Y or Nd were not detected there.

Fig.14.(a) The base alloy (Mg-4wt% Ga-4wt% Zn) sample after the immersion for 192h at 37°C in Hanks’ solution, (b) microstructure of sample area labeled A where the MgOH2 is predominant, (c) microstructure of sample area labeled B where the Ca3Mg3(PO4)4 is predominant, (d) microstructure of a transverse section.

Table 2The surface layer composition of base alloy (Mg-4wt% Ga-4wt% Zn) and its variants with 0.2wt% Ca, 0.3wt% Y or 0.3wt% Nd after ECAP processing(three passes) after 192h of immersion in Hanks’ solution at 37°C.

It is of interest to compare the performance of the alloys considered here with the literature data for their contenders.In Fig.15, the mechanical characteristics (YS, UTS and El) are plotted vs.the corrosion rate for alloys developed in our work vis-à-vis other biodegradable magnesium alloys [38,62-65].Only those alloys which were prepared from high purity raw materials and exhibit a reasonably low corrosion rate were included in the data set.The corrosion rate results for original simulated body flui [66] and modifie SBF (m-SBF)[67] were included alongside those for Hanks’ solution because they have about the same content of NaCl and a similar content of other constituents [68].It is seen that the ECAPprocessed alloys of the MgGa4Zn4 system populate the upper left part of the diagram.That is to say, they compare favorably with the competing systems.Thus, the alloys developed in this study have a low corrosion rate comparable with that of Mg-3 wt% Nd-0.4 wt% Zr-0.2 wt% Zn and Mg-3 wt%Gd-2.7 wt% Zn-0.4 wt% Zr-0.1 wt% Mn.However, the latter two alloys outperform the present ones in terms of their yield strength and are slightly better with regard to the ultimate tensile strength.The main problem with the RE-containing alloys is that the rare earth elements exhibit undesirable toxicity to human body when their concentration is high [29].Other alloys included in the data set exhibit mechanical characteristics similar to those developed in this work but are definitel inferior to the latter in terms of the corrosion resistance.

Fig.15.Mechanical characteristics vs.corrosion rate in Hanks’ and SBF solutions for a range of biodegradable magnesium alloys after deformation processing.

4.Conclusions

The phases and their composition in Mg-Ga-Zn alloys and the maximum solubility of Zn and Ga in (Mg) were determined.Guided by these results, an alloy with the composition Mg-4wt% Zn-4wt% Ga was chosen for investigation as a base alloy.Also, its modifie versions with small additions (0.2-0.3wt%) of Ca, Y, or Nd were investigated.The microstructure of the base alloy in the as-cast condition consisted of a solid solution of zinc and gallium in magnesium (Mg) and two eutectic phases: (Mg,Zn)5Ga2and(Mg,Ga)7Zn3.After heat treatment the eutectic phases disappeared, and diffusion pores emerged.Micro-alloying of the base alloy with Ca, Nd, or Y turned out to promote the formation of new phases in the microstructure in the as-cast condition, which were retained after heat treatment.A further effect of micro-alloying was the inhibition of porosity formation.To improve the mechanical characteristics of the alloys, severe plastic deformation of the heat-treated material by equal-channel angular pressing was employed.Microstructure of the base alloy after ECAP processing was found to be bi-modal.It comprised large grains 90μm in size surrounded by small grains with the average size of about 10μm.Small additions of 0.2 -0.3wt% of Ca, Y, or Nd to the base alloy promoted the formation of a fine and more uniform grain structure, with an average grain size of about 10μm.The mechanical properties of the base alloy and its modification through micro-alloying were found to be on par with or better than those of several Mg alloys considered as candidate materials for bone replacement implants.

The most valuable property of the alloys developed is their reasonably low corrosion rate.With the estimated values of the corrosion rate in Hanks’ solution in the range from 0.16mm/year for the base alloy to 0.22mm/year for its modifie version obtained by micro-alloying with Y, the alloys developed satisfy the typical requirements on the corrosion resistance of Mg based implants set by orthopedic surgeons[29,37,52].

Through this study were able to ascertain that the Mg alloy Mg-4wt% Ga-4wt% Zn processed by equal-channel angular pressing possesses a favorable property profil in terms of a combination of mechanical properties and biocorrosion rate,which makes it suitable for applications in bone implants.

Acknowledgments

The authors gratefully acknowledge financia support from the Ministry of Science and Higher Education of the Russian Federation under the Increasing Competitiveness Program of NUST “MISiS” (No.К2-2019-008) implemented according to the governmental decree N 211 dated 16 March 2013.