study on the mg-li-zn ternary alloy system with improved …lbmd.coe.pku.edu.cn/pdf/2017ab02.pdf ·...

Acta Biomaterialia 62 (2017) 418–433

Contents lists available at ScienceDirect

Acta Biomaterialia

journal homepage: www.elsevier .com/locate /ac tabiomat

Full length article

Study on the Mg-Li-Zn ternary alloy system with improved mechanicalproperties, good degradation performance and different responses tocells

http://dx.doi.org/10.1016/j.actbio.2017.08.0211742-7061/� 2017 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

⇑ Corresponding author at: Department of Materials Science and Engineering,College of Engineering, Peking University, Beijing 100871, China.⇑⇑ Co-corresponding author.

E-mail addresses: [email protected] (Y. Zheng), [email protected] (J. Zhou).

Yang Liu a, Yuanhao Wub, Dong Bian a, Shuang Gao b, Sander Leeflang c, Hui Guo a, Yufeng Zheng a,b,⇑,Jie Zhou c,⇑⇑aDepartment of Materials Science and Engineering, College of Engineering, Peking University, Beijing 100871, ChinabCenter for Biomedical Materials and Tissue Engineering, Academy for Advanced Interdisciplinary Studies, Peking University, Beijing 100871, ChinacDepartment of Biomechanical Engineering, Delft University of Technology, Mekelweg 2, 2628 CD Delft, The Netherlands

a r t i c l e i n f o

Article history:Received 26 April 2017Received in revised form 22 July 2017Accepted 16 August 2017Available online 17 August 2017

Keywords:Mg-Li-Zn alloyBiodegradable metalsCorrosionCytocompatibilityHemocompatibility

a b s t r a c t

Novel Mg-(3.5, 6.5 wt%)Li-(0.5, 2, 4 wt%)Zn ternary alloys were developed as new kinds of biodegradablemetallic materials with potential for stent application. Their mechanical properties, degradation behav-ior, cytocompatibility and hemocompatibility were studied. These potential biomaterials showed higherultimate tensile strength than previously reported binary Mg-Li alloys and ternary Mg-Li-X (X = Al, Y, Ce,Sc, Mn and Ag) alloys. Among the alloys studied, the Mg-3.5Li-2Zn and Mg-6.5Li-2Zn alloys exhibitedcomparable corrosion resistance in Hank’s solution to pure magnesium and better corrosion resistancein a cell culture medium than pure magnesium. Corrosion products observed on the corroded surfacewere composed of Mg(OH)2, MgCO3 and Ca-free Mg/P inorganics and Ca/P inorganics. In vitro cytotoxicityassay revealed different behaviors of Human Umbilical Vein Endothelial Cells (HUVECs) and HumanAorta Vascular Smooth Muscle Cells (VSMCs) to material extracts. HUVECs showed increasing nitric oxide(NO) release and tolerable toxicity, whereas VSMCs exhibited limited decreasing viability with time.Platelet adhesion, hemolysis and coagulation tests of these Mg-Li-Zn alloys showed different degreesof activation behavior, in which the hemolysis of the Mg-3.5Li-2Zn alloy was lower than 5%. These resultsindicated the potential of the Mg-Li-Zn alloys as good candidate materials for cardiovascular stentapplications.

Statement of significance

Mg-Li alloys are promising as absorbable metallic biomaterials, which however have not received signif-icant attention since the low strength, controversial corrosion performance and the doubts in Li toxicity.The Mg-Li-Zn alloy in the present study revealed much improved mechanical properties higher than mostreported binary Mg-Li and ternary Mg-Li-X alloys, with superior corrosion resistance in cell culturemedia. Surprisingly, the addition of Li and Zn showed increased nitric oxide release. The present studyindicates good potential of Mg-Li-Zn alloy as absorbable cardiovascular stent material.

� 2017 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction

From the viewpoint of materials science and engineering, anaddition of lithium (density: 0.534 g�cm3) to magnesium can makea light alloy with high specific strength. With increasing content of

lithium in the magnesium-lithium system, the matrix phasechanges from the single a-phase (0–5.7 wt% Li) with a hexagonalclosed-packed (hcp) crystal structure to the a + b dual phases(5.7–10.3 wt% Li) and then to the single b-phase (>10.3 wt% Li)with a body-centered cubic (bcc) crystal structure. It is well knownthat a material with a bcc structure is in general more ductile thana material with an hcp structure due to the activation of more slipsystems in the former. A crystal structure transition from hcpto bcc leads to an enhanced room temperature ductility of aMg-Li-based alloy up to 80% [1,2], which is superior to the ductility

http://crossmark.crossref.org/dialog/?doi=10.1016/j.actbio.2017.08.021&domain=pdfhttp://dx.doi.org/10.1016/j.actbio.2017.08.021mailto:[email protected]:[email protected]://dx.doi.org/10.1016/j.actbio.2017.08.021http://www.sciencedirect.com/science/journal/17427061http://www.elsevier.com/locate/actabiomat

Table 1Analyzed compositions of the Mg-Li-Zn alloys used in the present study.

Nominal composition (wt%) Analyzed composition (wt%)

Li Zn Mg

Mg-3.5Li-0.5Zn 3.7 ± 0.1 0.51 ± 0.04 BalanceMg-3.5Li-2Zn 3.7 ± 0.1 2.3 ± 0.2 BalanceMg-3.5Li-5Zn 3.6 ± 0.1 4.4 ± 0.1 BalanceMg-6.5Li-0.5Zn 6.2 ± 0.2 0.81 ± 0.04 BalanceMg-6.5Li-2Zn 6.8 ± 0.1 2.1 ± 0.1 BalanceMg-6.5Li-4Zn 6.8 ± 0.1 3.5 ± 0.1 Balance

Y. Liu et al. / Acta Biomaterialia 62 (2017) 418–433 419

of any other Mg-based alloys and significant for application whereductility is of primary importance, for example, for vascular stents.However, the corrosion performance of Mg-Li-based alloysremains a highly controversial issue in the papers published. Areduction in corrosion resistance is basically related to the highchemical reactivity of lithium, which limits the application ofMg-Li-based alloys [3,4]. However, the high corrosion susceptibil-ity of Mg-Li-based alloys can be suppressed to a certain extent. Aspecial thermal and mechanical treatment has recently been suc-cessfully applied to a Mg-Li-Al-Zr-Y alloy with a Li content of10.95 wt% to form complex surface layers, leading to corrosionresistance better than pure Mg or other hcp-structured Mgalloys [5].

In the biomedical field, lithium has been adopted in the treat-ment of bipolar disorder [6] and lithium salts are believed to behelpful in diagnosing schizoaffective disorder, cyclic major depres-sion [7] and cluster headaches [8]. In addition, the role of lithium incardiovascular treatment has in recent years been recognized. Forexample, Li has been found to induce the release of nitric oxide(NO) in brain vascular cells [9,10]. Mg-Li-Ca and Mg-Li-Al-RE alloyshave been reported to be promising biodegradable alloys,especially for stent application. The dual phases in theMg-9.29Li-0.88Ca alloy have been reported to be associated witha multi-layer structured surface film, leading to a corrosion processin Hank’s solution different from that of common Mg alloys withone phase being attacked and another being protected [11].Mg-Li-based alloys with the addition of rare earth elements whichwere intended for stent application have been found to performwell in corrosion resistance during static immersion tests andpossess good cytocompatibility [12].

Zinc is well known for being an essential element in humannutrition. It is associated with numerous biological functions suchas nucleic acid metabolism, signal transduction, gene expression,apoptosis regulation and endocrine regulation [13]. Furthermore,zinc has been found to possess physiological corrosion resistanceideal for biodegradable stents [14]. Zn-Li binary alloys have shownnot only good corrosion resistance but also good mechanical prop-erties, both of which are needed for stent applications [15]. More-over, age-hardening effects have been found after the combinedaddition of Li and Zn in Mg, due to stable h phase (MgLiZn) precip-itation in the a-phase alloy and metastable h’ phase (MgLi2Zn) pre-cipitation in the a + b dual phase alloy [16]. From the perspectiveof material science, an addition of zinc to magnesium may leadto improved mechanical properties and corrosion resistance aslong as its concentration does not exceed a certain limit [17,18].

In this study, a series of Mg-Li-Zn alloys were designed, andtheir microstructure, surface characteristics, mechanical proper-ties, corrosion performance, oxidation behavior, in vitro cytocom-patibility as well as in vitro hemocompatibility werecomprehensively investigated. The study was aimed at exploringthe feasibility to develop a Mg-Li-Zn alloy system specifically forstent applications where biodegradability, biocompatibility, hemo-compatibility, strength and ductility are all required.

2. Materials and methods

2.1. Material preparation

Mg-(3.5, 6.5 wt%)Li-(0.5, 2, 4 wt%)Zn alloys were prepared inZhengzhou Light Metals Research Institute, China Aluminum Cor-poration. The analyzed compositions are shown in Table 1. Highpurity Mg (99.98%) was used as a control group. Ingots of the alloysand pure Mg were extruded at 280 �C and at a reduction ratio of 16into bars. Disk-shaped samples for microstructure, surface chem-istry, corrosion tests, in vitro cell viability and hemocompatibility

(with sizes of U10 � 2 mm3) were cut perpendicular to the extru-sion direct from the extruded bars and mechanically polished, fol-lowed by cleaning and drying in air.

2.2. Microstructure and surface characterization

An optical microscope (BX51M, Olympus) was used to observethe microstructure of the alloys prepared. Polished samples wereetched in a 2% nitric acid alcohol solution and rinsed in distilledwater. An X-ray diffractometer (XRD DMAX 2400, Rigaku) withCu Ka radiation and at a scan rate of 4�/min was adopted to detectthe phase compositions of the samples. The surface chemistry wasdetermined with an X-ray photoelectron spectroscope (XPS AxisUltra, Kratos Analytical) with Al Ka radiation. High resolution nar-row scanning was conducted to determine the binding states of Mg2p, Li 1S, Zn 2p, O 1s and C 1s. Surface morphologies were observedwith an SEM coupled with EDS (S-4800 Emission Scanning ElectronMicroscopy, Hitachi).

2.3. Mechanical tests

Tensile specimens were cut from the extruded bars andmachined according to ASTM-E8-04a to have a gauge length of25 mm. Tensile tests were carried out at a crosshead speed of1 mm/min and at room temperature using an Instron 5969machine. At least three specimens for each material were tested.

2.4. Electrochemical evaluation

Electrochemical evaluation was performed by using a three-electrode cell with a platinum plate as the counter-electrode anda saturated calomel electrode (SCE) as the reference electrode.The electrochemical tests were carried out in Hank’s solution (NaCl8.00 g�L�1, KCl 0.40 g�L�1, CaCl2 20.14 g�L�1, NaHCO3 0.35 g�L�1,glucose 1.00 g�L�1, MgCl2�6H2O 0.10 g�L�1, MgSO4�7H2O 0.06 g�L�1,Na2HPO4�12H2O 0.06 g�L�1 and KH2PO4 0.06 g�L�1, pH = 7.4) atroom temperature with an electrochemical workstation (Autolab,Metrohm). The surface of a round specimen with a diameter of5 mm was exposed to Hank’s solution. Each specimen first under-went open-circuit potential (OCP) for 4800 s. Afterwards, potentio-dynamic polarization was performed at a scanning rate of1 mV s�1. Corrosion parameters including open-circuit potential(OCP), corrosion potential (Ecorr) and corrosion current density(icorr) were estimated from the polarization curves by means ofTafel analysis based on the polarization plots. Since the determina-tion of the Tafel slope might cause large variations [19], theTafel slopes were carefully determined in the potential range of130–300 mV away from Ecorr both on the cathodic curve and theanodic curve. At least three duplicates were tested for each alloy.

2.5. In vitro immersion and hydrogen evolution

Immersion tests with 3 duplicates were performed with a ratioof solution volume to sample surface area (V/S) at 20 ml�cm�2,

420 Y. Liu et al. / Acta Biomaterialia 62 (2017) 418–433

according to ASTM G31-72, in Hank’s solution at 37 �C. Hydrogenevolution tests with the same V/S ratio at 37 �C were conductedusing a setup similar to the schematic illustration presented else-where [20]. The pH value and the volume of hydrogen evolvedwere recorded at each time point. At 3 and 20 d, samples weretaken out of the solution, immersed in deionized water and driedin air. Surface and interface morphologies were observed withSEM and the ion concentrations of the solution after 20 days ofthe immersion tests were determined with an Inductively CoupledPlasma-Atomic Emission Spectrometer (ICP-AES, PROFILE SPEC,Leeman). In addition, surface morphologies as well as cross sectionmorphologies before and after natural oxidization were investi-gated with SEM coupled with EDS, operating in the second electron(SE) mode and the backscattering electron (BSE) mode. The func-tional groups in the corrosion products were detected by meansof microscopic Fourier Transform Infrared analysis (Spotlight 200Micro-FTIR, PerkinElmer). The spectra were recorded from 4000to 650 cm�1. Small angle X-ray diffraction was conducted by usingX’Pert PRO MDP (Panalytical) with Cu Ka radiation at 0.5� over ascan range from 10 to 90�.

2.6. Cell viability assay, cell morphology cell apoptosis and NO release

Cell viability and proliferation evaluation was conductedaccording to ASTM 10993-5: 2009. Human Umbilical VeinEndothelial Cells (HUVECs) and Human Aorta Vascular SmoothMuscle Cells (HA-VSMCs) were utilized. Cells were obtained fromAmerican Type Culture Collection (ATCC) and maintained in ourlaboratory. Cells were separately cultured in Dulbecco’s ModifiedEagle’s Medium (DMEM) supplemented with 10% fetal bovineserum (FBS), 100 U�ml�1 penicillin and 100 g�ml�1 streptomycinin a humidified atmosphere with 5% CO2 at 37 �C. Cell viability,morphology and NO release were all determined with the indirectmethod. Polished samples were washed, dried in air and sterilizedvia Co60 c ray radiation. Extracts were generated by incubating thesamples in DMEM supplemented with 10% FBS for 24 h at anextraction ratio of 1 cm2�ml�1 at 37 �C. Ion concentrations in thegathered extracts were determined using an Inductively CoupledPlasma Optical Emission Spectroscope (ICP-OES, iCAP6300,Thermo). The extract of the as-extruded pure Mg (99.98% purity)was utilized as the material control, the cell culture medium asthe negative control and the culture medium added with 10%dimethylsulfoxide (DMSO) as the positive control.

As to the in vitro viability assay, cells were seeded in 96-wellplates at a density of 5 � 103 cells per 100 lL medium and incu-bated for 24 h to allow attachment. Then, the cell culture mediumwas replaced by the extracts. At 1, 3 and 5 d before testing, theextracts were replaced by the cell culture medium to avoid inter-ference [21]. The cell viability was tested via Cell Counting Kit-8(CCK-8, Dojindo Molecular Technologies, Japan). 10 lL CCK-8 wasadded to each well and restored in an incubator for 1 h. The spec-trophotometric absorbance of each well was measured with amicroplate reader (Bio-RAD 680) at a single wavelength of450 nm. At least 6 duplicates were adopted for the in vitro viabilityassay. Cells for apoptosis evaluation were harvested and washedwith cold PBS. The cell density was adjusted to 106/ml with 1�Binding Buffer. 5 lL Annexin V-FITC and 10 lL Propidium Iodide(Biolegend, Beijing Korad Biotechnology) was added to 100 lL cellsuspension and reacted for 15 min at dark place and room temper-ature. 400 lL 1� Binding Buffer was added before testing on flowcytometry.

The influence of extracted ions on cell morphology was evalu-ated with a Laser Scanning Confocal Microscope (A1R-si, Nikon)and a glass slide with a thickness of 0.17 mm. 0.5 ml of cell suspen-sions in 24-well plate were placed on a glass slide at a density of5�104 cells per milliliter and incubated for 24 h to allow

attachment to the glass slide. Then, the media were replaced byextracts and incubated for 24 h. After that, the HUVEC cultureextracts were obtained for NO release tests and both HUVECsand VSMCs were thoroughly washed by PBS and fixed by 4%paraformaldehyde for 10 min at room temperature. 0.1% Tritonwas then added and reacted for 7 min. 5 lg�ml�1 FITC-phalloidindye for microfilament was added to the PBS-rinsed well and incu-bated for 30 min at room temperature. 10 lg�ml�1 DAPI dye fornuclei was then added and incubated for 10 min at 37 �C. The cellmorphology was observed with a confocal microscope at excitationwavelengths of 405 and 488 nm. Three duplicates were utilized forobservation. The NO release tests of the 24 h culture extract ofHUVECs were conducted using an NO detection kit (NanjingJiancheng Bioengineering Institute), following the protocol. Threeduplicates were used for the NO release tests.

2.7. Hemolysis and platelet adhesion

Human blood from a healthy volunteer was anticoagulated with3.8 wt% sodium citrate at a ratio of 9:1 and diluted with normalsaline at a volume ratio of 4:5. Samples were then incubated in10 mL normal saline at 37 �C for 30 min. Then 0.2 mL diluted bloodwas further incubated for another 60 min. Normal saline anddeionized water were set as the negative group and positive group,respectively. All tubes were centrifuged at 1000�g for 5 min andthe supernatant was carefully transferred to 96-well plate. Absor-bance (Optical Density, OD) was read by a microplate reader (Bio-RAD 680) at a single wavelength of 545 nm. The hemolysis was cal-culated according to the following equation:

Hemolysis rate ¼ OD ðtesting groupÞ � OD ðnegative groupÞOD ðpositive groupÞ � OD ðnegative groupÞ � 100%

Adhesion and activation performance of the platelets weredetermined through the platelet adhesion test. Platelet-RichPlasma (PRP) was prepared by centrifuging anticoagulated wholeblood at 112�g (1000 rpm on our centrifuge) for 10 min. 0.2 mLPRP was dropped on the surface of sample and incubated for60 min at 37 �C. Afterwards, PBS was used to rinse surface, fol-lowed by platelet fixation for 2 h in a 2.5 % glutaraldehyde-PBSsolution, dehydrating in gradient ethanol/distilled water solutions(50, 60, 70, 80, 90 and 100%) for 10 min each and drying in air. Themorphologies of adhered platelets were observed by using SEM(S4800, Hitachi).

2.8. Coagulation four

The activated partial thromboplastin time (APTT), prothrombintime (PT), fibrinogen (FIB-C) and thrombin time (TT) were mea-sured in a coagulometer (ACL TOP, Instrumentation LaboratoryCo). Fresh human whole blood was centrifuged at 1000�g for15 min to prepare platelet poor plasma (PPP). Afterwards, PPPwas transferred to an anticoagulative tube without sodium citrate.The four biochemical indexes were measured at Peking UniversityPeople’s Hospital.

2.9. Statistical analysis

Statistical analysis was performed with SPSS 18.0 for WindowsSoftware (SPSS Inc., Chicago, USA). One-way analysis of variance(ANOVA) followed by Tukey post hoc tests was used to statisticallyanalyze all the data. A p-value

Fig. 1. Microstructures of Mg-3.5Li-0.5Zn, Mg-3.5Li-2Zn, Mg-3.5Li-4Zn, Mg-6.5Li-0.5Zn, Mg-6.5Li-2Zn, and Mg-6.5Li-4Zn from (a) the cross-section perpendicular to theextrusion direction, (b) along with the extrusion direction, (c) XRD results and (d) EDS confirming the presence of Zn-containing particles.



3.Results

3.1. Microstructures

The microstructure of the materials is shown in Fig. 1a. Afterextrusion, all the Mg-Li-Zn showed a wavy structure in the crosssection perpendicular to the extrusion direction, which was causedby the severe deformation of extrusion. According to previousstudy, the microstructure presented in Fig. 1, with grains aggre-gated by the continuous interfolding of neighboring grains, wasnamed curling [22] or the Van Gogh Sky pattern [23,24], whichwas caused by diffuse strain heterogeneities arise from the con-comitant deformation of grains of different strength. Besides, withthe Li content being 3.5 wt%, the alloys are a phase-based withdark particles which increased with increasing Zn content. Thedark Zn-containing particles along the wavy structure changedfrom small particles and large aggregates in Mg-3.5Li-0.5Zn, tofinely distributed particles in Mg-3.5Li-2Zn to evenly distributedparticles in Mg-3.5Li-4Zn. As the content of Li increased to 6.5 wt%, both the light a-phase and dark b-phase were observed (Mg-6.5Li-xZn). As for the microstructure along with the extrusiondirection, all the alloys showed phase and particles elongatedalong with the extrusion direction. XRD results (Fig. 1b) confirmedthe single a phase in the Mg-3.5Li-xZn alloys and the dual a + bphases in the Mg-6.5Li-xZn alloys. However, no Zn-containingphase was detected in XRD even when the addition of Zn to Mgwas 4 wt.%. It might be because Zn has large solid solubility in bothMg and Li, leaving Zn-containing particles of a small volume frac-tion not detectable by XRD. Nevertheless, SEM coupled EDS con-firmed the presence of Zn-containing compounds in the form ofaggregated particles embedded in the heavily deformed matrix.As shown in Fig. 1c, all the small parallel particles and the largeraggregation marked by red arrow are Zn-containing particles.Compared current results with previously study [16], the Zn-containing particles may be h phase (MgLiZn) or h0 phase(MgLi2Zn).

3.2. Surface characteristics

The surface chemistry of the alloys was determined using XPS.The XPS files of Mg-3.5Li-2Zn and Mg-6.5Li-2Zn are shown inFig. 2, while the XPS spectra of the other Mg-Li-Zn alloys are shownin Figure S1. In general, C, O, Li and Mg were detected, with con-centration and chemical valence variations depending on alloycomposition. The relative broad peak of O 1s indicated a mixtureof O at different chemical states. Interestingly, from all of the XPSspectra, no sign of zinc was found on the surface, indicating a verylimited amount of Zn on the naturally formed surface film. In theMg-3.5Li-based alloys, more Mg was detected on the surface(Fig. 2a, Fig. S1d and S1h), whilst in the Mg-6.5Li-based alloys amuch stronger peak of Li 1s was detected (Fig. 2b, Fig. S1l andS1p). Although no Zn was detected on the surface, the surfaceMg/Li ratio reached its highest with Zn at 2.0 wt% in both theMg-6.5Li-xZn and Mg-3.5Li-xZn alloys. To reveal chemical statedifferences, the detected elements were all further investigatedwith high resolution scan. In the Mg-3.5Li-2Zn alloy, one Li 1s peakassigned to 55.1 eV, accompanied by C 1s at 290.1 eV and O 1s at531.7 eV, was detected, suggesting the presence of Li2CO3 [5,25].In addition, the broad peak range of O 1s suggested different chem-ical states of O present on the surface. The results also confirmedthe predominant presence of Mg-based compounds, includingMgO/Mg(OH)2 and/or MgCO3 [26,27]. With Li content increasing,more Li appeared on the surface and Li2O (Li 1s 54.0 eV) insteadof MgCO3 appeared on the surface of the Mg-6.5Li-2Zn alloy, inaddition to Li2CO3 and MgO [28]. The other two Mg-6.5Li-xZn

alloys showed the similar phase compositions on the surface. Itshould be mentioned that although no sign of LiOH was found, itwas considered to be existed due to the absorption of H2O toLi2O. In general, our XPS results confirmed the surface chemistry ofthe alloys with Li2CO3 and Mg-based compounds as the main phases.Furthermore, Zn seemed to act as a mediator to influence the surfaceMg/Li ratio but did not interfere the surface phase composition.

3.3. Mechanical properties

The tensile properties of the as-extruded Mg-Li-Zn alloys areshown in Fig. 3. The combined addition of Li and Zn greatlyimproved the yield strength (YS), ultimate tensile strength (UTS)and elongation of pure magnesium. All the as-extruded Mg-Li-Znalloys showed curling microstructure with high strain in the grainsand caused continuous interfolding of grains, which implied highwork hardening in the alloys. Besides the severe deformationcaused by extrusion, the existence of the different phases providesfundamental hardening effects after the addition of Li and Zn. Forexample, Zn-containing particles/precipitations provided harden-ing effects both in the a phase based Mg-Li-Zn alloys and a + bphase based Mg-Li-Zn alloys [16]. The YS and elongation valuesof the as-extruded Mg-Li-Zn alloys were at least two times as highas those of the control pure Mg. More specifically, the UTS values ofMg-3.5Li-2Zn and Mg-3.5Li-4Zn were around 250 MPa (i.e., 246 ± 2and 250 ± 10 MPa, respectively), up 48% from the value of pure Mg.With respect to elongation, both the Mg-6.5Li-2Zn and Mg-6.5Li-4Zn alloys reached high values, i.e., 29 ± 2% and 35 ± 1%, respec-tively, representing the increases of 2 and 3 times from the valueof pure Mg (12 ± 1%). The increase in mechanical properties couldbe attributed to the microstructural changes brought about by thealloying elements and by the extrusion processes that the alloyswent through. In the case of the Mg-3.5Li-xZn alloys, the matrixis still the hcp-structured a-Mg phase with solute Li andZn-containing compounds. Therefore, the improvements in tensilestrength and elongation must be due to the hot extrusion processwhich led to high strain residual, grain refinement, in addition tosolution strengthening and dispersion strengthening. TheMg-6.5Li-xZn alloys contain both the hcp-structured a-phase andbcc-structured b-phase. In the bcc-structured b-phase more slipsystems are activated at room temperature, as compared withthe hcp-structured a-phase, which made these alloys easy todeform and attain exceptionally large elongation values. Thegains of the alloys with the duplex structure in ductility areaccompanied by the losses in YS and UTS, but the overall levelsof YS and UTS are about 3 times as high as those of commonbiodegradable polymers [29].

3.4. Electrochemical evaluations

Table 2 summarizes the electrochemical parameters obtainedfrom the electrochemical tests in Hank’s solution. Since Li and Znrespectively has more negative (Li: �3.04 V) and nobler (Zn:�0.76 V) standard hydrogen electrode potentials than Mg(�2.37 V) [30], the combined addition of these two elementsresulted in OCP and corrosion potential values quite similar tothose of the control pure Mg. However, the addition of Zn and Lihad more effects on corrosion rate and thus different corrosionbehavior and corrosion resistance. Corrosion rate often showed lar-ger influence on the corrosion behavior than corrosion potential.The Mg-3.5Li-2Zn alloy showed the best corrosion resistance inthe Mg-3.5Li-xZn alloy system, whereas the Mg-6.5Li-2Zn alloyexhibited corrosion resistance inferior to the Mg-6.5Li-0.5Zn alloy,implying the joint influence of Li and Zn. With reference to themicrostructure shown in Fig. 1, the single a-phase together withfinely distributed Zn-containing particles may account for the best

Fig. 2. XPS spectra of (a) Mg-3.5Li-2Zn and (b) Mg-6.5Li-2Zn.

Fig. 3. Yield strength (YS), ultimate tensile strength (UTS) and elongation values ofas-extruded Mg-Li-Zn alloys.


corrosion resistance of the Mg-3.5Li-2Zn alloy determined fromthe electrochemical tests. Besides, the dual phases of a and bshowed higher corrosion rate with their single phase counterpart(fixed Zn addition with variation of Li content).

The potentiodynamic polarization curves of the Mg-Li-Zn alloystogether with the curve of pure Mg as the control group are shownin Fig. 4a. In addition to the shifts of corrosion potential, the Mg-Li-Zn alloys exhibited sharper cathodic reaction curves. From theresults of potentiodynamic polarization, it was clear that, amongall the Mg-Li-Zn alloys, the Mg-3.5Li-2Zn alloy possessed the low-est corrosion current density. However, the corrosion resistanceinferior to that of pure Mg in the electrochemical tests showed thatthe presence of a Li2CO3-containing film on the surface could notprovide effective protection from the variable potential applied.The other compounds in the surface film originating from thematrix must have weakened the protecting effect of the surface film.

Table 2Open circuit potential, corrosion potential (Ecorr), corrosion current density (Icorr) and corrosion rate values obtained from the electrochemical tests.

Open circuit potential (VSCE) Ecorr (VSCE) Icorr (lA�cm�2) Corrosion rate (mm�y�1)Pure Mg �(1.63 ± 0.04) �(1.56 ± 0.01) 7.8 ± 0.6 0.18 ± 0.01Mg-3.5Li-0.5Zn �(1.61 ± 0.03) �(1.55 ± 0.04) 15 ± 1 0.34 ± 0.03Mg-3.5Li-2Zn �(1.62 ± 0.01) �(1.59 ± 0.01) 10.6 ± 0.3 0.24 ± 0.01Mg-3.5Li-4Zn �(1.58 ± 0.01) �(1.54 ± 0.01) 14 ± 2 0.32 ± 0.04Mg-6.5Li-0.5Zn �(1.61 ± 0.01) �(1.56 ± 0.03) 15 ± 3 0.34 ± 0.07Mg-6.5Li-2Zn �(1.59 ± 0.01) �(1.55 ± 0.01) 13.1 ± 0.2 0.30 ± 0.01Mg-6.5Li-4Zn �(1.61 ± 0.02) �(1.57 ± 0.01) 12.4 ± 0.7 0.28 ± 0.02

Fig. 4. (a) Potentiodynamic polarization, (b) hydrogen evolution with immersion time, (c) changes of the pH value of Hank’s solution with immersion time and (d) ionconcentrations of Hank’s solution at 20 d with pure Mg as the control.


3.5. Immersion tests and hydrogen evolution

The results obtained from the in vitro immersion tests withrespect to hydrogen evolution, pH change and ion concentrationsare shown in Fig. 4b–d. It is obvious that when 4.0 wt%Zn wasadded to magnesium, the quantities of evolved hydrogen weremuch higher than those of the other Mg-Li-Zn alloys with lowerZn contents. The Mg-3.5Li-4Zn alloy showed a high reactivity inHank’s solution and produced over 8 ml�cm�2 hydrogen after240 h immersion. Being in accordance with the results from theelectrochemical tests, the corrosion resistance of the alloys didnot show a monotonic variation with the Zn content when the Licontent was kept unchanged. The alloys containing 0.5 wt%Znexhibited weaker corrosion resistance, compared to their counter-parts containing 2 wt%Zn. The total hydrogen evolved from

Mg-3.5Li-2Zn and Mg-6.5Li-2Zn alloy samples over a period of240 h were 0.11 ± 0.01 ml�cm�2 and 0.35 ± 0.01 ml�cm�2, respec-tively, both of which were lower than the hydrogen produced frompure Mg sample (0.41 ± 0.02 ml�cm�2) (p < 0.05).

The pH values of Hank’s solution with Mg-3.5Li-2Zn and Mg-6.5Li-2Zn alloy samples were quite comparable and the trend withimmersion time was similar to that of pure Mg (p > 0.05). The ionconcentrations after 20 d immersion in Hank’s solution were quiteconsistent with the results in hydrogen evolution. The Mg2+ con-centrations of Mg-3.5Li-2Zn and Mg-6.5Li-2Zn alloy samples werein the range from 80 lg�mL�1 to 90 lg�mL�1, which is comparableto the high purity Mg control. The Mg-3.5Li-4Zn alloy released thelargest amount of Mg2+, which was 673.4 ± 78.6 lg�mL�1 at day 20.It was only the Mg-3.5Li-4Zn alloy that released more Zn2+ than Li+.The Mg-6.5Li-4Zn alloy released more Li+ than Zn2+, the Zn2+


concentration (3.7 ± 0.9 lg�mL�1) was much higher that Zn2+ con-centrations from the other Mg-Li-Zn alloys (all being around1 lg�mL�1, with 0.5 ± 0.2 lg�mL�1 being the lowest from theMg-6.5Li-2Zn alloy). The Li+ concentrations from most of theMg-Li-Zn alloys were in the range from 10 lg�mL�1 to 20 lg�mL�1,with the lowest at 5.7 ± 0.9 lg�mL�1 from the Mg-3.5Li-2Znalloy. Comparison between the Mg-3.5Li-2Zn alloy and theMg-6.5Li-2Zn alloy revealed a opposite relationship in theconcentrations of Li+ and Zn2+; the Zn2+ concentration fromthe Mg-3.5Li-2Zn alloy was double the Zn2+ concentration from theMg-6.5Li-2Zn alloy (1.0 ± 0.5 lg�mL�1 versus 0.5 ± 0.2 lg�mL�1)but the Li+ concentration from the former was only halfthe Li+ concentration from the latter (5. 7 ± 0.9 lg�mL�1 versus12.1 ± 3.4 lg�mL�1).

The general surface morphologies of samples immersed inHank’s solution for 20 d are shown in Fig. S2, with a detaileddescription given in the Supplemental materials. Fig. 5 shows thesurface morphologies and elements detected on sample surface,as well as the morphologies of corrosion products observed onthe cross section. The Mg-3.5Li-2Zn alloy showed a surface withlimited amounts of corrosion products at 3 and 20 d. It’s obviousthat the deposition of Ca and P from Hank’s solution on sample sur-face increased over time, which was also observed on the surface ofpure Mg.

On the surfaces of Mg-3.5Li-0.5Zn and Mg-3.5Li-4Zn alloy sam-ples there were a lot of corrosion products aggregated at 3 d. After20 d immersion, however, the surfaces changed. Mg-3.5Li-0.5Znalloy sample had a thick coating layer mainly composed of Mg, Oand C, whilst the corrosion products on the surface of Mg-3.5Li-4Zn alloy sample were globular. On the cross-section, pure Mgsample showed typical characteristics of pitting corrosion throughthe surface layer, together with the formation of corrosion prod-ucts. Mg-3.5Li-2Zn sample exhibited a much clearer and uniformcorrosion layer, as compared with pure Mg sample, which was con-sidered to be the result of a change in corrosion mechanism.Despite similar macroscopic local corrosion morphologies betweenMg-3.5Li-2Zn and the other Mg-Li-Zn alloys (Fig. S2), the cross-section of Mg-3.5Li-2Zn sample had much shallower corrosionbeneath the surface, as compared with the other alloy samples thatsuffered from severer local corrosion. For Mg-3.5Li-0.5Zn, a fewsevere corrosion sites were visible, as shown in Fig. 5, and loosecorrosion products of large amounts were composed mainly of

Fig. 5. Surface morphologies, elements on surface detected at 3 and 20 d, an

Mg and O (data not shown). The visible holes in the corrosion prod-ucts indicated the possibility of the penetration of Cl� and otherions. By comparison, the corrosion layer of Mg-3.5Li-4Zn alloysample showed a flake-like structure with open space for ion pen-etration. Surprisingly, Mg-6.5Li-xZn alloy samples exhibited a sim-ilar trend to Mg-3.5Li-xZn alloy samples. Mg-6.5Li-2Zn alloysample had a shallow corrosion layer on the surface, whereasMg-6.5Li-0.5Zn and Mg-6.5Li-4Zn alloy samples showed the pene-tration of corrosion from the surface.

Fig. 6 shows the surface chemistry of Mg-Li-Zn alloy samplesafter 20 d immersion. The infrared spectra of the involved func-tional groups are shown in Fig. 6a. The results revealed the pres-ence of CO32� and PO43� (and/or HPO42�) in the corroded area onthe surface. In fact, it was difficult to distinguish the functionalgroups over a certain range of wavenumbers. The defined 830–890 cm�1 wavenumber range has been reported to be the CO32�

(v2) band [31-33], but other reports have shown the IR band ofHPO42� in a similar range [34,35]. Furthermore, an overlappingrange of 960–1150 cm�1 between PO43� and HPO42� [36,37] hasbeen reported. Further XRD investigation confirmed the presenceof Mg(OH)2 as well as MgCO3, as shown in Fig. 6b. The dashed lineand circled peaks in Fig. 6b might indicate the presence ofMgHPO4�3H2O (newberyite) and/or Mg3(PO4)2 since no other peaksat low 2h angles were found. Further XPS results proved the pres-ence of Mg, O, C and P in the corrosion products. High resolutionnarrow scan confirmed the presence of MgCO3 and Mg(OH)2, andthe detected peaks were in good agreement with the findings givenin the literature. P showed a single peak at 133.6 eV. Based on theinformation in the literature, this peak might be related either toPO43� [38,39] or to HPO42� [40], which made the P-containing com-pound uncertain. Nevertheless, no Ca was detected from XPS.

3.6. Cell morphology, cell viability assay, cell apoptosis and NO release

Cell morphologies of VSMCs cultured in different extracts areshown in Fig. 7a. VSMCs showed benign responses to all the testedmaterials at 24 h and had good spreading morphologies in numer-ous directions, as well as good cell-to-cell connection and visiblestained cytoplasmic filament. VSMCs all perform spindle or polyg-onal morphologies, which was in good coherence with previousreport [41]. However, the viability of VSMCs cultured in theextracts continuously decreased with time (Fig. 7c). This suggests

d the morphologies of corrosion products on the cross section at 20 d.

Fig. 6. Characterization of surface chemistry after 20 d immersion. (a) FTIR files, (b) XRD results of Mg-3.5Li-4Zn and Mg-6.5Li-4Zn, (c) XPS results of the corrosion productson Mg-6.5Li-4Zn with high resolution XPS files of C, O, Mg and P (no Ca found from XPS).


a continuing and lagged response of VSMCs. Previous cytotoxicitytests also indicated the toxicity of Mg-Li-based materials forVSMCs [12]. Ion concentrations in the DMEM extracts were deter-mined (Fig. 7e). The results showed lower amounts of releasedMg2+ in all the Mg-Li-Zn alloy extracts except that of the Mg-3.5Li-4Zn alloy, as compared with those in pure Mg extracts,implying relative better corrosion resistance of the Mg-Li-Zn alloysin the cell culture environment. The inhibition of VSMC is consid-ered to be ion-related. Despite Mg2+ concentrations (around 4 mM)is in the safe range (below 40 mM) which showed no side-effects tohuman VSMC cells [42], LiCl was reported to suppress the prolifer-ation of rat VSMC cells [43]. Our results suggested that 1 mM mayalso cause inhibition. The toxicity of all materials except Mg-3.5Li-4Zn are slight and acceptable (Grade 1 toxicity). Moreover, withregard to vascular stenting application, the late proliferation ofVSMCs leads to restenosis. Ideally, the material for stent shouldprovide benign response to VSMCs for recovery at first and then

inhibit the proliferation of VSMCs to prevent restenosis. The slighttoxicity of Mg-Li-Zn alloys may retard the recovery process, butprovide chances for later inhibition. Cell morphologies of HUVECscultured in different extracts for 24 h are shown in Fig. 7b. The cul-tured cells exhibited healthy cobblestone or polygonal morpholo-gies as those cultured in the culture medium, which was in goodcoherence with previous studies [44,45]. Furthermore, the 5 dresults indicated that the ion concentrations as well as the pH val-ues were tolerable for HUVECs, which is different from the resultsfrom the viability assay of VSMCs (compare Fig. 7c with Fig. 7d).Apoptosis results (Table S1) revealed very limited early apoptosis(LR) of HUVEC cultured in all the extracts except for Mg-6.5Li-4Zn. As for VSMC, apoptosis results showed a slight increase ofearly apoptosis percentage compared with the results via culturemedia.

Nitric oxide (NO) has been widely accepted as endothelium-derived relaxing factor (EDRF) to relax the surrounding smooth

Fig. 7. Actin-nucleus co-staining of (a) VSMC morphologies and (b) HUVEC morphologies after 24 h in the extracts. (c) VSMC viability, (d) HUVEC viability, (e) ionconcentrations of DMEM extracts and (f) NO release of HUVECs cultured in extracts for 24 h.


muscle [46], inhibit the growth of VSMCs and reduce the chancesof thrombosis and restenosis [47]. The NO release results revealedsignificant increase of NO in all the Mg-Li-Zn alloy extracts except

that of Mg-3.5Li-0.5Zn. Apart from the Mg-Li-Zn alloys, pure Mgalso showed increased NO release, which is in good accordancewith the findings obtained from the previous research [48].


Another interesting finding is that the Li-abundant Mg-Li-Zn alloys(i.e., Mg-6.5Li-0.5Zn, Mg-6.5Li-2Zn and Mg-6.5Li-4Zn) releasednearly double amounts of NO, as compared with the negative con-trol, which implied the potential improving effects of Li in thedeveloped Mg-Li-Zn alloys on NO release.

3.7. Platelet adhesion and hemolysis

Hemocompatibility test results of the Mg-Li-Zn alloys areshown in Fig. 8. In theory, aggregation begins after activation[49]. Pure Mg showed obvious aggregation of human platelets withsome platelets having extended pseudopodia. The Mg-Li-Zn alloyscaused different degrees of aggregation of the platelets. By compar-ison, the Mg-3.5Li-2Zn alloy showed less aggregation, which hap-pened in only two or three platelets. Besides, most of theplatelets remained separated and round (Fig. 8c). As for activation,3 stages, named early dendritic, early spread and spread, will bepassed through accompanying morphology change [50]. It is worthnoting that the platelets adhered on the surface of the Mg-6.5Li-2Zn alloy showed an apparent transition morphology, with morepseudopodia spreading but not becoming octopus-like morphol-ogy in early dendritic stage (Fig. 8f). A few platelets aggregatedwhile most of the platelets extended pseudopodia for furtheraggregation, indicating initial state of induction. Furthermore,released NO can provide another function to reduce the adhesionand activation of platelets [51]. The present results suggest goodhemocompatibility of the currently investigated Mg-Li-Zn alloysin vitro, which calls for in vivo evaluation for verification. Hemol-ysis results are shown in Fig. 8h. Only the Mg-3.5Li-2Zn alloyhad a hemolysis value lower than 5%. This result could be easilyunderstood because the hemolysis of biodegradable Mg alloys isstrongly dependent on corrosion resistance and related ion releaseand pH change. Previous study has demonstrated higher influenceof pH value rather than ions on the elevation of hemolysis rate[52]. Thus the hemolysis rate would be reduced due to the relativestable pH value in vivo.

3.8. Coagulation four

Four common clinical indexes, i.e., activated partial thrombo-plastin time (APTT), prothrombin time (PT), Fibrinogen with Claussmethod (FIB-C) and thrombin time (TT) are shown in Fig. 9. APTTand PT are used to show the speed of coagulation caused by intrin-sic coagulation pathway and extrinsic coagulation pathway,respectively. Fibrinogen is a glycoprotein in vertebrates that helps

Fig. 8. Platelet adhesion on (a) pure Mg, (b) Mg-3.5Li-0.5Zn, (c) Mg-3.5Li-2Zn, (d) Mg-3.5rate. The scale bar in the zoomed image is 2 lm.

the formation of blood clots, whilst TT measures the time of clotformation in the plasma of a blood sample containing anticoagu-lant [53], or known as thrombin-mediated fibrin formation [54].All the four indexes stayed within a range for healthy people. PureMg, Mg-3.5Li-0.5Zn, Mg-3.5Li-2Zn and all the Mg-6.5Li-xZn alloysshowed significant increases of APTT relative to the control, indi-cating the effects of these materials on the intrinsic pathway. Fur-thermore, the increased amounts of alloying elements showed atendency of prolonged APTT results, suggesting the effects of Li+

and Zn2+ in the prolonged APTT. The results were in good agree-ment with those obtained in a previous study where Zn2+ influ-enced the intrinsic pathway by binding to factor XII [55]. Besides,all the APTT results were lower than 34 s, which would lead tono abnormal clinical effects [56]. As for PT, pure Mg had a signifi-cantly lower value than the negative control, implying that pureMg affected the extrinsic clotting pathway but all the Mg-Li-Znalloys appeared to exert no influence. Fibrinogen before and afterincubating the materials revealed no significant differences. Asfor TT, both Mg-3.5Li-2Zn and Mg-6.5Li-4Zn showed significantincreases, as compared with the negative control group. For coag-ulation, the intracellular Ca2+ has been demonstrated to be criti-cally involved in the functioning of coagulation cascade [57]. Inthis case, Mg2+, as an antagonist to Ca2+, would delay the coagula-tion time [58]. Previous studies showed the tendency that Mgalloys with higher corrosion rates induced more delayed coagula-tion time [54,59]. In this study, the coagulation time seemed tobe related to not only corrosion rate but also the ions released,since some of the alloys with higher corrosion rates showed lessinterference, as well as the tendency in the delayed time withthe addition of alloying elements.

4. Discussion

4.1. Mechanical properties

In the present study, Mg-Li-Zn alloys subject to extrusion pro-cess have been found significant improvement on the mechanicalproperties. The microstructure of the material helps besides theextrusion process which provided severe deformation to improvemechanical properties. The combined addition of Li and Zn formedZn-containing particles/precipitations (MgLiZn phase) in Mg-3.5Li-Zn alloys to improve the strength, whilst the Zn-containing parti-cles (metastable MgLi2Zn phase) in a + b phase Mg-6.5Li-Zn alloyshelp improve the mechanical properties [16]. As for the room-temperature mechanical properties, a summary of binary Mg-Li

Li-4Zn, (e) Mg-6.5Li-0.5Zn, (f) Mg-6.5Li-2Zn, and (g) Mg-6.5Li-4Zn and (h) hemolysis

Fig. 9. APTT, PT, FIB-C and TT results of samples incubated in platelet-poor plasma (PPP).


alloys and ternary Mg-Li-X (X = Al, Zn, Ca, Y, Ce) alloys is shown inFig. 10 [1,11,60-77]. It is obvious that the mechanical properties ofmagnesium are significantly improved by the addition of Li and Zn.Moreover. Zn shows the best strengthening effect, as comparedwith the other alloying elements, such as Ca, Y and Ce. Althoughaluminum has a good strengthening effect in Mg-Li-Al alloys, it ispotentially neurotoxic. Therefore, our results confirm that Zn isindeed an effective element to replace aluminum to achieve a goodbalance between strength and ductility. As far as the choice ofmaterial for potential stent applications is concerned, the Mg-3.5Li-2Zn alloy (�250 MPa UTS, >20% elongation) is superior tothe other alloys currently studied and previously studied, includingMg-Li or Mg-Li-X alloys subjected to ECAP (equal channel angularpressing) or other post-processing [1,65,71,73-78], and includingthe Mg-Li-Ca alloys previously developed for biomedical applica-tions [11]. Besides ultimate tensile strength, plastic deformabilityis crucial for materials designed for stent. In this study, the yieldstrength r0.2 has been remarkably improved after the addition ofLi and Zn. Previously studies have shown 90–126 MPa in as-rolled Mg-Li-X alloys [67,70,71,73,75], 108–146 MPa in as-extruded Mg-Li-X alloys [11,64,65] and 150–155 MPa in ECAPed(Equal Channel Angular Pressing) Mg-Li-X alloys [1,60]. Mg-3.5Li-2.0Zn, Mg-3.5Li-4.0Zn and Mg-6.5Li-4.0Zn in this study showedsuperior plastic deformability than any Mg-Li-X ternary alloys

either by rolling, extrusion or ECAP. However, higher plasticdeformability is still desired to provide better performance of thestent. Obviously, any medical devices potentially made of theMg-Li-Zn alloys with specific amount of Li and Zn will have distinctadvantages in specific strength and exceptionally large ductility asneeded for vascular stents.

4.2. Biodegradation behavior

On the surface of the alloys studied, Li2CO3 is present in the pro-tecting layer. Moreover, at a low Li content, the concentrations of Liand Mg on the surface seem to be adjustable by the content of Zn.The Mg-3.5Li-2Zn and Mg-3.5Li-4Zn alloys show higher concentra-tion of Mg and lower concentration of Li on the surface, whilst theMg-3.5Li-0.5Zn alloy has a higher Li concentration on the surface.Moreover, the corrosion mechanism changes with increasingamounts of Li and Zn, which is reported for the first time. WhenZn content increased from 0.5 to 4 wt%, the corrosion mechanismchanges from local corrosion with a few pitting sites, to local cor-rosion with small pitting, and to severe local corrosion (Fig. S2).Furthermore, with the addition of Li and Zn, naturally oxidizedMg-Li-Zn alloys have different corrosion preferences, which isshown in Fig. S3.

Fig. 10. Summary of room temperature mechanical properties of binary Mg-Li alloys and ternary Mg-Li-X (X = Al, Zn, Ca, Y, Ce, Sc, Mn, Ag) alloys [1,11,60–77].


The reported Mg-Li-based alloys for biomedical applicationinclude the Mg-Li-(Al)-(RE) alloys [12], Mg-Li-Ca alloys [11] andMg-Li-Zn alloys in the present study. The as-cast and as-extrudedMg-9.29Li-0.88Ca alloy was subjected to short immersion testsup to 430 min and no longer immersion test was conducted. Mean-while, the hydrogen evolution data of Mg-Li-(Al)-(RE) are available.According to previous study, hydrogen evolution rate can reflectthe corrosion rate of Mg alloys to some degree [79]. Comparisonin hydrogen evolution after 240 h between the Mg-Li-(Al)-(RE)alloys and the alloys used in the present study is shown inFig. 11a. It is obvious that the Mg-Li-Zn alloys developed in thisstudy have generally better corrosion resistance in Hank’s solution.Corrosion current density shows the transient behavior of metalswith time. It is therefore interesting to compare the corrosion current

Fig. 11. Comparison in (a) hydrogen evolution between the previously studied Mg-Li-solution at 37 �C, and (b) corrosion current density between the previously studied Mg-

density values of the Mg-Li-(Al)-(RE) alloys, Mg-9.29Li-0.88Caalloy and the alloys used in the present study. It is obvious thatthe corrosion current density values of the alloys used in thepresent study are among the lowest ones, although the determina-tion of the Tafel slope may cause large variations [19].

4.3. In vitro cytocompatibility and hemocompatility

The HUVEC viability assay in the present study suggests the via-bility of cells cultured in extracts is first suppressed but then returnto normal. The suppression at 1 d was likely to be the result of highion concentration, large quantity of hydrogen evolution or alka-lized environment [80]. However, the concentration level of eitherMg2+, Li+ or Zn2+ will not cause the HUVEC viability to suppress

(Al)-(RE) alloys and the present Mg-Li-Zn alloys after 240 h immersion in Hank’sLi-Ca and Mg-Li-(Al)-(RE) alloys and the present Mg-Li-Zn alloys.


[42,81]. The maximal concentration of Mg2+ that does not inhibitthe viability of HCAECs has been found to be 30 mM, which corre-sponds to nearly 700 mg�L�1 [82]. This value is much larger thanMg2+ concentration encountered in this study. Furthermore, as ithas been mentioned, Mg-Li-Zn alloys produced limited hydrogento the environment. Therefore, the slight suppressing on day onemay be considered to be caused by a relatively high pH value. Pre-vious studies have demonstrated pH value of the extract higherthan 8 [12,17]. Nevertheless, the slight suppression at 1 d in thepresent study has been shown recoverable at 3 d and 5 d. Cellularmetabolism with acidic products may help balanced pH value.Besides, the artery is a dynamic environment with pH value around7.4. Thus the recoverable suppression as well as the low content ofreleased metallic ions and hydrogen confirmed the good viabilityof Mg-Li-Zn alloys to HUVECs. As for the VSMC viability, a low levelof magnesium ion concentration (below 40 mM) has been found toshow no side-effects on human VSMCs [42] and again this value ismuch larger than the Mg2+ concentration encountered in thisstudy. The extracts obtained in this study except for Mg-3.5Li-0.5Zn all had Li+ concentrations around 1 mM. As previously illus-trated, 5 mM LiCl pretreatment for 30 min was found to suppressthe proliferation of rat VSMCs [43]. In the present study, the con-centration of Li+ is around 1 mM but the treating time is 1, 3 and5 days, which is much longer than 30 min. Long treatment of Li+

can be the main reason for the suppression of VSMC proliferation.In addition to Li+ concentration, the alkaline environment causedby the biodegradation of magnesium may affect the cell viability.According to previous studies, Mg-Li-Al-RE and some other Fe-based and Zn-based alloys as well showed limited toxicity toVSMCs [12,83,84]. Commercial drug eluting coating for metallicstent is applied to inhibit the excessive proliferation of VSMC andthus prevent restenosis, but often accompanied by impairedendothelial regeneration and increased the rate of stent thrombo-sis [85]. The inhibition of VSMCs by iron has been considered toreduce restenosis [86]. Thus the current study provide evidencefor Mg-Li-Zn to reduce restenosis in vivo. Moreover, a relationshipbetween increasing NO release and increasing Li+ release has beenfound. This needs further investigation. As for hemolysis,biodegradable Mg alloys seem easy to cause high hemolysis [87].Previous report has demonstrated the high pH value has higherimpact on the high hemolysis than the composition. As for stentmaterial in dynamic blood environment, the pH value is stableand thus the hemolysis results would be better. To summarize,the current study provides evidence of the feasibility of Mg-Li-Znalloys especially Mg-3.5Li-2Zn as vascular stent material.

5. Conclusions

In this study, Mg-Li-Zn alloys were investigated to explore thefeasibility to be used as potential biodegradable stent materials.The results showed enhanced mechanical properties (bothstrength and ductility, relative to pure magnesium), good corrosionresistance and good biocompatibility, indicating the potential ofthe Mg-Li-Zn alloys as good candidates for cardiovascular stentapplications. The main findings are as follows.

(1) A protecting surface layer containing Li2CO3/MgCO3 wasformed on the alloy surface, which was dependent on theLi/Mg ratio of the alloy. Zn seemed to be located along theb-phase and was not detected in the protecting layer.

(2) The Mg-Li-Zn alloys exhibited significantly improvedmechanical properties due to the existence of a phase andZn-containing particles/precipitations (h phase and/or h’phase) as well as extrusion process. The Mg-3.5Li-2Zn alloy

developed in the present study showed superior mechanicalproperties to most of reported binary Mg-Li and ternary Mg-Li-X (X = Al, Zn, Ca, Y, Ce, Sc, Mn and Ag) alloys.

(3) Both electrochemical tests and in vitro corrosion evaluationshowed comparable corrosion resistance of Mg-3.5Li-2Znand Mg-6.5Li-2Zn to that of pure Mg in Hank’s solutionand superior corrosion resistance in the cell culture medium.The corrosion products were found to be composed of Mg(OH)2, MgCO3 and Ca-free Mg/P inorganics and Ca/Pinorganics.

(4) In vitro cytotoxicity assay revealed different behaviors ofHUVECs and VSMCs. HUVECs showed increased NO releaseand tolerable toxicity, while VSMCs had limited decreasingviability with time.

(5) Hemocompatibility tests showed different degrees of activa-tion behavior of the Mg-Li-Zn alloys and the hemolysis ofMg-3.5Li-2Zn was lower than 5%. Further coagulation fourresults showed good reaction of PPP to the Mg-Li-Zn alloys.

Acknowledgments

This work was supported by the National Natural Science Foun-dation of China (Grant No. 51431002), National Key Research andDevelopment Program of China (Grant No. 2016YFC1102402),NSFC/RGC Joint Research Scheme (Grant No. 51361165101 and5161101031) and NSFC-RFBR Cooperation Project (Grant No.51611130054).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.actbio.2017.08.021.

References

[1] T.-C. Chang, J.-Y. Wang, C.-L. Chu, S. Lee, Mechanical properties andmicrostructures of various Mg–Li alloys, Mater. Lett. 60 (2006) 3272–3276.

[2] S.R. Agnew, M.H. Yoo, C.N. Tomé, Application of texture simulation tounderstanding mechanical behavior of Mg and solid solution alloyscontaining Li or Y, Acta Mater. 49 (2001) 4277–4289.

[3] Y. Song, D. Shan, R. Chen, E.-H. Han, Corrosion characterization of Mg–8Li alloyin NaCl solution, Corros. Sci. 51 (2009) 1087–1094.

[4] D. Cao, L. Wu, Y. Sun, G. Wang, Y. Lv, Electrochemical behavior of Mg–Li, Mg–Li–Al and Mg–Li–Al–Ce in sodium chloride solution, J. Power Sour. 177 (2008)624–630.

[5] W. Xu, N. Birbilis, G. Sha, Y. Wang, J.E. Daniels, Y. Xiao, M. Ferry, A high-specific-strength and corrosion-resistant magnesium alloy, Nat. Mater. 14(2015) 1229–1235.

[6] J.R. Geddes, S. Burgess, K. Hawton, K. Jamison, G.M. Goodwin, Long-termlithium therapy for bipolar disorder: systematic review and meta-analysis ofrandomized controlled trials, Am. J. Psychiatry. 161 (2004) 217–222.

[7] S. Kean, The Disappearing Spoon and Other True Tales of Madness, Love, andthe History of the World from the Periodic Table of the Elements, RandomHouse, New York, 2011.

[8] J. Lieb, A. Zeff, Lithium treatment of chronic cluster headaches, Br. J. Psychiatry133 (1978) 556–558.

[9] S. Guo, K. Arai, M.F. Stins, D.M. Chuang, E.H. Lo, Lithium upregulates vascularendothelial growth factor in brain endothelial cells and astrocytes, Stroke 40(2009) 652–655.

[10] M. Ghasemi, A.R. Dehpour, The NMDA receptor/nitric oxide pathway: a targetfor the therapeutic and toxic effects of lithium, Trends Pharmacol. Sci. 32(2011) 420–434.

[11] R.C. Zeng, L. Sun, Y.F. Zheng, H.Z. Cui, E.H. Han, Corrosion and characterisationof dual phase Mg–Li–Ca alloy in Hank’s solution: the influence ofmicrostructural features, Corros. Sci. 79 (2014) 69–82.

[12] W.R. Zhou, Y.F. Zheng, M.A. Leeflang, J. Zhou, Mechanical property,biocorrosion and in vitro biocompatibility evaluations of Mg-Li-(Al)-(RE)alloys for future cardiovascular stent application, Acta Biomater. 9 (2013)8488–8498.

http://dx.doi.org/10.1016/j.actbio.2017.08.021http://dx.doi.org/10.1016/j.actbio.2017.08.021http://refhub.elsevier.com/S1742-7061(17)30524-X/h0005http://refhub.elsevier.com/S1742-7061(17)30524-X/h0005http://refhub.elsevier.com/S1742-7061(17)30524-X/h0010http://refhub.elsevier.com/S1742-7061(17)30524-X/h0010http://refhub.elsevier.com/S1742-7061(17)30524-X/h0010http://refhub.elsevier.com/S1742-7061(17)30524-X/h0015http://refhub.elsevier.com/S1742-7061(17)30524-X/h0015http://refhub.elsevier.com/S1742-7061(17)30524-X/h0020http://refhub.elsevier.com/S1742-7061(17)30524-X/h0020http://refhub.elsevier.com/S1742-7061(17)30524-X/h0020http://refhub.elsevier.com/S1742-7061(17)30524-X/h0025http://refhub.elsevier.com/S1742-7061(17)30524-X/h0025http://refhub.elsevier.com/S1742-7061(17)30524-X/h0025http://refhub.elsevier.com/S1742-7061(17)30524-X/h0030http://refhub.elsevier.com/S1742-7061(17)30524-X/h0030http://refhub.elsevier.com/S1742-7061(17)30524-X/h0030http://refhub.elsevier.com/S1742-7061(17)30524-X/h0035http://refhub.elsevier.com/S1742-7061(17)30524-X/h0035http://refhub.elsevier.com/S1742-7061(17)30524-X/h0035http://refhub.elsevier.com/S1742-7061(17)30524-X/h0035http://refhub.elsevier.com/S1742-7061(17)30524-X/h0040http://refhub.elsevier.com/S1742-7061(17)30524-X/h0040http://refhub.elsevier.com/S1742-7061(17)30524-X/h0045http://refhub.elsevier.com/S1742-7061(17)30524-X/h0045http://refhub.elsevier.com/S1742-7061(17)30524-X/h0045http://refhub.elsevier.com/S1742-7061(17)30524-X/h0050http://refhub.elsevier.com/S1742-7061(17)30524-X/h0050http://refhub.elsevier.com/S1742-7061(17)30524-X/h0050http://refhub.elsevier.com/S1742-7061(17)30524-X/h0055http://refhub.elsevier.com/S1742-7061(17)30524-X/h0055http://refhub.elsevier.com/S1742-7061(17)30524-X/h0055http://refhub.elsevier.com/S1742-7061(17)30524-X/h0060http://refhub.elsevier.com/S1742-7061(17)30524-X/h0060http://refhub.elsevier.com/S1742-7061(17)30524-X/h0060http://refhub.elsevier.com/S1742-7061(17)30524-X/h0060


[13] K.M. Hambidge, N.F. Krebs, Zinc deficiency: a special challenge, J. Nutr. 137(2007) 1101–1105.

[14] P.K. Bowen, J. Drelich, J. Goldman, Zinc exhibits ideal physiological corrosionbehavior for bioabsorbable stents, Adv. Mater. 25 (2013) 2577–2582.

[15] S. Zhao, C.T. McNamara, P.K. Bowen, N. Verhun, J.P. Braykovich, J. Goldman, J.W. Drelich, Structural characteristics and in vitro biodegradation of a novel Zn-Li alloy prepared by induction melting and hot rolling, Metall. Mater. Trans. A.48 (2017) 1204–1215.

[16] A. Yamamoto, T. Ashida, Y. Kouta, K. Kim, et al., Precipitation in Mg-(4–13)%Li-(4–5)%Zn ternary alloys, Mater. Trans. 44 (2003) 619–624.

[17] X. Gu, Y. Zheng, Y. Cheng, S. Zhong, T. Xi, In vitro corrosion andbiocompatibility of binary magnesium alloys, Biomaterials 30 (2009) 484–498.

[18] S. Zhang, X. Zhang, C. Zhao, J. Li, Y. Song, C. Xie, H. Tao, Y. Zhang, Y. He, Y. Jiang,Y. Bian, Research on an Mg–Zn alloy as a degradable biomaterial, ActaBiomater. 6 (2010) 626–640.

[19] N.T. Kirkland, N. Birbilis, M.P. Staiger, Assessing the corrosion of biodegradablemagnesium implants: a critical review of current methodologies and theirlimitations, Acta Biomater. 8 (2012) 925–936.

[20] R.C. Zeng, L.Y. Cui, K. Jiang, R. Liu, B.D. Zhao, Y.F. Zheng, In vitro corrosion andcytocompatibility of a microarc oxidation coating and poly(L-lactic acid)composite coating on Mg-1Li-1Ca alloy for orthopedic implants, ACS Appl.Mater. Interfaces 8 (2016) 10014–10028.

[21] J. Fischer, M.H. Prosenc, M. Wolff, N. Hort, R. Willumeit, F. Feyerabend,Interference of magnesium corrosion with tetrazolium-based cytotoxicityassays, Acta Biomater. 6 (2010) 1813–1823.

[22] J.G. Sevillano, P. van Houtte, E. Aernoudt, Large strain work hardening andtextures, Prog. Mater. Sci. 25 (1980) 69–134.

[23] J.G. Sevillano, C. García-Rosales, J.F. Fuster, Texture and large–straindeformation microstructure, Philos. Trans. R. Soc. A 357 (1999) 1603–1619.

[24] M.R. Barnett, A. Ghaderi, I. Sabirov, B. Hutchinson, Role of grain boundarysliding in the anisotropy of magnesium alloys, Scr. Mater. 61 (2009) 277–280.

[25] T. Eriksson, A.M. Andersson, A.G. Bishop, C. Gejke, T.R. Gustafsson, J.O. Thomas,Surface analysis of LiMn2O4 electrodes in carbonate-based electrolytes, J.Electrochem. Soc. 149 (2002) A69–A78.

[26] K.B. Garg, T. Chatterji, S. Dalela, M. Heinonnen, J. Leiro, B. Dalela, R.K. Singhal,Core level photoemission study of polycrystalline MgB2, Solid State Commun.131 (2004) 343–347.

[27] Y. Bouvier, B. Mutel, J. Grimblot, Use of an Auger parameter for characterizingthe Mg chemical state in different materials, Surf. Coat. Technol. 180 (2004)169–173.

[28] S. Tanaka, M. Taniguchi, H. Tanigawa, XPS and UPS studies on electronicstructure of Li2O, J. Nucl. Mater. 283 (2000) 1405–1408.

[29] D.Y. Kwon, J.I. Kim, D.Y. Kim, H.J. Kang, B. Lee, K.W. Lee, M.S. Kim,Biodegradable stent, J. Biomed. Sci. Eng. 5 (2012) 208–216.

[30] Petr Vanýsek, Electrochemical series, in: William M. Haynes, CRC Handbook ofChemistry and Physics, CRC Press Inc., Boca Raton, 2010, pp. 8–20 – 8–28.

[31] P.K. Bowen, J. Drelich, J. Goldman, Magnesium in the murine artery: probingthe products of corrosion, Acta Biomater. 10 (2014) 1475–1483.

[32] M. Lebon, I. Reiche, F. Fröhlich, J.-J. Bahain, C. Falguères, Characterization ofarchaeological burnt bones: contribution of a new analytical protocol based onderivative FTIR spectroscopy and curve fitting of the m1m3 PO4 domain, Anal.Bioanal. Chem. 392 (2008) 1479–1488.

[33] A. Brangule, K. Agris Gross, Importance of FTIR spectra deconvolution for theanalysis of amorphous calcium phosphates, IOP Conf. Ser. Mater. Sci. Eng. 1(2015) 012027.

[34] E.C. Victoria, F. Gnanam, Synthesis and characterization of biphasic calciumphosphate, Trends Biomater. Artif. Organs. 16 (2002) 12–14.

[35] B. Boonchom, Kinetic and thermodynamic studies of MgHPO4�3H2O by non-isothermal decomposition data, J. Therm. Anal. Calorim. 98 (2009) 863–871.

[36] Y. Leung, M.A. Walters, R.Z. Legeros, Second derivative infrared spectra ofhydroxyapatite, Spectrochim. Acta Mol. Spectrosc. 46 (1990) 1453–1459.

[37] E.P. Paschalis, E. Dicarlo, F. Betts, P. Sherman, R. Mendelsohn, A.L. Boskey, FTIRmicrospectroscopic analysis of human osteonal bone, Calcif. Tissue Int. 59(1996) 480–487.

[38] W.H. Chang, B. Qu, A.D. Liao, S.F. Zhang, R.F. Zhang, J.H. Xiang, In vitrobiocompatibility and antibacterial behavior of anodic coatings fabricated in anorganic phosphate containing solution on Mg–1.0Ca alloys, Surf. Coat. Technol.289 (2016) 75–84.

[39] X. Zhu, J. Chen, L. Scheideler, R. Reichl, J. Geis-Gerstorfer, Effects of topographyand composition of titanium surface oxides on osteoblast responses,Biomaterials 25 (2004) 4087–4103.

[40] F. Barrère, A. Lebugle, Calcium phosphate interactions with titanium oxide andalumina substrates: an XPS study, J. Mater. Sci. Mater. Med. 14 (2003) 419–425.

[41] Y. Liu, W.-M. Wang, X.-L. Zhang, H.-Q. He, X.-L. Sun, H. Zeng, et al., AGE/RAGEpromotes thecalcification of human aortic smooth muscle cells via the Wnt/b-catenin axis, Am. J. Transl. Res. 8 (2016) 4644–4656.

[42] J. Ma, N. Zhao, D. Zhu, Biphasic responses of human vascular smooth musclecells to magnesium ion, J. Biomed. Mater. Res. A. 104 (2016) 347–356.

[43] Z. Wang, X. Zhang, S. Chen, D. Wang, J. Wu, T. Liang, et al., Lithium chlorideinhibits vascular smooth muscle cell proliferation and migration and alleviatesinjury-induced neointimal hyperplasia via induction of PGC-1a, PLoS One 8(2013) e55471.

[44] K. Bala, K. Ambwani, N.K. Gohil, Effect of different mitogens and serumconcentration on HUVEC morphology and characteristics: implication on useof higher passage cells, Tissue Cell 43 (2011) 216–222.

[45] H. Inoguchi, T. Tanaka, Y. Maehara, T. Matsuda, The effect of gradually gradedshear stress on the morphological integrity of a huvec-seeded compliantsmall-diameter vascular graft, Biomaterials 28 (2007) 486–495.

[46] R.M. Palmer, D.S. Ashton, S. Moncada, Vascular endothelial cells synthesizenitric oxide from L-arginine, Nature 333 (1988) 664–666.

[47] K.S. Bohl, J.L. West, Nitric oxide-generating polymers reduce platelet adhesionand smooth muscle cell proliferation, Biomaterials 21 (2000) 2273–2278.

[48] G. Wang, S. Ge, Y. Shen, H. Wang, Q. Dong, Q. Zhang, J. Gao, Y. Wang, Study onthe biodegradability and biocompatibility of WE magnesium alloys, Mater. Sci.Eng. C 32 (2012) 2190–2198.

[49] J. Yip, S. Yang, M.C. Berndt, R.K. Andrews, Primary platelet adhesion receptors,IUBMB Life 57 (2005) 103–108.

[50] P. Matarrese, E. Straface, G. Palumbo, M. Anselmi, L. Gambardella, B. Ascione,et al., Mitochondria regulate platelet metamorphosis induced by opsonizedzymosan A – activation and long-term commitment to cell death, FEBS J. 276(2009) 845–856.

[51] J. Loscalzo, Nitric oxide and restenosis, Clin. Appl. Thromb. Hemost. 2 (1996)7–10.

[52] Z. Zhen, X. Liu, T. Huang, T. Xi, Y. Zheng, Hemolysis and cytotoxicitymechanisms of biodegradable magnesium and its alloys, Mater. Sci. Eng. C46 (2015) 202–206.

[53] M. Lordkipanizé, G.C. Lowe, P. Harrison, Tests of platelet function, in: N. Key,M. Makris, D. O’Shaughnessy, D. Lillicrap (Eds.), Practical hemostasis andthrombosis, John Wiley & Sons Ltd, West Sussex, 2016, pp. 42–62.

[54] J. Wang, Y. He, M.F. Maitz, B. Collins, K. Xiong, L. Guo, Y. Yun, G. Wan, N. Huang,A surface-eroding poly(1,3-trimethylene carbonate) coating for fullybiodegradable magnesium-based stent applications: toward betterbiofunction, biodegradation and biocompatibility, Acta Biomater. 9 (2013)8678–8689.

[55] I. Schousboe, Contact activation in human plasma is triggered by zinc ionmodulation of factor XII (Hageman factor), Blood Coagul. Fibrinolysis 4 (1993)671–678.

[56] S. Scarpelini, S. Rhind, B. Nascimento, H. Tien, P. Shek, H. Peng, H. Huang, R.Pinto, V. Speers, M. Reis, S. Rizoli, Normal range values forthromboelastography in healthy adult volunteers, Braz. J. Med. Biol. Res. 42(2009) 1210–1217.

[57] J.L. Daniel, C. Dangelmaier, J. Jin, B. Ashby, J.B. Smith, S.P. Kunapuli, Molecularbasis for ADP-induced platelet activation I. Evidence for three distinct ADPreceptors on human platelets, J. Biol. Chem. 273 (1998) 2024–2029.

[58] J. Adams, J. Mitchell, The effect of agents which modify platelet behaviour andof magnesium ions on thrombus formation in vivo, Thromb. Haemost. 42(1979) 603–610.

[59] P. Lu, L. Cao, Y. Liu, X. Xu, X. Wu, Evaluation of magnesium ions release,biocorrosion, and hemocompatibility of MAO/PLLA-modified magnesium alloyWE42, J. Biomed. Mater. Res. B 96 (2011) 101–109.

[60] J.Y. Wang, T.C. Chang, L.Z. Chang, S. Lee, Effect of Al and Mn content on themechanical properties of various ECAE processed Mg-Li-Zn alloys, Mater.Trans. 47 (2006) 971–976.

[61] H. Takuda, S. Kikuchi, T. Tsukada, K. Kubota, N. Hatta, Effect of strain rate ondeformation behaviour of a Mg–8.5Li–1Zn alloy sheet at room temperature,Mater. Sci. Eng. A 271 (1999) 251–256.

[62] H. Dong, L. Wang, Y. Wu, L. Wang, Effect of Y on microstructure andmechanical properties of duplex Mg–7Li alloys, J. Alloys Compd. 506 (2010)468–474.

[63] R. Wu, Z. Qu, M. Zhang, Effects of the addition of Y in Mg–8Li–(1,3)Al alloy,Mater. Sci. Eng. A 516 (2009) 96–99.

[64] H. Dong, F. Pan, B. Jiang, Y. Zeng, Evolution of microstructure and mechanicalproperties of a duplex Mg–Li alloy under extrusion with an increasing ratio,Mater. Des. 57 (2014) 121–127.

[65] Y.N. Lin, H.Y. Wu, G.Z. Zhou, C.H. Chiu, S. Lee, Mechanical and anisotropicbehaviors of Mg–Li–Zn alloy thin sheets, Mater. Des. 29 (2008) 2061–2065.

[66] A. Sanschagrin, R. Tremblay, R. Angers, D. Dubé, Mechanical properties andmicrostructure of new magnesium—lithium base alloys, Mater. Sci. Eng. A 220(1996) 69–77.

[67] D.K. Xu, T.T. Zu, M. Yin, Y.B. Xu, E.H. Han, Mechanical properties of theicosahedral phase reinforced duplex Mg–Li alloy both at room and elevatedtemperatures, J. Alloys Compd. 582 (2014) 161–166.

[68] M.L. Zhang, R.Z. Wu, T. Wang, Microstructure and mechanical properties ofMg–8Li–(0–3)Ce alloys, J. Mater. Sci. 44 (2009) 1237–1240.

[69] P. Crawford, R. Barrosa, J. Mendez, J. Foyos, O.S. Es-Said, On the transformationcharacteristics of LA141A (Mg-Li-Al) alloy, J. Mater. Process. Technol. 56(1996) 108–118.

[70] K. Kubota, M. Mabuchi, K. Higashi, Review processing and mechanicalproperties of fine-grained magnesium alloys, J. Mater. Sci. 34 (1999) 2255–2262.

[71] H. Takuda, H. Matsusaka, S. Kikuchi, K. Kubota, Tensile properties of a few Mg-Li-Zn alloy thin sheets, J. Mater. Sci. 37 (2002) 51–57.

[72] T. Liu, Y.D. Wang, S.D. Wu, R.L. Peng, C.X. Huang, C.B. Jiang, S.X. Li, Textures andmechanical behavior of Mg–3.3%Li alloy after ECAP, Scr. Mater. 51 (2004)1057–1061.

[73] H. Takuda, T. Enami, K. Kubota, N. Hatta, The formability of a thin sheet of Mg–8.5Li–1Zn alloy, J. Mater. Process. Technol. 101 (2000) 281–286.

[74] M. Karami, R. Mahmudi, Work hardening behavior of the extruded and equal-channel angularly pressed Mg–Li–Zn alloys under tensile and sheardeformation modes, Mater. Sci. Eng. A 607 (2014) 512–520.

http://refhub.elsevier.com/S1742-7061(17)30524-X/h0065http://refhub.elsevier.com/S1742-7061(17)30524-X/h0065http://refhub.elsevier.com/S1742-7061(17)30524-X/h0070http://refhub.elsevier.com/S1742-7061(17)30524-X/h0070http://refhub.elsevier.com/S1742-7061(17)30524-X/h0075http://refhub.elsevier.com/S1742-7061(17)30524-X/h0075http://refhub.elsevier.com/S1742-7061(17)30524-X/h0075http://refhub.elsevier.com/S1742-7061(17)30524-X/h0075http://refhub.elsevier.com/S1742-7061(17)30524-X/h0080http://refhub.elsevier.com/S1742-7061(17)30524-X/h0080http://refhub.elsevier.com/S1742-7061(17)30524-X/h0085http://refhub.elsevier.com/S1742-7061(17)30524-X/h0085http://refhub.elsevier.com/S1742-7061(17)30524-X/h0090http://refhub.elsevier.com/S1742-7061(17)30524-X/h0090http://refhub.elsevier.com/S1742-7061(17)30524-X/h0090http://refhub.elsevier.com/S1742-7061(17)30524-X/h0095http://refhub.elsevier.com/S1742-7061(17)30524-X/h0095http://refhub.elsevier.com/S1742-7061(17)30524-X/h0095http://refhub.elsevier.com/S1742-7061(17)30524-X/h0100http://refhub.elsevier.com/S1742-7061(17)30524-X/h0100http://refhub.elsevier.com/S1742-7061(17)30524-X/h0100http://refhub.elsevier.com/S1742-7061(17)30524-X/h0100http://refhub.elsevier.com/S1742-7061(17)30524-X/h0105http://refhub.elsevier.com/S1742-7061(17)30524-X/h0105http://refhub.elsevier.com/S1742-7061(17)30524-X/h0105http://refhub.elsevier.com/S1742-7061(17)30524-X/h0110http://refhub.elsevier.com/S1742-7061(17)30524-X/h0110http://refhub.elsevier.com/S1742-7061(17)30524-X/h0115http://refhub.elsevier.com/S1742-7061(17)30524-X/h0115http://refhub.elsevier.com/S1742-7061(17)30524-X/h0120http://refhub.elsevier.com/S1742-7061(17)30524-X/h0120http://refhub.elsevier.com/S1742-7061(17)30524-X/h0125http://refhub.elsevier.com/S1742-7061(17)30524-X/h0125http://refhub.elsevier.com/S1742-7061(17)30524-X/h0125http://refhub.elsevier.com/S1742-7061(17)30524-X/h0125http://refhub.elsevier.com/S1742-7061(17)30524-X/h0125http://refhub.elsevier.com/S1742-7061(17)30524-X/h0130http://refhub.elsevier.com/S1742-7061(17)30524-X/h0130http://refhub.elsevier.com/S1742-7061(17)30524-X/h0130http://refhub.elsevier.com/S1742-7061(17)30524-X/h0130http://refhub.elsevier.com/S1742-7061(17)30524-X/h0135http://refhub.elsevier.com/S1742-7061(17)30524-X/h0135http://refhub.elsevier.com/S1742-7061(17)30524-X/h0135http://refhub.elsevier.com/S1742-7061(17)30524-X/h0140http://refhub.elsevier.com/S1742-7061(17)30524-X/h0140http://refhub.elsevier.com/S1742-7061(17)30524-X/h0140http://refhub.elsevier.com/S1742-7061(17)30524-X/h0145http://refhub.elsevier.com/S1742-7061(17)30524-X/h0145http://refhub.elsevier.com/S1742-7061(17)30524-X/h0155http://refhub.elsevier.com/S1742-7061(17)30524-X/h0155http://refhub.elsevier.com/S1742-7061(17)30524-X/h0160http://refhub.elsevier.com/S1742-7061(17)30524-X/h0160http://refhub.elsevier.com/S1742-7061(17)30524-X/h0160http://refhub.elsevier.com/S1742-7061(17)30524-X/h0160http://refhub.elsevier.com/S1742-7061(17)30524-X/h0160http://refhub.elsevier.com/S1742-7061(17)30524-X/h0160http://refhub.elsevier.com/S1742-7061(17)30524-X/h0160http://refhub.elsevier.com/S1742-7061(17)30524-X/h0165http://refhub.elsevier.com/S1742-7061(17)30524-X/h0165http://refhub.elsevier.com/S1742-7061(17)30524-X/h0165http://refhub.elsevier.com/S1742-7061(17)30524-X/h0170http://refhub.elsevier.com/S1742-7061(17)30524-X/h0170http://refhub.elsevier.com/S1742-7061(17)30524-X/h0175http://refhub.elsevier.com/S1742-7061(17)30524-X/h0175http://refhub.elsevier.com/S1742-7061(17)30524-X/h0175http://refhub.elsevier.com/S1742-7061(17)30524-X/h0175http://refhub.elsevier.com/S1742-7061(17)30524-X/h0180http://refhub.elsevier.com/S1742-7061(17)30524-X/h0180http://refhub.elsevier.com/S1742-7061(17)30524-X/h0185http://refhub.elsevier.com/S1742-7061(17)30524-X/h0185http://refhub.elsevier.com/S1742-7061(17)30524-X/h0185http://refhub.elsevier.com/S1742-7061(17)30524-X/h0190http://refhub.elsevier.com/S1742-7061(17)30524-X/h0190http://refhub.elsevier.com/S1742-7061(17)30524-X/h0190http://refhub.elsevier.com/S1742-7061(17)30524-X/h0190http://refhub.elsevier.com/S1742-7061(17)30524-X/h0195http://refhub.elsevier.com/S1742-7061(17)30524-X/h0195http://refhub.elsevier.com/S1742-7061(17)30524-X/h0195http://refhub.elsevier.com/S1742-7061(17)30524-X/h0200http://refhub.elsevier.com/S1742-7061(17)30524-X/h0200http://refhub.elsevier.com/S1742-7061(17)30524-X/h0200http://refhub.elsevier.com/S1742-7061(17)30524-X/h0205http://refhub.elsevier.com/S1742-7061(17)30524-X/h0205http://refhub.elsevier.com/S1742-7061(17)30524-X/h0205http://refhub.elsevier.com/S1742-7061(17)30524-X/h0210http://refhub.elsevier.com/S1742-7061(17)30524-X/h0210http://refhub.elsevier.com/S1742-7061(17)30524-X/h0215http://refhub.elsevier.com/S1742-7061(17)30524-X/h0215http://refhub.elsevier.com/S1742-7061(17)30524-X/h0215http://refhub.elsevier.com/S1742-7061(17)30524-X/h0215http://refhub.elsevier.com/S1742-7061(17)30524-X/h0220http://refhub.elsevier.com/S1742-7061(17)30524-X/h0220http://refhub.elsevier.com/S1742-7061(17)30524-X/h0220http://refhub.elsevier.com/S1742-7061(17)30524-X/h0225http://refhub.elsevier.com/S1742-7061(17)30524-X/h0225http://refhub.elsevier.com/S1742-7061(17)30524-X/h0225http://refhub.elsevier.com/S1742-7061(17)30524-X/h0230http://refhub.elsevier.com/S1742-7061(17)30524-X/h0230http://refhub.elsevier.com/S1742-7061(17)30524-X/h0235http://refhub.elsevier.com/S1742-7061(17)30524-X/h0235http://refhub.elsevier.com/S1742-7061(17)30524-X/h0240http://refhub.elsevier.com/S1742-7061(17)30524-X/h0240http://refhub.elsevier.com/S1742-7061(17)30524-X/h0240http://refhub.elsevier.com/S1742-7061(17)30524-X/h0245http://refhub.elsevier.com/S1742-7061(17)30524-X/h0245http://refhub.elsevier.com/S1742-7061(17)30524-X/h0250http://refhub.elsevier.com/S1742-7061(17)30524-X/h0250http://refhub.elsevier.com/S1742-7061(17)30524-X/h0250http://refhub.elsevier.com/S1742-7061(17)30524-X/h0250http://refhub.elsevier.com/S1742-7061(17)30524-X/h0255http://refhub.elsevier.com/S1742-7061(17)30524-X/h0255http://refhub.elsevier.com/S1742-7061(17)30524-X/h0260http://refhub.elsevier.com/S1742-7061(17)30524-X/h0260http://refhub.elsevier.com/S1742-7061(17)30524-X/h0260http://refhub.elsevier.com/S1742-7061(17)30524-X/h0265http://refhub.elsevier.com/S1742-7061(17)30524-X/h0265http://refhub.elsevier.com/S1742-7061(17)30524-X/h0265http://refhub.elsevier.com/S1742-7061(17)30524-X/h0265http://refhub.elsevier.com/S1742-7061(17)30524-X/h0265http://refhub.elsevier.com/S1742-7061(17)30524-X/h0265http://refhub.elsevier.com/S1742-7061(17)30524-X/h0265http://refhub.elsevier.com/S1742-7061(17)30524-X/h0270http://refhub.elsevier.com/S1742-7061(17)30524-X/h0270http://refhub.elsevier.com/S1742-7061(17)30524-X/h0270http://refhub.elsevier.com/S1742-7061(17)30524-X/h0270http://refhub.elsevier.com/S1742-7061(17)30524-X/h0270http://refhub.elsevier.com/S1742-7061(17)30524-X/h0275http://refhub.elsevier.com/S1742-7061(17)30524-X/h0275http://refhub.elsevier.com/S1742-7061(17)30524-X/h0275http://refhub.elsevier.com/S1742-7061(17)30524-X/h0280http://refhub.elsevier.com/S1742-7061(17)30524-X/h0280http://refhub.elsevier.com/S1742-7061(17)30524-X/h0280http://refhub.elsevier.com/S1742-7061(17)30524-X/h0280http://refhub.elsevier.com/S1742-7061(17)30524-X/h0285http://refhub.elsevier.com/S1742-7061(17)30524-X/h0285http://refhub.elsevier.com/S1742-7061(17)30524-X/h0285http://refhub.elsevier.com/S1742-7061(17)30524-X/h0290http://refhub.elsevier.com/S1742-7061(17)30524-X/h0290http://refhub.elsevier.com/S1742-7061(17)30524-X/h0290http://refhub.elsevier.com/S1742-7061(17)30524-X/h0295http://refhub.elsevier.com/S1742-7061(17)30524-X/h0295http://refhub.elsevier.com/S1742-7061(17)30524-X/h0295http://refhub.elsevier.com/S1742-7061(17)30524-X/h0300http://refhub.elsevier.com/S1742-7061(17)30524-X/h0300http://refhub.elsevier.com/S1742-7061(17)30524-X/h0300http://refhub.elsevier.com/S1742-7061(17)30524-X/h0305http://refhub.elsevier.com/S1742-7061(17)30524-X/h0305http://refhub.elsevier.com/S1742-7061(17)30524-X/h0305http://refhub.elsevier.com/S1742-7061(17)30524-X/h0310http://refhub.elsevier.com/S1742-7061(17)30524-X/h0310http://refhub.elsevier.com/S1742-7061(17)30524-X/h0310http://refhub.elsevier.com/S1742-7061(17)30524-X/h0315http://refhub.elsevier.com/S1742-7061(17)30524-X/h0315http://refhub.elsevier.com/S1742-7061(17)30524-X/h0320http://refhub.elsevier.com/S1742-7061(17)30524-X/h0320http://refhub.elsevier.com/S1742-7061(17)30524-X/h0320http://refhub.elsevier.com/S1742-7061(17)30524-X/h0325http://refhub.elsevier.com/S1742-7061(17)30524-X/h0325http://refhub.elsevier.com/S1742-7061(17)30524-X/h0330http://refhub.elsevier.com/S1742-7061(17)30524-X/h0330http://refhub.elsevier.com/S1742-7061(17)30524-X/h0330http://refhub.elsevier.com/S1742-7061(17)30524-X/h0335http://refhub.elsevier.com/S1742-7061(17)30524-X/h0335http://refhub.elsevier.com/S1742-7061(17)30524-X/h0335http://refhub.elsevier.com/S1742-7061(17)30524-X/h0340http://refhub.elsevier.com/S1742-7061(17)30524-X/h0340http://refhub.elsevier.com/S1742-7061(17)30524-X/h0345http://refhub.elsevier.com/S1742-7061(17)30524-X/h0345http://refhub.elsevier.com/S1742-7061(17)30524-X/h0345http://refhub.elsevier.com/S1742-7061(17)30524-X/h0350http://refhub.elsevier.com/S1742-7061(17)30524-X/h0350http://refhub.elsevier.com/S1742-7061(17)30524-X/h0350http://refhub.elsevier.com/S1742-7061(17)30524-X/h0355http://refhub.elsevier.com/S1742-7061(17)30524-X/h0355http://refhub.elsevier.com/S1742-7061(17)30524-X/h0360http://refhub.elsevier.com/S1742-7061(17)30524-X/h0360http://refhub.elsevier.com/S1742-7061(17)30524-X/h0360http://refhub.elsevier.com/S1742-7061(17)30524-X/h0365http://refhub.elsevier.com/S1742-7061(17)30524-X/h0365http://refhub.elsevier.com/S1742-7061(17)30524-X/h0370http://refhub.elsevier.com/S1742-7061(17)30524-X/h0370http://refhub.elsevier.com/S1742-7061(17)30524-X/h0370


[75] G. Sha, X. Sun, T. Liu, Y. Zhu, T. Yu, Effects of Sc addition and annealingtreatment on the microstructure and mechanical properties of the as-rolledMg-3Li alloy, J. Mater. Sci. Technol. 27 (2011) 753–758.

[76] H. Li, H. Ji, G.C. Yao, Y.H. Liu, Z.Q. Guo, Z.G. Liu, Microstructure and mechanicalproperties of Mg-Li-Mn alloy, Chin. J. Process. Eng. 6 (2006) 491–494.

[77] Zhang, M.F.M. Elkin, Microstructure and properties of Mg-Li alloy (in Chinese),in: Zhang, M.F.M. Elkin (Eds.), Mg-Li ultra-light alloy, Science Press, Beijing,2010, pp. 171–252.

[78] W.R.D. Jones, G.V. Hogg, The stability of mechanical properties of beta-phasemagnesium-lithium alloys, J. Inst. Metals 85 (1957) 255–261.

[79] Z. Shi, A. Atrens, An innovative specimen configuration for the study of Mgcorrosion, Corros. Sci. 53 (2011) 226–246.

[80] Y.F. Zheng, X.N. Gu, F. Witte, Biodegradable metals, Mater. Sci. Eng. R. 77(2014) 1–34.

[81] F. Feyerabend, J. Fischer, J. Holtz, F. Witte, R. Willumeit, H. Drucker, C. Vogt, N.Hort, Evaluation of short-term effects of rare earth and other elements used inmagnesium alloys on primary cells and cell lines, Acta Biomater. 6 (2010)1834–1842.

[82] N. Zhao, D. Zhu, Endothelial responses of magnesium and other alloyingelements in magnesium-based stent materials, Metallomics 7 (2015)118–128.

[83] A. Francis, Y. Yang, S. Virtanen, A.R. Boccaccini, Iron and iron-based alloys fortemporary cardiovascular applications, J. Mater. Sci. Mater. Med. 26 (2015)138.

[84] H.F. Li, X.H. Xie, Y.F. Zheng, Y. Cong, F.Y. Zhou, K.J. Qiu, et al., Development ofbiodegradable Zn-1X binary alloys with nutrient alloying elements Mg, Ca andSr, Sci. Rep. 5 (2015) 10719.

[85] P.K. Bowen, E.R. Shearier, S. Zhao, R.J. Guillory, F. Zhao, J. Goldman, et al.,Biodegradable metals for cardiovascular stents: from clinical concerns torecent Zn-Alloys, Adv. Healthc. Mater. 5 (2016) 1121–1140.

[86] P.P. Mueller, T. May, A. Perz, H. Hauser, M. Peuster, Control of smooth musclecell proliferation by ferrous iron, Biomaterials 27 (2006) 2193–2200.

[87] N. Li, Y. Zheng, Novel magnesium alloys developed for biomedical application:a review, J. Mater. Sci. Technol. 29 (2013) 489–502.

http://refhub.elsevier.com/S1742-7061(17)30524-X/h0375http://refhub.elsevier.com/S1742-7061(17)30524-X/h0375http://refhub.elsevier.com/S1742-7061(17)30524-X/h0375http://refhub.elsevier.com/S1742-7061(17)30524-X/h0380http://refhub.elsevier.com/S1742-7061(17)30524-X/h0380http://refhub.elsevier.com/S1742-7061(17)30524-X/h0385http://refhub.elsevier.com/S1742-7061(17)30524-X/h0385http://refhub.elsevier.com/S1742-7061(17)30524-X/h0385http://refhub.elsevier.com/S1742-7061(17)30524-X/h0385http://refhub.elsevier.com/S1742-7061(17)30524-X/h0385http://refhub.elsevier.com/S1742-7061(17)30524-X/h0390http://refhub.elsevier.com/S1742-7061(17)30524-X/h0390http://refhub.elsevier.com/S1742-7061(17)30524-X/h0395http://refhub.elsevier.com/S1742-7061(17)30524-X/h0395http://refhub.elsevier.com/S1742-7061(17)30524-X/h0400http://refhub.elsevier.com/S1742-7061(17)30524-X/h0400http://refhub.elsevier.com/S1742-7061(17)30524-X/h0405http://refhub.elsevier.com/S1742-7061(17)30524-X/h0405http://refhub.elsevier.com/S1742-7061(17)30524-X/h0405http://refhub.elsevier.com/S1742-7061(17)30524-X/h0405http://refhub.elsevier.com/S1742-7061(17)30524-X/h0410http://refhub.elsevier.com/S1742-7061(17)30524-X/h0410http://refhub.elsevier.com/S1742-7061(17)30524-X/h0410http://refhub.elsevier.com/S1742-7061(17)30524-X/h0415http://refhub.elsevier.com/S1742-7061(17)30524-X/h0415http://refhub.elsevier.com/S1742-7061(17)30524-X/h0415http://refhub.elsevier.com/S1742-7061(17)30524-X/h0420http://refhub.elsevier.com/S1742-7061(17)30524-X/h0420http://refhub.elsevier.com/S1742-7061(17)30524-X/h0420http://refhub.elsevier.com/S1742-7061(17)30524-X/h0425http://refhub.elsevier.com/S1742-7061(17)30524-X/h0425http://refhub.elsevier.com/S1742-7061(17)30524-X/h0425http://refhub.elsevier.com/S1742-7061(17)30524-X/h0430http://refhub.elsevier.com/S1742-7061(17)30524-X/h0430http://refhub.elsevier.com/S1742-7061(17)30524-X/h0435http://refhub.elsevier.com/S1742-7061(17)30524-X/h0435

Study on the Mg-Li-Zn ternary alloy system with improved mechanical properties, good degradation performance and different responses to cells1 Introduction2 Materials and methods2.1 Material preparation2.2 Microstructure and surface characterization2.3 Mechanical tests2.4 Electrochemical evaluation2.5 In vitro immersion and hydrogen evolution2.6 Cell viability assay, cell morphology cell apoptosis and NO release2.7 Hemolysis and platelet adhesion2.8 Coagulation four2.9 Statistical analysis

3.Results3.1 Microstructures3.2 Surface characteristics3.3 Mechanical properties3.4 Electrochemical evaluations3.5 Immersion tests and hydrogen evolution3.6 Cell morphology, cell viability assay, cell apoptosis and NO release3.7 Platelet adhesion and hemolysis3.8 Coagulation four

4 Discussion4.1 Mechanical properties4.2 Biodegradation behavior4.3 In vitro cytocompatibility and hemocompatility

5 ConclusionsAcknowledgmentsAppendix A Supplementary dataReferences

study on the mg-li-zn ternary alloy system with improved …lbmd.coe.pku.edu.cn/pdf/2017ab02.pdf ·...

Documents