Linxi Chen,Cheng Guo,Crsten Blwert,Junjie Yng,Donghu Chen,Xiojin Wng,*,Zhento Yu,Mikhil L.Zheludkevih,Wei Li,*

aSchool of Materials Science and Hydrogen Energy,Foshan University,Foshan,Guangdong 528000,China

b Institute of Advanced Wear & Corrosion Resistant and Functional Materials,Jinan University,Guangzhou 510632,China

c Institute of Surface Science,Helmholtz-Zentrum Hereon,Geesthacht 21502,Germany

d College of Mechatronics and Control Engineering,Shenzhen University,Shenzhen 518060,China

Abstract Magnesium(Mg)alloys are attractive biodegradable implant materials.The degradation products on Mg alloys play a critical role in the stability of the interface between implant and surrounding tissue.In the present study,the effects of dynamic deformation on the interface layer of biomedical Mg-1Zn alloy were investigated using the constant extension rate tensile tests(CERT)coupled with electrochemical impedance spectroscopy(EIS).The deformation of the Mg-1Zn alloy had an adverse influenc on the impedance of the surface degradation layer formed in simulated body flui that only containing inorganic compounds.However,the surface degradation layer with improved corrosion resistance was obtained for the strained samples tested in protein-containing simulated body fluid The spontaneous or enhanced adsorption of protein into the degradation product led to a fl xible and stable hybrid anti-corrosive layer.A relationship between the dynamic deformation of Mg alloy and the impendence of the degradation layer was established,which demonstrates the necessity for in situ characterisation of the evolution of the surface layer under dynamic condition.? 2021 Chongqing University.Publishing services provided by Elsevier B.V.on behalf of KeAi Communications Co.Ltd.This is an open access article under the CC BY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/)Peer review under responsibility of Chongqing University

Keywords:Degradation layer;Protein;Biodegradable Mg alloy;Dynamic deformation;Impedance.

1.Introduction

Magnesium(Mg)alloys have been extensively investigated as implant materials,owing to their biocompatibility and good mechanical properties compared to those of bio-polymers and bio-ceramics[1-3].The Young’s modulus of Mg alloys is similar to that of human bone,which can moderate“stress shielding effects”[1].It is a remarkable fact that the degradability of Mg alloys can avoid some negative effects related to a second surgery or to a permanent metallic implant after tissue healing[4,5].

However,the degradable Mg implant would undergo certain mechanical deformation and bio-stress during the implementation operations or service period after fixatioin vivo.For example,the bone plate would suffer from bending deformation or dynamic loading during the implanting operation or normal human movement[6].Cardiovascular stent undertakes larger crimping and expansion deformation to support the stenotic vessels or cyclic stress from cardiac impulse[7].The deformation can cause variable dimension of the bulk implant materials,often resulting in a premature failure due to stress corrosion failure in aqueous physiological environment[8,9].Macroscopic elastic and/or plastic deformation could lead to an irreversible change of the microstructure of metallic materials[10,11].It is usually accompanied with the formation of twins,dislocations and residual stress in the hexagonal close-packed(HCP)structure of polycrystalline Mg alloys[12].The deformation twinning and residual strain could result in a high corrosion rate,localised corrosion and fast deterioration of mechanical properties of Mg alloys[13].

Deformation strain is related to a dynamic evolution of the material surface,affecting the formation,stability or physical shielding properties of the surface layer[14,15].For example,the adhesion of surface natural passivation fil is significantl affected by the structural mismatch between the oxide fil and HCP Mg lattice.Slip steps that occur during deformation process can cause the rupture of oxide fil where become the corrosion initiation sites[16].With respect to the ceramic coated Mg alloy,such as Ca-P[14]and micro-arc oxidation(MAO)coatings[17],the elastic and/or plastic deformation of Mg substrate normally cause microcracks or small fragments falling off of the coating.Even for PLGA coating[18]and rapamycin-eluting poly(D,Llactic acid)coating(PDLLA/RAPA)[19],nano/micro-scale cracks are also observable after the deployment deformation of the polymer coated Mg alloy.The mechanical defects provide channel for corrosive ions penetration,then forming the micro-galvanic corrosion with surrounding coating and resulting in severe localised corrosion,which ultimately accelerate the deterioration of the mechanical properties of the coated Mg alloys.

Thein vitrodegradation process and surface degradation layer evolution of Mg alloys in different inorganic solutions,such as simulated body flui(SBF),Hanks’solution and phosphate-buffered saline(PBS)have been widely studied[1,20,21].However,the presence of organic components such as biomolecules,proteins or cells,would lead to a more complex case for the degradation process of Mg alloys[22].The spontaneous adsorption process of proteins presented in biological solutions usually take place on the surface of implants by Van Der Waals and electrostatic interactions,followed by cell attachment,cell proliferation and growth[23,24].The organic-containing degradation layer formed on Mg implant surface can determine the interaction between bulk material and surrounding tissues[25].Wagener et al.[26]confirme a decreased corrosion rate for Mg alloy in bovine serum albumin(BSA)containing solution at the early stage of immersion tests.The authors recently demonstrated that the biomedical Mg-1Zn alloy exhibited better mechanical properties and corrosion cracking resistance in proteins-containing media[27].However,the chelating reaction between proteins and metal ions can cause the formation of colloidal organometallic complexes,whose migration away from the metals surface leads to an increasing corrosion rate of Mg alloys[28,29].

In a static physiological environment,many conventional tests can be applied to measure the separate effect of deformation strain or protein on the degradation and surface properties of Mg alloys[13,21,30].However,the coupling influenc of dynamic deformation strain and natural organic albumin on the degradation layer of Mg alloy during the deployment process or service period was seldom investigated.The degradation product layer was mainly composed of conversion products of dissolved Mg and solution compounds.Additionally,deposition or adsorption of biological components simultaneously occurs,and both together are determining the primary corrosion rate and biocompatibility of Mg-based implantsin vivo[31].Therefore,investigating the surface evolution under dynamic deformation in protein-containing biological environment is essential to further analyse the degradation behaviour of bio-Mg alloys.

In order to minimise the adverse effects of secondary phases in Mg alloys on corrosion resistance,a low alloying,wrought and heat-treatable Mg-1Zn alloy was selected.A constant extension rate tensile(CERT)test combined with electrochemical impedance spectroscopy(EIS)testing was performed in a protein-containing biological environment.Thus,this work firstl attempts to characterise the bio-corrosion behaviour of Mg-1Zn alloy under dynamic strain,and may provide a way to quantitatively explore the corrosion resistance of surface layers under dynamic conditions in proteincontaining physiological media.

2.Experimental

2.1.Materials and samples preparation

All the samples used in the present study were manufactured from an extruded Mg-1Zn magnesium alloy,as described in our previous work[27].Cubic samples in a size of 2×10×10 mm3for microstructure,immersion and electrochemical tests,and dumbbell-shaped specimen with 3 mm diameter and 16 mm gauge length for tensile tests were ground using SiC paper from 600 to 2500 grid,ultrasonically washed in distilled water and anhydrous alcohol.The polished cubic specimens were etched using a picric acid-based etchant(5 g picric acid,70 mL ethanol,10 mL acetic acid and 20 mL distilled water)and 3% nitric acid for optical microscopy(OM)measurement.

2.2.Microstructural and composition characterisation

The microstructure and composition of the Mg-1Zn alloy were observed using a metallographic microscope(Leica,DM3000),scanning electron microscope(SEM;Phenom XL)equipped with energy-dispersive spectrometer(EDS).Phases were identifie by X-ray diffraction(XRD,D/Max-2400,Japan)using monochromatic Cu-Kαradiation,at a step size of 0.02°within 2θ=20-90°and a scan rate of 4°?min?1.Electron back-scattered diffraction(EBSD,JeoL JSM-7800F Prime+EDS+EBSD,JEOL Ltd.,Japan)was used at 15 kV accelerating voltage with a step size of 0.25μm.The EBSD samples were taken from the gauge dimension in the unstrained and 4% extension strained samples,mounted with epoxy resin,and polished by a cross section polisher(IB-19530CP,JEOL Ltd.,Japan)for 15 min using Ar gas.

The composition and structure of the corrosion layer on Mg alloy were investigated by Fourier transform infrared spectra(FTIR,Nicolet iS50,Thermo Fisher Scientific after 36 h immersion.To further investigate the chemical compositions,X-ray photoelectron spectroscopy(XPS,ESCALAB 250Xi,Thermo Fisher Scientific was used with Al Kαsource at 1486.6 eV.Binding energies of XPS were analysed depending on carbon(C1 s)at 284.8 eV.

2.3.Immersion and electrochemical tests

Thein-vitrophysiological solution used is the buffered modifie simulated body flui(m-SBF)as proposed in literature[32].The chemical compositions of m-SBF,which only contains inorganic ions,were composed of 5.403 g NaCl,0.426 g KCl,0.504 g NaHCO3,0.230 g K2HPO4?3H2O,0.293 g CaCl2,0.072 g Na2SO4,0.311 g MgCl2?6H2O,17.892 g HEPES(2-(4-(2-hydroxyethyl)?1-piperazinyl)ethanesulfonic acid)[32].The pH=7.4 was adjusted by 1 mol?L?1NaOH solution.Bovine serum albumin(BSA,Shanghai Yuanye Bio-Technology Co.,Ltd),as a natural organic component and as a model protein,has been plentifully used to study the influenc of protein on biocorrosion of biomedical metallic implants[33].Then BSA was added into m-SBF in a physiological concentration of 40 g?L?1based on the concentration of albumin in human serum(35-52 g?L?1)[34](named as m-SBF+BSA).The m-SBF+BSA and m-SBF media were sterile filtere with 0.1 um micro-porous filtration before tests.

The immersion tests were performed in m-SBF and m-SBF+BSA at 37 °C under sterilised conditions.The dissolution of Mg during the immersion tests is related to the volume of hydrogen(H2).The volume of accumulated H2gas of samples was tested using an apparatus,which was introduced in the authors’previous work[35].The ratio of the exposed surface area(cm2)of the specimens to the volume of solution(mL)was kept at 1:40.The average corrosion rate from hydrogen evolution(PAH,mm?y?1)was denoted as[36,37]:

where the hydrogen evolution rate,VH(mL?cm?2?d?1),was assessed as the total volume of evolved hydrogen per unit area divided by total immersion time.After the immersion tests,surface morphologies of Mg alloys were observed using SEM,and the surface composition was measured at 5 different locations by EDS.

An electrochemical workstation(PARSTAT4000,AMETEK,USA)was used to study the potentiodynamic polarisation(PDP)behaviour of the Mg alloy in the solutions at 37±1 °C.It was performed from?0.25 V to 0.25 mV with regard to open circuit potential(OCP)at a scanning rate of 0.5 mV?s?1.The tests were performed in a conventional three-electrode system,where the testing sample,a platinum sheet and a saturated calomel electrode(SCE)acted as the working electrode(1 cm2),the counter electrode and the reference electrode,respectively.The corrosion current density(icorr)and corrosion potential(Ecorr)were derived from the cathodic braches of the PDP slopes[27].

2.4.Constant extension rate tensile tests combined with electrochemical impedance spectroscopy

The evolution of the surface layer on the Mg-1Zn alloy was measured using a constant extension rate tensile(CERT)device coupled with electrochemical workstation,which was presented in our previous work[38].It is used toin-situmonitor the impendence evolution of the samples under a uniaxial increasing tensile strain.The dumbbell-shaped specimen was used in a universal testing machine(UTM;5504X,Shenzhen).Due to the samples immersed in solutions,the strain was not directly measured by a clip gauge.Instead,the crosshead speed(1×10?3mm?min?1)was kept constant assuring a strain rate of 1×10?6s?1in the gauge length.This mode using a very slow strain rate could be regarded as a quasistatic testing,which is conducive to acquire test data.In case that the deformation strain was centralised on the gauge part in tensile direction,the slight deformation of the stainless steel clamps and nylon nuts were not considered.Therefore,the strain-time plots based on crosshead travel were assumed as constant,offering a constant dynamically increasing tensile strain in the gauge length portion of this experiment.

During the CERT tests,electrochemical impedance spectroscopy(EIS)was simultaneously measured at different times without stopping the tensile tests.The impedance measurement range was from 1×105to 1×10?1Hz with 5 mV peak amplitude.The amplitude of 5 mV was chosen to avoid an influenc of the applied potential perturbation on the corrosion of the samples as much as possible.The frequency of 1×10?1Hz was chosen to avoid non-stationarity and pseudo-inductive response trigged at low frequencies,and to shorten the measurement time[39].To guarantee the reproducibility of the data,two or three duplicate samples were tested.

The samples after CERT tests were washed with distilled water,then dried and observed using SEM.ZSimpWin 3.30 software(AMETEK,USA)was used to fi the EIS data.In order to ensure a constant exposed area in the gauge length and to avert galvanic corrosion with other parts of the combined device,the remaining area of the CERT samples in solution was wrapped by using Teflo tapes.The ratio of medium volume(mL)to the exposed surface zone(cm2)was about 50:1.

3.Results

3.1.Microstructure of the Mg-1Zn alloy

The etched and polished samples were analysed using optical microscopy(OM),SEM and XRD,to investigate the basic characteristics and microstructure of the Mg-1Zn alloy,as shown in Fig.1.The average grain size was around 16±13μm,which could endow permissible mechanical properties for the Mg-1Zn alloy.SEM and EDS analysis(inset in Fig.1b)identifie main elements on the surface of samples were Mg,O and Zn in the alloy.α-Mg matrix was identifie by XRD as the main phase(Fig.1c),which indicate only a negligible amount of other secondary phases in the matrix[7].

Fig.1.(a)Metallographic images,(b)SEM,EDS(dashed area)and(c)XRD measurement of the Mg-1Zn alloy.

Fig.2.(a)Change of average corrosion rate(PAH)of the Mg-1Zn alloy with immersion time at 37±1 °C in m-SBF and m-SBF+BSA;Inset:total volume of evolved hydrogen gas;(b)Electrochemical polarisation curves after 0.5 h immersion in solutions at 37 °C;Inset:corrosion current density(icorr)and corrosion potential(Ecorr).

3.2.Degradation rate tests

The average corroson rate(PAH)and total evolved hydrogen gas volume with time(the inset)of the Mg-1Zn alloy immersed in m-SBF and m-SBF+BSA at 37±1 °C are shown in Fig.2a.The result displays a higherPAHof the samples immersed in m-SBF compared to m-SBF+BSA in the preliminary stage and after 15 h of immersion,indicating an overall lower degradation rate in m-SBF+BSA.For a shorter time(2-15 h),thePAHin m-SBF was lower than in m-SBF+BSA.The results suggest that the corrosion rate of the Mg-1Zn alloy is a dynamic parameter during the immersion tests.With prolonged corrosion process,a distinct drop ofPAHwas obversed as a function of time.This reason can be contributed to degradation products forming a protective layer on the Mg-1Zn alloy.

The potentiodynamic polarisation(PDP)curves of the Mg-1Zn alloy after 0.5 h immersion in m-SBF and m-SBF+BSA,as well as corrosion potential(Ecorr)and corrosion current density(icorr)drived from the PDP measurement are exhibited in Fig.2b.It can be seen that the addition of BSA in m-SBF shifted theEcorrfrom?1.63±0.02 V to?1.6±0.01 V,and theicorrdecreased from 105±12μA?cm?2to 33±4μA?cm?2correspondingly.The results suggested that the presence of proteins in physiological environment could decrease the corrosion rate of the Mg-1Zn alloy.

3.3.Constant strain rate tensile tests coupled with impedance analysis

Fig.3 shows the engineering strain/strain rate-time and engineering stress-time plots of the Mg-1Zn alloys tested in m-SBF and in m-SBF+BSA based on cross head movement.The mechanical properties obtained on the samples tested in aqueous solution were different.Thus,the ultimate tensile strength(UTS)and elongation to fracture(ε)of the Mg-1Zn alloy tested in m-SBF+BSA were significan higher than that in m-SBF.

To further investigate the evolution of the degradation product layer during constant extension rate tensile(CERT)tests in m-SBF and m-SBF+BSA solutions,thein-situelectrochemical impedance spectroscopy(EIS)of unstrained and strained Mg alloys were established.

For all the unstrained samples,three capacitive loops can be seen in low,mid and high frequency range from Nyquist and Bode plots,meanwhile only two real capacitive loops present in all strained samples,as shown in Fig.4.To understand the corrosion characterisation of the samples in detail,the impedance spectra data were fitte with respective equivalent circuit models(Fig.5).In the equivalent circuit,Rsis the solution resistance.R2is accounted for the resistance of outer partially protective layer,which grown on the parallel to the capacitance of such layer described by a constant phase element(CPE2).R1and CPE1refer to the resistance and capacitance of interfacial oxide layer.For the strained samples,R2and CPE2do not show in Fig.5b for the strained samples because of the disappearance of the high frequency time constant.Rctrepresents the charge transfer resistance[40].The electric double layer between the Mg alloy and electrolyte is related toCPEdl.In general,CPE are employed to illustrate a non-ideal capacitive behaviour because of surface roughness,variation of fil layer composition,slow adsorption reactions or the inhomogeneity of the system[41,42].

Fig.3.(a)Engineering strain/strain rate-time and(b)engineering stress-time/strain plots of tensile strained Mg alloy in m-SBF and m-SBF+BSA solution at 1×10?6 s?1 constant strain rate.

The corresponding fitte results are listed in Table S1 in supplementary data.To compare and avoid possible controversy,theR1,R2andRctas well as the total resistance(Rsum)of the surface layer are displayed(Rsum=R1+R2+Rct),as shown in Fig.6.For the unstrained tests(shown in Fig.6a and c),it can be seen thatRsumof samples in m-SBF+BSA rapidly increased in the initial firs hour of immersion.In comparison,Rsumin m-SBF increased with immersion time.HigherRsumvalues in m-SBF+BSA than in m-SBF very at the beginning of immersion suggested initial better corrosion resistance of samples in m-SBF+BSA.It is in good agreement with the results obtained from potentiodynamic polarisation(PDP)tests after 0.5 h immersion.With prolonged immersion time(after～1 h),an interesting trend was observed in m-SBF+BSA,whereRsumdropped and then quickly climbed up again after 5.5 h immersion.On the other hand,Rsumvalues in m-SBF gradually increased during immersion time in unloaded condition.The gradual formation,deposition and adsorption of corrosion products or components of the media determine the evolution ofRsumfor all samples as a function of testing time[43].

In order to further illustrate the effect of BSA on the degradation process under corrosion and dynamic constant deformation,the plots of engineering stress-time/strain were inserted in Fig.6b and d.Based on the evolution of impendence and stress,the plots ofRsumcould be divided into 3 periods(Ⅰ,Ⅱ,Ⅲ)for strained samples tested in m-SBF and m-SBF+BSA,respectively.They are related to the evolution of impendence during the low applied stress stage(Ⅰ,0-1 h),the resistance value increasing stage(Ⅱ,about 1-10 h for m-SBF or 1-19 h for m-SBF+BSA),as well as the initiation and propagation of cracks and before fracture stage(Ⅲ,about 10-22 h for m-SBF or 19-32 h for m-SBF+BSA)during constant strain rate process.The results show that the sum impedance of the dynamic strained samples were significantl different compared to the values obtained from the unstrained samples tested in unloaded condition.TheRsumvalues of the strained samples in m-SBF increased slowly duringⅡstage before 10 h,and were lower than that of the unstrained samples(Fig.6b).It can be assumed that the slight deformation of Mg alloys could make a disadvantageous effect on the protective performance of surface layer in m-SBF.

In contrast,Rsumvalues of samples in m-SBF+BSA were higher than that of the strained samples in m-SBF duringⅡstage(Fig.6d).The results demonstrated that the corrosion resistance was enhanced under dynamic deformation of Mg alloys in m-SBF+BSA.This result indicated that the corrosion resistance of the samples tested in m-SBF+BSA was better than that in m-SBF during deformation process.Before fracture failure(Ⅲperiod)of the strained samples,higher decline rate of the impedance in m-SBF than in m-SBF+BSA was observed.It was suggested that large deformation could further deteriorate the corrosion resistance,while BSA could partly reduce the detrimental influenc for Mg alloys.Considering the better mechanical properties obtained from the strained samples tested in m-SBF+BSA,it could make a conclusion that the adsorbed BSA into the surface degradation layer could improve the corrosion and stress corrosion cracking(SCC)of biomedical Mg alloys under the dynamic deformation of samples.

3.4.Electron back-scattered diffraction(EBSD)analysis before and after deformation

Fig.7 shows electron back-scattered diffraction(EBSD)results of the Mg-1Zn alloys before and after plastic deformation.Due to the elongation to failure in m-SBF was about 8%,enough deformation magnitude(4%)was taken to ensure that deformation microstructure could be clearly observed and analysed.Compared to the unstrained sample,plastic deformation sample possessed a higher density of twin boundary,and stored strain zones presented in the twins and their surroundings[13].However,the Volta potential difference for the twinning and its adjacent region can affect the corrosion kinetics,leading to localised corrosion and preferential sites for anodic dissolution during constant strain rate tensile process[44].The orientation maps(as shown in Fig.7b and d)illustrate a slight distinction in the crystallographic orientation of the grains between the unstrained and plastic deformed samples.Therefore,the results demonstrated that the plastic deformation was related to the microstructure evolution and accelerated corrosion rate of HCP structured Mg alloys[45-47].

Fig.4.The in-situ electrochemical impedance spectroscopy(EIS)during the constant extension rate tensile(CERT)testing:(a,b,c)unstrained samples and(d,e,f)strained samples in m-SBF;(g,h,i)unstrained samples and(j,k,l)strained samples in m-SBF+BSA.

3.5.Fourier transformed infrared(FTIR)spectra and X-ray photoelectron spectroscopy(XPS)analysis

In order to further investigate the corrosion composition formed on the surface of Mg alloys tested in m-SBF and m-SBF+BSA,Fourier transformed infrared(FTIR)spectra(Fig.8a)and X-ray photoelectron spectroscopy(XPS)(Fig.8b-g)were performed.The adsorption bands at 550 cm?1,1650 cm?1and 3470 cm?1were related to OH?and/or H2O stretching vibrations,which indicated hydrated species formed in the corrosion products[48,49].Hydrated matter characteristic adsorption bands and stronger peaks at PO43?(at 1030 cm?1)both confirme the presence of phosphates species on the surface of samples in m-SBF[49,50].The bands at 850 cm?1(v2 mode),and from 1350 cm?1to 1500 cm?1corresponding to the stretching vibrations and bending of CO32-ions[49,51],manifesting carbonate compounds formed in the corrosion products.The band around 1650 cm?1is also regarded as the amide I feature of organic molecules[52].Combining with the band at 1545 cm?1(amide II,only detected in m-SBF+BSA),the results confirme the existence of BSA components in the corrosion layer on the Mg alloy surface.

Fig.5.Equivalent circuit models applied to fi the electrochemical impedance spectra(EIS)of the(a)unstrained and(b)strained samples in m-SBF and m-SBF+BSA at different times.

Fig.6.The evolution of the R1,R2,Rct,and sum impedance value(Rsum)of the CERT samples under(a)unstrained or(b)strained in m-SBF,and under(c)unstrained or(d)strained in m-SBF+BSA as a function of testing time or strain.

Fig.8b-g show the analysis results of the XPS spectra.Carbon,oxygen,magnesium,phosphorus and nitrogen elements were existent.Nitrogen(N)was barely detectable for sample immersed in m-SBF,but a high intensity was found in m-SBF+BSA(Fig.8b).Meanwhile,the intensity of phosphorus(P)element was strong in m-SBF,compared to a moderate value in m-SBF+BSA(Fig.8d).As shown in Fig 8c and e,the high resolution spectra ofC1 s can be assigned toC-1(C=O/C=N),C-2(C-O)and C-3(C-C,C=C and C-H)bonds around 284.6,285.9 and 287.8 eV,as well as theN1 s spectra(Fig.8f)to N-1(C-N)and N-2(NH2)around 399.6 and 400.6 eV for the samples immersed in m-SBF+BSA.In comparison,the low resolution C 1 s was measured in m-SBF probably corresponds to C-O and CO32-ions[49,53].These results demonstrate the adsorption of BSA on the corroding Mg alloy surface,which are in agreement with the FTIR results and other literatures[54,55].High intensity of Mg 1 s and Mg 2p spectra(as shown in Fig.8b)confir the existence of Mg2+compounds(such as MgO and Mg(OH)2)for both samples after immersion tests[48,56].These results showed that the composition of the testing solutions determined the chemical elements and phase content of the corrosion layer on the same Mg alloy surface.

Fig.7.Orientation map and twins(red arrow)distributions of the unstrained(a and b)and the samples after 4%plastic deformation(c and d)(For interpretation of the references to color in this figur legend,the reader is referred to the web version of this article.).

3.6.EDS analysis and surface morphologies

Fig.9 displays the EDS analysis and surface morphologies of the Mg-1Zn alloy after 36 h immersion in the two solutions.The main elements detected in the corrosion products were oxygen(O),magnesium(Mg),calcium(Ca)and phosphorus(P).Ca and P elements were necessary to form the hydroxyapatite or calcium phosphates(Ca-P compound).However,the intensity of Ca and P for the samples tested in m-SBF+BSA was less than that in m-SBF.The content of Mg shows the opposite trend.In general,Ca-P compoundscontaining degradation layer could enhance the corrosion resistance of bio-Mg alloys in physiological environment[49].The results suggested that the Ca-P is not the key role for the low degradation rate in m-SBF+BSA.Especially,pitting corrosion was more severe for the Mg alloy exposed to m-SBF(after removing the corrosion products,Fig.9).The surface morphology of the Mg-1Zn samples changes from quite severe localised corrosion mode in SBF to a milder localised attack in SBF+BSA.Uniform corrosion is conducive to avoiding a sudden and sharp drop in the mechanical properties of biodegradable metals[24].The results are consistent with the data presented in FTIR and XPS,confirmin that BSA added in m-SBF determines the chemical composition of the surface corrosion layer.

3.7.Micro-cracks features

In order to further analyse the formation and propagation of micro-cracking in the constant extension rate tensile(CERT)testing,SEM images of cross sections were identifie and shown in Fig.10a and b.The corrosion product around micro-cracking sites were measured using EDS.Narrow micro-cracks perpendicular to the strain direction were all shown for the CERT samples.The branches(as shown by arrows in Fig.10a)and main micro-cracks(Fig.10b)suggested the preferential anodic dissolution and trans-granular cracking behaviour[57].Although the localised corrosion and preferential anodic dissolution might result in emerging of stress concentration points in the deformed Mg alloy[38],the EDS results show that the micro-cracks were sealed with corrosion products for the sample tested in m-SBF+BSA,but none in m-SBF,suggesting the adsorption of protein or corrosion product into the crack tunnels.Adsorbed organic components could retard the formation and propagation of corrosion-assisted cracking,thanks to their physical shielding function.Liu et al.[58]illustrated that the corrosion pits of biodegradable metal would be fille with adsorbed protein,hindering the ions exchange in the micro-cracks.As a result,better mechanical properties were obtained for the CERT samples tested in m-SBF+BSA in comparison with m-SBF(Fig.3b).

Fig.8.(a)FTIR spectra and(b-f)XPS analysis of the surface layer on the Mg-1Zn alloy after 36 h immersion in m-SBF and m-SBF+BSA:(b)entire range of the binding energy survey;(c)representative C 1 s spectral;(d)representative P 2p spectral(inset graph:the at% of P);(e)C 1 s spectra for the samples after immersion in m-SBF+BSA;(f)N 1 s spectra for the sample immersed in m-SBF+BSA.

4.Discussion

In actual applications,biomedical Mg devices are usually exposed to more complicated strain caused by bend,torsion and compression dynamic stresses[59].However,these stresses can cause strains,change of dimensions and an accelerated degradation rate as well as stress corrosion of Mg alloys[47,60].Low degradation rate and stress corrosion susceptibility for the Mg alloys are attributed to the grain refinement high intensity of grain boundaries,as well as the homogenised redistribution and little of the second phase[61,62].Generally,the degradation layer formed on the surface plays a key role on the bio-corrosion behaviour of Mg alloy.Surprisingly,enhanced mechanical properties and corrosion resistance of biomedical Mg-1Zn alloy were detected in organic components-containing synthetic or natural media[27].Therefore,analysing the surface degradation layer of Mg alloys in protein-containing physiological environment is imperative to understand the degradation mechanism under dynamic deformation.

Fig.9.The EDS analysis of the corrosion products and the surface morphologies of the Mg-1Zn alloy after 36 h immersion in m-SBF and m-SBF+BSA.

Fig.10.The cross section SEM features of the constant extension rate tensile(CERT)samples tested in(a)m-SBF and(b)m-SBF+BSA with the EDS analysis of micro-cracking.

In testing media that free of protein or organic components,the aggressive environment especially in presence of chlorides can lead to corrosion of Mg alloys[63].Interestingly,the dissolved Mg2+and inorganic composition from the m-SBF can form the surface corrosion layer,which is mainly composed of insoluble inorganic salts such as Ca-P,Ca/Mg-PO4and(Mg,Ca)-CO3[43,64],often acting as a physical barrier fil[22,65].The total resistance for the unstrained samples tested in m-SBF is increased with time(Fig.6),owing to the formation of this outer layer and a consequent stabilisation of the interface oxide film which confirme by the appearance of an additional time constant at high frequencies(as shown in Fig.4).This oxide fil determines the impedance value at low frequencies.Nevertheless,these inorganic compounds exhibit low ductility and bonding strength[22],and can be delaminated by the plastic deformation.

Fig.11.The speculated schematic illustration of the strain-assisted corrosion mechanism of the unstrained or strained Mg-1Zn alloys tested in m-SBF and m-SBF+BSA:the deformation strain of Mg alloys accelerated corrosion in m-SBF,while deformation strain enhanced the adsorption of protein then improving the corrosion resistance in m-SBF+BSA.

As shown in thein-situelectrochemical impedance spectra(EIS)and constant extension rate tensile(CERT)tests,the disruption of the surface fil on the Mg surface was attributed to the fluctuatio of impendence during the initial deformation stage.These lead to the disappearance of the respective relaxation process from the strained samples.It is worth noting that the increase rate and value in impedance of strained samples tested in m-SBF is significantl lower than that of unstrained samples in the firs 10 h of testing(Fig.6a and b).Activated and fresh surfaces were induced by slip/deformation twinning and grain rotation,which experienced preferential corrosion leading to a high dissolution rate of Mg alloys.The impedance of strained samples even further decreased after 10 h tensile deformation,compared to that of unstrained samples.The results demonstrated that the elastic or plastic deformation strain of the Mg-1Zn alloy had a negative effect on the protective ability of both the inorganic salts layer and the interfacial thin nano sized oxide film

The protein or organic components in natural physiological environment act a pivotal role in the degradation of biomedical Mg alloys[54,64,66].The lower content of Ca and P element in the corrosion products(Fig.9)confirme that the presence of protein affected the formation of the protective co-precipitation layer MgmCan(PO4)x(CO3)y(OH)z[34].Chelate complex formed by the proteins migrated away from the surfaces leading to a decrease of impendence,an increase of corrosion rate and corrosion products at early immersion tests.Thus,a thick corrosion product layer was formed on the surface of the Mg-1Zn alloy,which is consistent with the results of pure Mg tested in protein-containing media[67].The results in the present study indicate that the proteincontaining surface layer exhibits better stability and corrosion resistance,compared with the samples without protein.Zhang et al.demonstrated that the(RCH(NH2)COO)2Mg combined by BSA and Mg2+improved the corrosion resistance of Mg alloy substrate and micro-arc oxidation coating[68].

Notably,the phenomenon that the total impedance value(before 15 h,Fig.6d)of the strained sample was larger than that of the unstrained sample tested in m-SBF+BSA,could be attributed to the mechanically-induced microstructure evolution.It is assumed that the proteins adsorbed on the interface stabilised the primary thin interfacial film The values ofR1andR2in Fig.6 clearly show that the outer inorganic layer was strongly affected by the deformation.So,the BSA should really act at the interface,which then dictated the values ofRsumvia the integrity of the interfacial film In the case of dynamic deformation,the reason why the effect of BSA is more visible can be related to the fact that deformation decreased the barrier properties of the outer layer by defects.Therefore,protein molecules could diffuse into the primary fil much easier,and stabilise the interface by a competitive adsorption on the interfacial MgO layer.

The dynamic evolution of Mg surface enhanced the adsorption of protein on the strained samples,and then improved the corrosion resistance of the surface layer.Hou et al.found that surface properties are the critical factor for protein adsorption on the surface of magnesium,whereas the solution parameters such as pH and temperature are not so effective[31].In addition,the adsorption of proteins on the surface of biological materials is not always harmful.Satzer et al.reported that the chelating or adsorption organic layers may also form a biocompatible and protective film thus improving the cell adhesion and differentiation[69].Our results are in agreement with previous works[22,24,34,70],indicating that protein might be a key role in determining the corrosion behaviour of biodegradable Mg alloys.The speculated schematic illustration of the corrosion mechanism of the unstrained or strained Mg-1Zn samples tested in m-SBF and m-SBF+BSA is shown in Fig.11.

5.Conclusions

In this work,the evolution of degradation layers on Mg-1Zn alloy during dynamic deformation strain in modifie simulated body flui(m-SBF)and protein-containing(Bovine serum albumin,BSA)solution(m-SBF+BSA)was investigated.A lower corrosion rate and a more homogenous degradation of the Mg alloy in m-SBF+BSA was obtained,compared to the samples immersed in m-SBF.The mechanism of improved corrosion resistance in m-SBF+BSA was attributed to the enhanced adsorption of proteins into the surface inorganic-organic degradation layer.The adsorbed protein into the inner fil was improved after the defects in the outer layer formed by the mechanical deformation.The adsorbed protein was in favour for stabilizing the interfacial MgO layer.Competitive adsorption of organic molecules into micro-cracks also can retard the propagation of stress corrosion cracking.A relationship between the initiation and propagation of cracks stages and the respective deformation strain of the Mg alloy,and the evolution of the corrosion resistance of the degradation product layer was established.

6.Data availability

The raw/processed data required to reproduce these find ings cannot be shared at this time as the data also forms part of an ongoing study.

Declaration of Competing Interest

The authors declare no competing financia interest.

CRediT authorship contribution statement

Lianxi Chen:Conceptualization,Formal analysis,Writing-original draft.Cheng Guo:Data curation,Formal analysis.Carsten Blawert:Writing-review & editing.Junjie Yang:Investigation,Methodology.Dongchu Chen:Supervision.Xiaojian Wang:Conceptualization,Resources,Writing-review & editing.Zhentao Yu:Resources.Mikhail L.Zheludkevich:Writing-review & editing.Wei Li:Project administration,Supervision.

Acknowledgment

This work was supported by National Key R&D Program of China(2017YFB0305100,2017YFB0305104),the Science and Technology Planning Project of Guangdong Province No.2017B090903005.Xiaojian Wang acknowledges the finan cial support from Jinan University(No.21620110).Zhentao Yu acknowledges the financia support from Science and Technology Planning Project of Guangdong Province(No.2021A0505030042).Lianxi Chen acknowledges the financia support from Guangdong Basic and Applied Basic Research Foundation(2019A1515110580).

Supplementary materials

Supplementary material associated with this article can be found,in the online version,at doi:10.1016/j.jma.2021.07.002.