In Vitro Corrosion and Cytocompatibility of a Microarc

In Vitro Corrosion and Cytocompatibility of a Microarc OxidationCoating and Poly(L‑lactic acid) Composite Coating on Mg−1Li−1CaAlloy for Orthopedic ImplantsRong-Chang Zeng,*,†,‡ Lan-yue Cui,†,‡ Ke Jiang,†,‡ Rui Liu,§ Bao-Dong Zhao,*,§ and Yu-Feng Zheng∥

†College of Materials Science and Engineering, Shandong University of Science and Technology, Qingdao 266590, China‡State Key Laboratory of Mining Disaster Prevention and Control Co-founded by Shandong Province and the Ministry of Scienceand Technology, Shandong University of Science and Technology, Qingdao 266590, China§Affiliated Hospital of Qingdao University, Qingdao University, Qingdao 266590, China∥State Key Laboratory for Turbulence and Complex Systems and Department of Materials Science and Engineering, College ofEngineering, Peking University, Beijing 100871, China

ABSTRACT: Manipulating the degradation rate of biomedical magnesium alloys poses a challenge. The characteristics of amicroarc oxidation (MAO), prepared in phytic acid, and poly(L-lactic acid) (PLLA) composite coating, fabricated on a novelMg−1Li−1Ca alloy, were studied through field emission scanning electron microscopy (FE-SEM), electron probe X-raymicroanalysis (EPMA), energy dispersive X-ray spectroscopy (EDS) and X-ray diffraction (XRD). The corrosion behaviors ofthe samples were evaluated via hydrogen evolution, potentiodynamic polarization and electrochemical impedance spectroscopyin Hanks’ solution. The results indicated that the MAO/PLLA composite coatings significantly enhanced the corrosion resistanceof the Mg−1Li−1Ca alloy. MTT and ALP assays using MC3T3 osteoblasts indicated that the MAO/PLLA coatings greatlyimproved the cytocompatibility, and the morphology of the cells cultured on different samples exhibited good adhesion.Hemolysis tests showed that the composite coatings endowed the Mg−1Li−1Ca alloys with a low hemolysis ratio. The increasedsolution pH resulting from the corrosion of magnesium could be tailored by the degradation of PLLA. The degradationmechanism of the composite coatings was discussed. The MAO/PLLA composite coating may be appropriate for applications ondegradable Mg-based orthopedic implants.

KEYWORDS: magnesium alloy, degradation, microarc oxidation, poly(L-lactic acid), biomaterial

1. INTRODUCTION

The regeneration of bone defects using biomedical materialssuch as bioceramics, stainless steel, titanium, polymers andcomposites has been extensively studied in recent years.1−4 Incomparison with these materials, magnesium (Mg) alloys showgreat potential in clinical applications because their goodbiocompatibility, degradability and mechanical properties aresimilar to those of human bone.5−7 However, the poorcorrosion resistance of Mg alloys leads to a rapid increase inthe pH value and hydrogen collection in the microenvironmentsurrounding the implant.8−10 Therefore, it is of importance tocontrol the degradation of Mg alloys to an appropriate rate.

Usually, the corrosion resistance of Mg alloys can beimproved by means of purifying and alloying,2,11−13 refinementof microstructure by postprocessing14,15 and surface modifica-tion, such as phosphate conversion coatings,16,17 microarcoxidation (MAO) coatings,18,19 polymeric coatings,19 silane-modified coatings,20 layer-by-layer assembly21−25 and layereddouble hydrotalcite.26,27

Numerous studied have been conducted on biomedical Mgalloys, such as Mg−Al (AZ31), Mg−Zn (Mg−6Zn, Mg−Zn−

Received: January 15, 2016Accepted: March 29, 2016Published: March 29, 2016

Research Article

www.acsami.org

© 2016 American Chemical Society 10014 DOI: 10.1021/acsami.6b00527ACS Appl. Mater. Interfaces 2016, 8, 10014−10028

www.acsami.org

http://dx.doi.org/10.1021/acsami.6b00527

Nd), Mg−Mn, Mg-RE (WE43), Mg−Ca (Mg−0.8Ca) andMg−Li (LAE442) alloys.11−15 Ca-bearing Mg alloys (Mg−Ca)have attracted particular attention because Ca is a majorelement of human bone and can refine the microstructure ofMg alloys.11 However, Mg−Ca alloys exhibit poor corrosionresistance.11,13

Additionally, Mg−Li alloys are remarkably malleable andultralight due to the alloying element Li. Li is the only elementthat can change the microstructure of Mg alloys from HCPcrystalline into α (-Mg), α + β (-Li) and β phases withincreasing Li concentration.10,11,14 Usually, β (-Li) phase has abody-centered cubic structure and better ductility than α (-Mg)phase. Higher Li content leads to an enhancement indeformability of binary Mg−Li alloys. The introduction of Cainto Mg−Li alloys gives rise to an increase in tensile strength.In addition, Li possesses higher activity than Mg, and exerts apronounced influence on the corrosion resistance. Liconcentration below 9 wt % in Mg is beneficial for thecorrosion resistance; however, increasing Li concentrationsignificantly accelerates the corrosion of Mg alloys.28

Trace of Li element has a necessary and beneficial functionfor human body. Mg−Li−Al−RE alloy (LAE442) demonstratessuperior in vivo corrosion resistance and good biocompati-bility.10 The Li concentration in bone slightly increased afterimplantation. The contents of Li in liver and bone of rabbit are1.4 μg·kg−1 and 0.08 mg·kg−1, respectively; far below anyphysiological risk level.29 However, the alloy has highercontents of Al and RE elements. The Al element is reportedto be predominantly accumulated in the nervous system andhas been implicated in the pathogenesis of Alzheimer’s disease,and the RE elements including Ce, Pr and Y exhibithepatotoxicity.15 Thus, Al- and RE-free Mg−Li−Ca alloysmay be more suitable for degradable biomaterials.The earlier research of our group shows that the Mg−Li−Ca

alloy exhibits a higher in vitro corrosion resistance than Mg−Caalloys due to the formation of a compact corrosion productlayer.11,14 The microstructure of the dual phase Mg−9Li−1Caalloy is characterized by α (-Mg) and β (-Li) phases andintermetallic compound Mg2Ca particles, and the oxide films,composed of Mg- and Li-bearing compounds acting as barrierlayers, greatly improve the corrosion performance of the alloy.14

Among the various surface modifications, i.e., the MAOcoating, also known as plasma electrolytic oxidation (PEO)coatings, have been widely employed on Mg alloys. MAO/PEOrefers to a high-voltage plasma-assisted anodic oxidationprocess in which plasma discharges cause the partial short-term melting of the oxide layer and promote the formation of ahighly adherent ceramic oxide coating. This coating possessesadvantages such as high hardness, good wear resistance andmoderate corrosion resistance.30

MAO coatings prepared in different solutions, e.g., silicate,phosphate and fluoride electrolytes, exhibit different morpho-logical characteristics, porosities, thicknesses and corrosionrates.31,32 MAO coatings prepared in phytic acid (C6H18O24P6)solution exhibit a better pore uniformity and corrosionresistance than those produced in silicate solution.33,34 Despitethe remarkably advantageous properties of the MAO coating,the presence of micropores and cracks generated during themicroarc discharges provides paths for aggressive ions topenetrate into and react with the substrate, thereby acceleratingcorrosion. Fischerauer et al.35 studied the degradation rates ofuncoated and MAO-coated ZX50 Mg alloy implants in ratfemurs and found that in the first 4 weeks, the degradation rate

for the uncoated Mg alloys is higher than for the MAO-coatedones. However, the reverse trend happens in the subsequent 4weeks. The result reveals that porous MAO coatings onlyimprove corrosion performance for a limited period.36−38 Inother words, it is necessary to seal the pores of the MAOcoatings to enhance the corrosion resistance and provide alonger protection.Poly(lactic acid) (PLA), as a biodegradable and biocompat-

ible polymer, has been approved for human clinical usesincluding small load-bearing bone implants and cardiovascularscaffolds.39−41 It undergoes a degradation process by the simplehydrolysis of an ester bond, and the final products are waterand carbon dioxide. A significant drawback of PLA is that itsdegradation generates acidic products that lower the localsolution pH, accelerating further degradation and triggeringinflammatory and foreign body reactions in vivo.42−44

Substantial investigations, involving those on the corrosionperformance of PLA coatings on Mg and its alloys, have beencarried out.Taking the degradation features of Mg alloys and PLA into

consideration, PLA polymeric coatings on Mg alloys havebecome a potential option to maintain the pH values at anormal level as well as to control the corrosion rate. Abdal-hayet al.21,45 found that polymeric membranes of pristine andhydroxyapatite-doped PLA strongly enhance the corrosionresistance and bending strength of the Mg substrate andameliorates cell growth. Xu et al.46,47 prepared poly(L-lacticacid) (PLLA) and poly(ε-caprolactone) (PCL) coatings ofboth high molecular weight and low molecular weight on pureMg, revealing that both films significantly improve cytocompat-ibility and suppress the pH increase of the medium during Mgdegradation. Chen et al.48 reported that PLA and PCL coatingson high-purity Mg (HP-Mg) improve the corrosion resistance.In addition, some attempts have been made to apply organiccoatings for sealing the micropores and cracks of MAO coatingson Mg alloys. MAO/PLLA composite coatings on WE42, AZ31and Mg−Li−Ca−Y alloys present excellent cytocompatibilityand corrosion resistance.49−52 Nevertheless, the degradationmechanism of the MAO/PLLA film on Mg alloys remainsunclear.The purpose of this investigation is to fabricate a porous

MAO/PLLA composite coating on the novel Mg alloy Mg−1Li−1Ca and to obtain insight into the degradation mechanismand cytocompatibility of the Mg−1Li−1Ca alloy with andwithout the MAO/PLLA coating.

2. MATERIALS AND METHODS2.1. Samples and Coating Preparation Procedures. An as-

extruded Mg−1Li−1Ca (1.26 wt % Li, 0.95 wt % Ca, and balancedMg) alloy was chosen as the substrate. The material was cut intosquare specimens with dimensions of 20 mm × 20 mm × 4 mm for invitro degradation in Hanks’ balanced salt solution (HBSS, 8 g·L−1

NaCl, 0.4 g L−1 KCl, 0.14 g L−1 CaCl2, 0.35 g L−1 NaHCO3, 1 g L−1

glucose, 0.1 g L−1 MgCl2·6H2O, 0.1 g L−1 MgSO4·7H2O, 0.06 g L−1

KH2PO4 and 0.126 g L−1 Na2HPO4·12H2O) with pH = 7.4 and 10mm × 10 mm × 4 mm for cytocompatibility tests. Prior to the coatingpreparation, the samples were mechanically ground with 400−2500grit SiC paper and then ultrasonically cleaned in acetone for 2 min,washed with deionized (DI) water and dried in a dry chamber.

The pretreated specimens were anodized by self-made MAOequipment, consisting of a power supply, a stainless steel barrel, and astirring and cooling system. The MAO treatment was carried out for600 s using the stainless steel barrel as the cathode at a frequency of100 Hz alternating current with a bridge-rectifier waveform, a

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.6b00527ACS Appl. Mater. Interfaces 2016, 8, 10014−10028

10015


breakdown voltage of 170−180 V, and a duty cycle of 50%. Theelectrolyte, comprising 10 g L−1 NaOH and 8 g L−1 phytic acid, wascontinuously stirred during the anodization process. After thepreparation of the coating, the samples were rinsed with DI waterand dried by warm air.PLLA (MW = 200 000, Shandong Institute of Medical Instruments,

China) was dissolved in dichloromethane with continuously stirringfor a 5 wt % solution. The MAO-treated substrates were thenimmersed into the polymeric solution for 60 s, withdrawn and dried ina self-made freeze-drying device. The prepared samples were hung in aprecooled airtight container using dryice. The air was then evacuated,and the container was put in a dry chamber at 55 °C for 12 h to allowsolvent evaporation as well as to achieve a porous film. Prior to thedipping process, all of the specimens were placed in a dry chamber at120 °C for 10 min to remove the moisture from the porous MAOcoating.2.2. Characterization of Coatings and in Vitro Degradation.

2.2.1. Surface Analysis. The surface and cross-sectional morphologiesof the MAO/PLLA coating were investigated using a field-emissionscanning electron microscope (FE-SEM, Nova NanoSEM 450, USA).All samples for the SEM observation were sputtered with gold. Boththe coated and bare samples before and after the in vitro degradationtests were examined by a X-ray diffraction meter (XRD, Rigaku D/Max 2500PC, Japan) with a Cu target (λ = 0.154 nm) at a scanningrate of 0.02 s−1 in the 2θ range of 10°−90°. The chemical compositionof the coating was inspected through energy-dispersive X-rayspectroscopy (EDS, Oxford Isis) affiliated with electron probe X-raymicroanalysis (EPMA, JXA-8230, Japan).2.2.2. Electrochemical Corrosion Test. The potentiodynamic

polarization (PDP) and electrochemical impedance spectroscopy(EIS) were performed on a potentiostat (PARSTAT 2273) inHBSS. All of the electrochemical tests were conducted in a classicalthree-electrode system that consists of the sample as the workingelectrode (1 cm2), a platinum plate as the counter electrode, and asaturated calomel electrode (SCE) as the reference electrode. ThePDP curves were recorded with a sweep rate of 2 mV·s−1, and EISmeasurements were acquired from 105 to 10−2 Hz using a 5 mVamplitude perturbation.2.2.3. In Vitro Immersion Tests. The substrate, MAO and MAO/

PLLA-coated substrates were prepared in triplicate and immersed in300 mL of HBSS. The ratio of the sample surface area to solutionvolume was 1:25 cm2·mL−1, and the temperature was controlled at 37± 0.2 °C by a water bath. The samples were taken out after immersionfor 140 h, with the pH values of the solution and the hydrogenevolution volume were recorded every hour during the immersion.The schematic illustration of the hydrogen evolution test was shown inFigure 1. The hydrogen evolution rate (HER) as a function ofimmersion time can be related through

=Vst

HER H(1)

where VH is the hydrogen evolution volume (mL), s and t are theexposed area (cm2) and immersion time (h), respectively.The pH value of the solution was measured by a pH meter (PH-

400). Immersion period of 140 h was predetermined upon the possibleobservation on the failure of the MAO/PLA coating, and for theinsight into the corrosion mechanism of the MAO/PLA coating.2.3. Cytocompatibility and Hemolysis Tests. 2.3.1. Cell

Culture and Preparation of Extracts. Murine MC3T3-E1 osteo-blast-like cells (Cells Resource Centre of Shanghai Institute forBiological Science, Shanghai, China) were utilized for cell culturestudies. Cells were cultured in minimum essential medium alpha (α-MEM, Hyclone) containing 1% penicillin/streptomycin antibioticsand 10% fetal bovine serum (Gibco, Australia) at 37 °C in ahumidified atmosphere of 5% CO2. At 78−80% confluence, the cellswere enzymatically detached with a minimum amount of trypsin−EDTA (Sigma-Aldrich, USA) and resuspended in culture medium.The cells were then recultured until the third passage, which was usedfor the experiments. The specimens were sterilized by ethylene oxideand then placed in clean centrifuge tubes with added α-MEM at a ratio

of 3 mL·cm−2. After 24 h of incubation at 37 °C, the samples wereremoved, and the media containing the extracts of different sampleswere collected for the following tests.

2.3.2. MTT Assay. The 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was used to determine theproliferation and viability based on the reductive cleavage of MTT(a yellow salt) to formazan (a dark blue/purple compound) by themitochondrial dehydrogenase of living cells. The cell suspension wasadjusted to a cell density of 1 × 105 mL−1, and then 200 μL was addedinto each well of 96-well plates, which were sealed by phosphatebuffered solution (pH 7.4, PBS, Solibo) in the outer 40 wells. Thewells with the cell suspension were divided into 4 groups: the Mg−1Li−1Ca substrate group, MAO group, MAO/PLLA group, controlgroup (culture media containing cell suspension) and blank group(culture media alone), each with 5 wells. After 24 h of incubation, theculture media were removed, together with the leaching liquor of thesubstrates. The bare substrate, MAO and MAO/PLLA-coated sampleswere added into the wells of the corresponding groups in volumes of100 μL; whereas equivalent volumes of culture media were supplied tothe control group and blank group. The culture medium of each wellwas changed every other day during the testing period, and 20 μLMTT solution was injected on each of days 1, 3 and 5. The 96-wellplates were kept in the incubator for 4 h. After incubation, the mediumwas removed, and the formazan crystals were solubilized in DMSO.The absorbance (A) was measured by a universal microplatespectrophotometer (722, Shanghai Precision Science Instrument,China) at a wavelength of 490 nm, and the relative growth rate(RGB %) of the cells was calculated by the relation below

= ×AA

RGB(%) 100%t

c (2)

where At is the absorbance of the testing specimen and Ac is theabsorbance of the control groups. The results are expressed as themean ± standard deviation of five samples per group and werecompared using Student’s t-test, with a significance level of P < 0.05.

2.3.3. Alkaline Phosphatase (ALP) Assay. An alkaline phosphatasecolorimetric assay kit (Nanjing Jiancheng, China) was used to evaluatethe osteoblast differentiation. The cell suspension was adjusted to a

Figure 1. Schematic illustration of the hydrogen evolution test.



10016


cell density of 1 × 105 mL−1, and 2 mL of the suspension was addedinto the first four wells of a 6-well plate. After being seeded for 24 h,the culture media were removed and rinsed twice by PBS. Then, α-MEM and the extracts of the bare substrates, the MAO-coatedsamples, and the MAO/PLLA-coated samples were added to each wellin sequence, and the plate was incubated under standard cell cultureconditions. The culture media were discarded, and the wells wererinsed twice by PBS after 5 and 7 days of incubation. Then, 500 μL of1.5% Triton X-100 solution was added into each well, and the platewas placed into a refrigerator for 10 h at 4 °C. The four groups of theMg−1Li−1Ca substrates, the MAO and MAO/PLLA coated samples,and the control were settled with 5 wells. The correspondingsupernatants were supplemented by the volume of 50 μL per well in a96-well plate. The addition of reagents was following the instructionsof the ALP kit, and the absorbance was measured using a universalmicroplate spectrophotometer at a wavelength of 520 nm andconverted into the amount of ALP (U·L−1).2.3.4. Cellular Adhesion and Morphology Imaging. A 1.5 mL

suspension was seeded onto the substrates, the MAO and MAO/PLLA-coated samples after sterilization by ethylene oxide at a celldensity of 5 × 104 cells per well in 24-well plates. After 24 h ofculturing in a cell incubator, the specimens were taken out andimmobilized by 4% glutaraldehyde for 12 h and then dehydrated withethanol step by step. Lyophilization was applied in the dehydration tomaintain the structure and morphology of the cells. Following thedrying process, the samples were coated with platinum, and the cellmorphology was observed using FE-SEM.2.3.5. Hemolysis Test. The hemolysis test in vitro was conducted as

follows: 10 mL fresh rabbit arterial blood from a healthy New Zealandwhite rabbit, containing 0.5 mL potassium oxalate (20 g·L−1)anticoagulant was added into the culture media with different extractsprepared in the previous step, kept at 37 °C for 60 min, and thencentrifuged at 3000 rpm for 5 min. Three sets of parallel tests weresettled, and the A of the supernatant solution was measured using thespectrophotometer at a wavelength of 545 nm. The hemolytic ratio(HR) was obtained by the following equation

=−−

×A A

A AHR(%) 100%t nc

pc nc (3)

where At is the A of the testing specimen, Anc is the A of the negativecontrol group (10 mL of culture medium with 0.2 mL of dilutedblood) and Apc is the A of the positive control group (10 mL distilledwater with 0.2 mL diluted blood).

3. RESULTS AND DISCUSSION3.1. Microstructure of the Alloy and Morphologies of

the Coatings. The microstructure of the Mg−1Li−1Ca alloyafter etching was characterized by the α-Mg phase and theintermetallic compounds Mg2Ca dispersed in the grain interiorsand at the grain boundaries (Figure 2), in which the Li atoms

are solid-soluted. The Mg2Ca phase, with a higher corrosionpotential than the α-Mg matrix, may lead to microgalvaniccorrosion between the Mg2Ca phase and its neighboring α-Mgmatrix.14 The typical morphology of the MAO coating inFigure 3a,c demonstrates the presence of micropores andmicrocracks. Interestingly, a porous PLA film was also preparedon the MAO coating, as depicted in Figure 3b,d. The EDSresult of the MAO coating, as shown in the inset of Figure 3a,demonstrates that the MAO coating is mainly composed of Mg,O, and traces of Ca and P, indicating the presence of MgO andcalcium phosphates.51 It is likely that the Ca and P of the MAOcoating are from the substrate and electrolytes used in theformation of the coating, respectively, while the presence of Cand O designates the MAO/PLLA coating, as shown in theinset of Figure 3b.Figure 4 shows the cross-sectional SEM micrographs and

elemental mapping of the substrates with MAO/PLLA films.The MAO and the PLLA coatings have thicknesses ofapproximately 3 and 10 μm, respectively. As illustrated inFigure 4b, the intensity of C decreased when the EDS scan lineapproached the PLLA/MAO interface, comparing with thereverse trend for Mg and O. This is due to the differentchemical compositions of each layer, which are consistent withthe EDS spectrum in the insets of Figure 3. Note that in Figure3a the MAO coating contains traces of Ca and P elements,which originate from the Mg−1Li−1Ca alloy and the solutionwith phytic acid, respectively. Thus, the line scan of both Caand P can be ignored; and the concentration profiles of the C,O and Mg elements from the substrate to resin can be evident.Nevertheless, it can be recognized that C diffused into theMAO layer, indicating that the PLLA sealed the porous filmand the PLLA coating and MAO coating interlocked.

3.2. Hydrogen Evolution Rates and pH Measure-ments. The hydrogen evolution rate (HER) vs immersiontime curves of the coated and uncoated samples are shown inFigure 5. In the initial stage of immersion, the HER of thesubstrates is much higher than that of the coated samples, andsome bubbles can be seen on the surfaces of the substrates,whereas no bubbles were observed on the coated ones.However, the HER of the MAO-coated specimens began toincrease rapidly at an immersion of 50 h and finally becameslightly higher than that of the substrates after 85 h ofimmersion, compared with the constant HER of the MAO/PLLA-coated samples.It is noted that simialr degradation behavior occurs on the

MAO coating on Mg−1.21Li−1.12Ca−1.0Y alloy based on ourprevious study.51 Namely, the HERs of the MAO coating onMg−1.21Li−1.12Ca−1.0Y alloy (0.142 mL·cm−2·h−1) sur-passes that of the substrate after an immersion time of 9 h.Interestingly, the HERs of the MAO coating on the Mg−1Li−1Ca alloy (0.004 mL·cm−2·h−1) is much lower than that on theMg−1.21Li−1.12Ca−1.0Y alloy, and it is until 85 h that theHERs of the MAO coating surpasses that of the substrate,although the MAO thickness (about 3 μm) on the Mg−1Li−1Ca alloy is much lower than that (approximately 10 μm) onthe former alloy.51 The result designates that the MAO coatingon the Mg−1Li−1Ca alloy possesses superior corrosionresistance to that on the Mg−1.21Li−1.12Ca−1.0Y alloy, thatis, the thickness of MAO is not a critical factor influcencing thecorrosion resistance. This abnormal scenario may be attribut-able to the different chemicals used in the prepared solutions.The MAO coatings were prepared in a silicate electrolytecontaining 20 g·L−1 NaOH, 20 g·L−1 Na2SiO3, 15g·L

−1 NaB4O7

Figure 2. Optical morphology of the microstructure of the Mg−1Li−1Ca alloy.



10017


and 10 g·L−1 C6H5Na3O7 together with 7.5 g·L−1 phytic acidfor the Mg−1.21Li−1.12Ca−1.0Y alloy, and in an electrolytecomprising 10 g·L−1 NaOH and 8 g L−1 phytic acid for theMg−1Li−1Ca alloy. As a result, a thinner and more compactMAO coating formed on the Mg−1Li−1Ca alloy (Figure 4a,c).After 140 h of immersion, the average HER of the substrate,

the MAO and MAO/PLLA-coated samples were 2.12, 1.58 and0.56 mL·cm−2·h−1, respectively. The HER indicate that the

MAO/PLLA coating greatly enhanced the corrosion resistanceof the Mg−1Li−1Ca alloy and also reveal that the disadvantageof the MAO coatings in the corrosion process that could beimproved by PLLA modification are consistent with Fischera-uer’s research.35

Figure 6 designates the changes in the pH value of thesolutions with the immersion time. The process can be dividedinto four stages. In stage one, there was a significant increase inthe solution pH for each sample, resulting from the corrosionof the Mg alloys. For the bare substrate and the MAO-coatedsamples, this was followed by a decrease that led to a period ofstable pH values, which can be defined as stage two. During thisstage, stable oxide films have been formed, so the reactionslowed down after the initial 10 h of immersion. However, it isnoteworthy that H2O and Cl− can diffuse into the PLLAcoating and react with the MAO coating and substrate beforethe hydrolysis degradation of the polymer films in the first 10 hof immersion, which can explain the slight increased state of the

Figure 3. SEM morphologies of (a) the MAO coatings; (b) the MAO/PLLA composite coatings formed on the Mg−1Li−1Ca alloys, (c and d) thehigh magnification of panels a and b, respectively. The insets in panels a and b correspond to the EDS spectra of the MAO and MAO/PLLAcoatings, respectively.

Figure 4. (a) SEM cross-sectional morphology, and (b) the EDS linescan mapping across MAO/PLLA coating.

Figure 5. HER of the substrate, MAO coating and MAO/PLLAcoating immersed in HBSS for 140 h.



10018


pH value of the MAO/PLLA samples. Interestingly, for thePLLA coating, the solution pH decreased to the lowest value of7.3 after 82 h of immersion, followed by a slow continuousincrease, which can be defined as stage three. This can be theresult of the chemical corrosion of the MAO coating and theelectrochemical corrosion of the substrate, which releasedalkaline corrosion products into the solution after thedelamination and collapse of the PLLA coating. After 110 hof immersion, the pH value of the MAO samples surpassed thatof the bare substrates, which exhibited a similar trend to theHER. However, the average pH value of the MAO/PLLA-coated sample is lower than that of the bare substrates and theMAO-coated samples during the immersion. This result isprobably attributable to the degradation of PLLA, whichneutralizes the hydroxyls released during the corrosion of Mgwith the lactic acid produced after the hydrolysis of the esterbond. Fortunately, our previous study, regarding the corrosionof MAO/PLLA on Mg−Li−Ca−Y alloy, has confirmed theacidicity of PLLA coating; a linear decrease in solution pH overimmersion time for the PLLA coating on the PTFE platediscloses the formation of acidic products during itsdegradation.51

3.3. Electrochemical Behavior. PDP curves obtained forthe coated and uncoated samples in HBSS with and withoutimmersion are shown in Figure 7, showing that the structuresand properties of the surfaces changed during the immersiontime. The free corrosion potential (Ecorr) and corrosion currentdensity (Icorr) were derived from these data by the Tafel

extrapolation method and are summarized in Table 1. It can beseen that the Ecorr of the samples without the immersion follows

this consequence: MAO/PLLA > MAO > bare substrate. Thisindicates that the coated samples exhibited a lower thermody-namic tendency toward electrochemical corrosion, the MAO/PLLA coating showed the lowest tendency to participate in theanodic reaction, and the Icorr of the MAO/PLLA-coated sampleis about 2 orders of magnitude lower than that of the baresubstrate, which indicates that the corrosion resistance iseffectively improved. After immersion in HBSS for 140 h, theEcorr decreases in the order of MAO/PLLA > substrate > MAO,so that the MAO-coated sample exhibits a stabilizedthermodynamic stability during the immersion time. Comparedwith the substrate, an increase in Icorr for the MAO coating is inpronounced agreement with the results from HER (Figure 5)and the reverse change in solution pH (Figure 6), implying thatthe MAO coating suffered from severe attack from the solutionand the delamination of the MAO coating. A similar scenario isalso discerned on the degradation of the MAO/PLLA coatingon the Mg−Li−Ca−Y alloy.51 The Ecorr and Icorr of the baresubstrate were, however, enhanced and lowered, respectively,which is probably related to the protection of the compactcorrosion product layer. The MAO/PLLA-coated sampleretains significant corrosion resistance, which is consistentwith the HERs. The PDP curve of the MAO/PLLA coating,neverthelss, show no significant difference among the threesamples after immersion in HBSS for 140 h, which are ascribedto the peeling-off of the MAO coating and the swelling of thePLLA coating after a long period of immersion.53

Figure 8a shows Nyquist plots of the uncoated and coatedsamples with and without immersion in HBSS for 140 h. All ofthe Nyquist plots were fitted with the corresponding equivalentcircuits, which are also designated in Figure 8b−g. As shown inthe Nyquist plots, the MAO/PLLA-coated samples exhibitedthe best corrosion resistance, and the impedance of the MAO-coated and immersion treated MAO/PLLA-coated samplesdecreased, while that of the bare substrate increased after beingimmersed for 140 h. The experimental data were processedusing ZSimpWin 3.4, and the χ2 values of the fitting results arewithin a permissible range (Table 2).Figure 8b,d,f discloses the Nyquist plots and corresponding

equivalent circuits before immersion. It can be seen that theNyquist plots consist of two capacitive loops at high and middlefrequencies, followed by an inductive loop in the low-frequencyrange. The capacitive loop in the high-frequency rangecorresponds to the charge transfer resistance (Rt) and interfacediffusion constant phrase element (CPE2) in Figure 8b and theinterface double-layer capacity (Cdl) in Figure 8d,f. The latterrepresents constant phase elements, defined as

ω= −Z Y j[ ( ) ]nCPE 0

1(4)

Figure 6. Changes in the pH value of HBSS with immersion for thecoated and uncoated samples.

Figure 7. PDP curves obtained for the substrate, MAO coating andMAO/PLLA coating with and without immersion of 140 h.

Table 1. Parameters of the PDP Curves of the Samples withand without Immersion of 140 h

sample immersion time, h Ecorr, V vs SCE Icorr, A cm−2

substrate-0 h 0 −1.66 2.48×10−5

MAO-0 h 0 −1.50 1.60×10−6

MAO/PLLA-0 h 0 −1.46 2.96×10−7

substrate-140 h 140 −1.61 7.21×10−6

MAO-140 h 140 −1.62 1.03×10−5

MAO/PLLA-140 h 140 −1.60 4.19×10−6



10019


where Y0 is the constant of the CPE, jω is the complex variablefor sinusoidal perturbations with ω = 2πf, and n is the exponentof CPE with values between 0 and 1. The CPE acts as a pureresistor when n equals 0 and as an ideal capacitor when n equals1. The value of n is associated with the nonuniform distributionof current as a result of both the degree of roughness of thesurface as well as the degree of the oxide’s heterogeneity. Themiddle capacitive loop is attributed to the film resistance (Rf)and the film capacitance (CPE1) of the electrolyte penetratingthrough the corrosion product layer on the surface of thesubstrate. The increased Rf from 908.9 to 6248 and 75490 Ω·cm2 confirms that the MAO/PLLA-coated sample retains goodcorrosion resistance. The inductive loop, corresponding to aninductor L and a resistor RL at the low frequency, is ascribed tothe absorption and peeling of the corrosion products such asMg(OH)2.

54 Figure 8d,f shows similar Nyquist plots, which canbe fitted with one equivalent circuit, exhibiting the multiple-layer structure of both coatings. The circuit contains the Rt, Cdl,

Rf, and the polarization resistance (Rp), the outer porous layercapacitance (Cf) and the inner barrier layer capacitance (CPE1).Nyquist plots and equivalent circuits of the bare substrate

and coated samples with 140 h of immersion in HBSS areshown in Figure 8c,e,g, respectively. The Nyquist plot of thesubstrates demonstrates a capacitive loop at a high frequency inFigure 8c, representing the charge transfer process, and anoblique line in a middle low frequency range, corresponding tothe corrosion reaction being controlled by the diffusionprocess.55 Therefore, as depicted in Figure 8c, the equivalentcircuit of the substrate after immersion is similar to that of thesubstrate without immersion, with replacement of the inductorwith a Warburg resistance. The reason for this change is that asthe immersion time is prolonged, the produced corrosionproducts cover the surface of the substrate, leading to theformation of a compact corrosion product layer that isolates thesolution from the substrates. Therefore, the diffusion processplays an important role in parallel with the absorption anddelamination of the corrosion products without immersion,

Figure 8. (a) Nyquist curves for the coated and uncoated samples; (b−g) fitted curves and equivalent circuits for the (b) substrate, (d) MAO coatingand (f) MAO/PLLA coating without immersion; (c) substrate, (e) MAO coating and (g) MAO/PLLA coating with immersion of 140 h.



10020


which can be equivalent to an inductor. As for the Nyquist plotsin Figure 8e,g, an inductor L and a resistor RL simultaneouslyappear at the low frequencies due to the formation of corrosionpits on the coatings. However, the MAO/PLLA coatingmaintained its integrity, whereas serious corrosion occurredon the MAO coatings due to a hydration reaction and thedissolution of Mg, resulting in localized destruction and the lossof the protective layer.

3.4. Cytocompatibility and Hemolysis Tests. Table 3shows the result of the hemolysis test of the bare substrate,

MAO and MAO/PLLA coated samples. Compared with thebare substrate, the HR of the MAO coating was lowered by upto 12.48%, indicating that it may lead to severe hemolysis whencontacted with the blood. The MAO/PLLA coating exhibited aHR of 0.17%, which is much lower than 5% (a judging criterionfor excellent blood compatibility), indicating the excellentblood compatibility of the composite coatings.The cytotoxicity of the MAO and MAO/PLLA-coated

samples was determined using MC3T3-E1 cells and theMTT assay (Figure 9). The relative growth ratio (RGR) of

the three groups is less than 100% after 24 h of culture, whereasthe RGR of the MAO/PLLA coating slightly exceeds 100%after a culture of 72 h. MC3T3-E1 cells in the extract media ofthe MAO/PLLA coating showed an obviously higherproliferation rate and vitality than those on the Mg−1Li−1Casubstrates and the MAO coating. It is noteworthy that thelowest cytotoxicity of the MAO-coated samples may beascribed to the influence of the electrolytes, which persistedin the micropores after the MAO process. It is also interestingto note that the RGR of the Mg−1Li−1Ca substrates is close tothat of the composite coatings due to the protective filmgenerated during the extraction process, which is consistentwith the results of the ions released during immersion, such asMg2+, Ca2+ and Li+, which are of small volume less than thetolerated contents of osteoblasts.28T

able

2.Electrochem

ical

DataObtainedby

Equ

ivalentCircuitFittingof

EIS

Curvesof

UncoatedandCoatedSamples

withandwitho

utIm

mersion

for140h

Samples

Rs,Ω·cm

2Rt,Ω·cm

2Rp,Ω·cm

2Rf,Ω·cm

2Cdl,F

·cm

−2

Cf,F·cm

−2

CPE

1,Ω

−1 ·S

n ·cm

−2

CPE

2,Ω

−1 ·S

n ·cm

−2

CPE

3,Ω

−1 ·S

n ·cm

−2

RL,Ω·cm

2L,

H·cm

2W,Ω

−1 ·S

0.5 ·cm

−2

χ2

substrate-0h

70.2

384.5

908.9

1.5×

10−8

1.7×

10−3

2.6×

10−5

122

586.4

2.0×

10−4

MAO-0

h127.3

2.3×

104

1174

6248

5.4×

10−7

2.9×

10−8

1.5×

10−6

2.6×

10−4

MAO/PLL

A-0

h95.4

1.2×

104

1.9×

104

7.5×

104

1.3×

10−5

3.6×

10−10

3.7×

10−7

4.3×

10−4

substrate-140h

179.9

2722

4810

1.1×

10−4

3.5×

10−7

3.7×

10−4

3.8×

10−4

MAO-140

h142.3

670.6

1556

4.4×

10−6

6.1×

10−5

3139

2.6×

104

5.5×

10−4

MAO/PLL

A-140

h184.7

2.3×

104

334

2895

2.2×

10−9

9.8×

10−6

1.6×

10−5

3.2×

104

1.0×

105

2.0×

10−4

Table 3. HR (%) of the Mg−Li−Ca Substrate, MAO, andMAO/PLLA (n = 5)

samples HR, %

substrate 61.35 ± 0.36MAO 12.48 ± 0.18MAO/PLLA 0.17 ± 0.04

Figure 9. RGR of MC3T3-E1 cells cultured for different times inextract media from the substrates, MAO and MAO/PLLA coatedsamples compared to in α-MEM culture medium. Error bars represent± S for n = 5 and P < 0.05, as indicated by the asterisk (*).



10021


ALP activity is one of the most widely used markers for earlyosteoblastic differentiation.56,57 ALP is a ubiquitous enzymethat catalyzes the hydrolysis of phosphate esters at alkaline pH.The differentiation of MC3T3 cells on differently treatedsamples was evaluated in up to 7 days of culture using ALP asan early phase marker. Figure 10 shows the ALP activity of the

osteoblast culture in the extract media of the Mg−1Li−1Casubstrates, MAO and MAO/PLLA-coated samples, as well aspure culture media as a control. The osteoblasts cultured in theextracts of the MAO/PLLA coatings displayed a highest ALPexpression among the samples, which indicates a strongerability for cell differentiation.56,57 The ALP expression of thesubstrates and the MAO coatings shows a parallel trend to theMTT assay.The morphologies of MC3T3-E1 cells cultured on the

coated and uncoated samples for 24 h are presented in Figure11. After 24 h of the culture, it is clear that the cells cultured onthe MAO/PLLA coatings exhibited good adhesion, which canbe identified by the number of spreading cells as well as thespreading shape and areas. Additionally, the cells displayednumerous filopodia extensions (Figure 11e,f). It is especiallynoteworthy that the edges of the pores on the MAO/PLLAcoatings can be easily attached to by the filopodia, whichstrongly enhances the cell adhesion. Although the cells culturedon the substrates and the MAO coatings also adhered to thesurface and filopodia protrusions were also observed in Figure11a−d, their growth seemed to be inhibited.

4. DISCUSSIONThe surface appearances of the coated and uncoated samplesafter immersion in HBSS for 140 h are shown in Figure 12. Acompact corrosion layer with micro cracks is formed on thesurface of the substrate (Figure 12a). Corrosion pits on theMAO coating destroyed its integrity (Figure 12b). However,the surface of the MAO/PLLA coating showed no obviousvariation on most of its surface (Figure 12c) with few corrosionpits covered by white corrosion products. With the integrityand mechanical property of the duplex coating beingundermined during the immersion by the diffusion of H2O,solution ions and the release of H2, the corrosion productscentralized under the polymeric PLLA film. Then, thedelamination and collapse of the PLLA coating occurred,which was demonstrated by the distinct shearing boundary on

the MAO/PLLA surface (Figure 12d). The EDS spectra of eachsample after immersion in HBSS for 140 h is also demonstratedby the insets in Figure 12, revealing that the MAO/PLLAcoating had similar elemental compositions on the substrateand the MAO coating, mainly containing O, Mg, P and Ca(Figures 12a,b).Interestingly, the centralization of corrosion products around

the corrosion pit in Figure 12d was destroyed by the stream ofelectrons released during the immersion, exposing thecorrosion pit to the microscope. An opening pore can beclearly observed in Figure 13, and C, O, Mg, Ca and P appearto be uniformly distributed in the corrosion products. It isnoteworthy that around the corrosion pit, the concentrations ofMg and O are much higher than those of Ca and P, and thedistribution of Ca corresponds to that of P. The elementaldistribution resulted from the formation of different corrosionproducts, including Mg(OH)2, which centralized around thecorrosion pit, and hydroxyapatite (HA).16

Figure 14 shows the cross-sectional SEM morphology andEDS elemental mapping of the MAO/PLLA coating afterimmersion. The composite film remained integrated andcompact after immersed in HBSS for 140 h, with a smallquantity of corrosion pits beneath the polymeric coating, thatwere produced by the diffusion of H2O and Cl− through thePLLA and MAO coating into the substrate and the dissolutionof Mg, thus resulting in the blistering of the PLLA coating dueto the hydrogen evolution. As shown in the EDS elementalmapping, C, O and Mg can be detected on the cross-section ofthe corrosion pit, suggesting the formation of the hydrolysatesof the PLLA coating and the Mg(OH)2 corrosion products ofthe alloy.Figure 15 shows the XRD patterns of the coated and

uncoated samples with and without immersion for 140 h. Thepresence of MgO indicates that the ceramic coatings are formedduring MAO treatment in electrolytes containing phytic andNaOH, which is different from the coatings formed in silicateelectrolytes.27 Interestingly, the presence of calcium lactate((C3H5O3)2Ca) can be detected on the MAO/PLLA coatingsamples after immersion in HBSS, demonstrating the reactionsbetween chemical species such as Ca2+ from the corrosionmedium and the dissolution of the Mg−1Li−1Ca alloy, andoligomers or monomers such as lactide and lactic acid releasedby the degradation of the PLLA film.The cytotoxicity and biocompatibility of clinical materials are

also greatly influenced by the corrosion. During the corrosionprocess, local alkalization and enrichment in Mg2+ exert asignificant impact on the physiological balance. It is postulatedthat Mg-based materials degrade too fast, and the highconcentration of Mg2+ in the solution is responsible for thehigh hemolysis rate.58 Although the MAO coatings reduced theHR to 12.85%, which is much lower than that of the Mg−1Li−1Ca substrates (61.35%), it is not suitable for utilization asblood-contacting materials for the potential danger toerythrocytes. After being sealed with PLLA, MAO/PLLAcoatings, with a HR of 0.17%, would have little influence on theblood.It is well-known that cells are very sensitive to changes in the

microsurrounding environment, such as the sharp changes inthe pH value and Mg2+ concentration. The sharp increase in thepH value and concentration of Mg2+ released by thedegradation of the substrates and the MAO coatings mayhinder cell growth (Figures 9−11). As a biodegradable material,PLLA not only works as a physical barrier layer but also

Figure 10. Proliferation and differentiation of MC3T3-E1 cellscultured for different times in extracts from the substrates, MAOand MAO/PLLA-coated samples as well as α-MEM culture medium.Error bars represent ± S for n = 3 and P < 0.05, as indicated by theasterisk (*).



10022


provides a suitable condition for cell adhesion and proliferationon the Mg−1Li−1Ca alloys (Figure 11e,f). Combining MAOwith PLLA can take advantage of the superiority of the twocoatings on biodegradable Mg alloys.As is well-known, PLA-based materials belong to the family

of aliphatic polyesters, so their ester groups are hydrolyticallydegraded in the presence of water according to the followingreaction:59

− + → − +− −COO H O COOH OH2 (5)

Thus, PLA-based materials degrade via a hydrolyticautocatalytic degradation process at the chain end and in themiddle of the chain, which are called exo-chain cleavage andendo-cleavage, respectively. However, the cleavage mechanismof PLA dissolved in solution depends on the media pH suchthat the hydrolytic degradation proceeds with a chain-endscission mechanism in lactyl monomer unit, forming lactic acidin acidic solution, whereas in alkaline solution, hydrolyticdegradation takes place via backbiting to form a lactide unit,which is further hydrolyzed to give lactic acid.60 Degradation ofthe PLA coating in HBSS at 37 °C and pH 7.4 is similar to invivo degradation. The degradation in neutral solution seems tobe selective and often occurs in amorphous regions, producing

hydrophilic groups (−OH and −COOH). The catalytic group(−COOH) is subsequently condensed in the amorphousregion between crystalline regions, resulting in the accelerationof the hydrolytic degradation in a restricted area.61

The degradation mechanism of PLA-based materials includethree models:62 a surface corrosion mechanism, bulk corrosionmechanism and core-accelerated bulk corrosion mechanism,with a critical thickness (Lcritical) of polymeric films that can beused as a criterion index to evaluate which corrosion takesplace. With the decrease of the thickness of the polymer films,the hydrolytic degradation rate of the material surface slowsdown, compared with the diffusion rate of water molecules orcatalytic substances within the material, and the mechanismmodel turns from surface corrosion to bulk corrosion. ThePLLA coating in this experiment has a thickness (approximately10 μm) that is lower than Lcritical of 0.5−2 mm. Therefore, thebulk corrosion occurred during the immersion test, which canbe divided into three stages:63 (1) initial water diffusion,absorption and hydration of the PLLA coatings, (2) gradualdecrease in molecular weight and (3) formation and dissolutionof water-soluble oligomers and monomers. Accompanied with adecrease in the molecular weight and crystallinity, the thicknessof the films decreased, causing a release of formed oligomers

Figure 11. SEM morphologies of MC3T3-E1 cells cultured on (a,b) the Mg−Li−Ca alloys, (c,d) the MAO coating and (e,f) the MAO/PLLAcoating for 24 h.



10023


and monomers to the solution, resulting in a reducedautocatalytic effect and a lower degradation rate.Furthermore, the MAO film can significantly influence the

degradation behavior of the polymer film. As is well-known, Mgis highly susceptible to galvanic corrosion,64 leading to theformation of Mg(OH)2 during the chemical reactions, and thepresence of Cl− can dissolve Mg(OH)2, leading to the releaseof OH− to the corrosion media. These reactions follow as65

→ ++ −Mg Mg 2e2(6)

+ → + ↑− −2H O 2e 2OH H2 2 (7)

+ → + ↑Mg 2H O Mg(OH) H2 2 2 (8)

+ → +− −Mg(OH) 2Cl MgCl 2OH2 2 (9)

Figure 12. SEM morphologies and EDS element analysis of surface of the samples after 140 h of immersion: (a) the substrate; (b) the MAOcoatings; (c, d) the MAO/PLLA coatings.

Figure 13. SEM morphology of a corrosion pit on the MAO/PLLA surface after 140 h of immersion: (a) SEM micrographs; the elemental mappingsof EDS: (b) C, (c) O, (d) Mg, (e) Ca and (f) P.



10024


Therefore, the degradation of Mg eventually results in alocally alkaline environment, which can result in the

consumption of the acidic products released by the degradationof the PLLA coatings, accelerating and promoting thehydrolytic degradation of the polymers as well as the loss ofthe Mg(OH)2 layers. Our earlier study

14 demonstrates that themicrostructure of Mg−1Li−1Ca is composed of an α-Mg phaseand intermetallic compound Mg2Ca particles, which result in aninitial corrosion around the Mg2Ca particles, generatinghydrogen bubbles and Mg(OH)2. The corrosion productsgradually concentrate on the interface of the MAO coatings andsubstrate during the reaction and destroy the MgO layer as

+ →MgO H O Mg(OH)2 2 (10)

Simultaneously, the H2 liberated also increases the internalstress under the MAO coatings and PLLA coatings andundermines the mechanical properties of the coatings, leadingto the generation of peeled off pores and cracks on the PLLAfilms and finally disintegration into fragments (Figure 12).65,66

The detachment of the PLLA coating may be detrimental to

Figure 14. Cross-sectional morphology of MAO/PLLA coating after 140 h of immersion in HBSS: (a, b) cross-sectional morphologies; EMPAimage of cross section of blister: (c) elemental mappings of EDS: (d) C, (e) O and (f) Mg.

Figure 15. XRD patterns of the substrate, MAO coatings and MAO/PLLA coatings before and after immersion for 140 h.



10025


application as stent materials, but the composite coating stillshows great potential in bone regeneration applications. Aboveall, the MAO/PLLA composite coating degraded predom-inately at the inner layer and exhibited features of localizedcorrosion and pitting corrosion, resulting in the concentrationof corrosion products inside and on the surface of coatings,finally leading to localized destruction and collapse. Thus, adegradation mechanism is schematically illustrated in Figure16a,b, indicating that the degradation of PLLA on the Mg−1Li−1Ca alloy experienced four steps: (1) water diffusingthrough the PLLA film onto the MAO coating and thesubstrate due to the diffusion rate of water being faster than thedegradation of the PLLA film, (2) chemical corrosion of theMAO coating (eq 10), (3) electrochemical corrosion of thesubstrate (eq 6) and (4) the swelling, hydrolysis (eq 5) andrupture of the PLLA film.

5. CONCLUSIONS

Porous MAO/PLLA composite coatings were fabricated bydip-coating and freeze-drying to improve the corrosionresistance of Mg−1Li−1Ca alloys, and the corrosion behaviorsof the coatings in HBSS were investigated by SEM, EPMA,XRD and corrosion measurements. The conclusions are asfollows: (1) The MAO/PLLA coatings significantly enhancedthe corrosion resistance, increasing the Ecorr from −1.66 to−1.44 V and decreasing the Icorr by approximately 2 orders ofmagnitude. (2) The corrosion predominantly occurred at theinterface of the MAO coating and the substrate, and theaccumulation of corrosion products as well as hydrogenbubbles caused swelling and blistering of the PLLA coatingdue to the faster penetration rate of water into the PLLAcoating than its degradation rate. (3) The acidic and alkalineproducts, respectively resulting from the degradation of PLLAand the corrosion of Mg, neutralized each other to maintain astable and appropriate pH value in HBSS. (4) The MAO/PLLA coating on Mg alloys greatly enhances its cytocompat-ibility and blood compatibility, and provides an appropriatesolution pH and porous microstructure, which is helpful for theattachment of cells. The MAO/PLLA coating may be apromising approach for bone tissue regeneration engineering.

■ AUTHOR INFORMATION

Corresponding Authors*E-mail address: [email protected] (R.-C. Zeng). Tel. +86-532-80681226. Fax: +86-532-86057920.*E-mail address: [email protected] (B.-D. Zhao).

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

This work was supported by the National Natural ScienceFoundation of China (51571134), SDUST (ShandongUniversity of Science and Technology) Research Fund(2014TDJH104), Joint Innovative Centre for Safe and EffectiveMining Technology and Equipment of Coal Resources,Shandong Province.

■ REFERENCES(1) Zheng, Y. F.; Gu, X. N.; Witte, F. Biodegradable Metals. Mater.Sci. Eng., R 2014, 77, 1−34.(2) Witte, F. The History of Biodegradable Magnesium Implants: AReview. Acta Biomater. 2010, 6, 1680−1692.(3) Zeng, R. C.; Dietzel, W.; Witte, F.; Hort, N.; Blawert, C. Progressand Challenge for Magnesium Alloys as Biomaterials. Adv. Eng. Mater.2008, 10, B3−B14.(4) Staiger, M. P.; Pietak, A. M.; Huadmai, J.; Dias, G. Magnesiumand Its Alloys as Orthopedic Biomaterials: A Review. Biomaterials2006, 27, 1728−1734.(5) Baino, F. Biomaterials and Implants for Orbital Floor Repair. ActaBiomater. 2011, 7, 3248−3266.(6) Witte, F.; Feyerabend, F.; Maier, P.; Fischer, J.; Stormer, M.;Blawert, C. Biodegradable Magnesium-hydroxyapatite Metal MatrixComposites. Biomaterials 2007, 28, 2163−2174.(7) Witte, F.; Hort, N.; Vogt, C.; Cohen, S.; Kainer, K. U.; Willumeit,R.; Feyerabend, F. Degradable Biomaterials Based on MagnesiumCorrosion. Curr. Opin. Solid State Mater. Sci. 2008, 12, 63−72.(8) Witte, F.; Ulrich, H.; Rudert, M.; Willbold, E. BiodegradableMagnesium Scaffolds: Part 1: Appropriate Inflammatory Response. J.Biomed. Mater. Res., Part A 2007, 81A, 748−756.(9) Gu, X. N.; Zheng, Y. F.; Zhong, S. P.; Xi, T. F.; Wang, J. Q.;Wang, W. H. Biocorrosion and Cytotoxicity Evolution of Mg-Zn-CaBulk Metallic Glass. Biomaterials 2010, 31, 1093−1103.(10) Witte, F.; Kaese, V.; Haferkamp, H.; Switzer, E.; Meyer-Lindenberg, A.; Wirth, C. J.; Windhagen, H. In Vivo Corrosion of FourMagnesium Alloys and the Associated Bone Response. Biomaterials2005, 26, 3557−3563.

Figure 16. Schematic illustrations of degradation mechanism of the porous MAO/PLLA composite coatings on the Mg−1Li−1Ca alloys in HBSS:(a) the swelling of PLLA and corrosion of the substrate at the initial stage, (b) the blistering and final peeling-off of PLLA under the pressure ofhydrogen gas and corrosion products.



10026

mailto:[email protected]

mailto:[email protected]


(11) Zeng, R. C.; Guo, X.; Liu, C.; Cui, H. Z.; Tao, W.; Liu, Y.; Li, B.Study on Corrosion of Medical Mg-Ca and Mg-Li-Ca Alloys. ActaMetall. Sin. 2011, 47, 1477−1482.(12) Zeng, R. C.; Wang, L. D.; Zhang, F.; Cui, H. Z.; Han, E. H. InVitro Corrosion of Mg-6Zn-1Mn-4Sn-1.5Nd/0.5Y Alloys. Front.Mater. Sci. 2014, 8, 230−243.(13) Gu, X.; Zheng, Y.; Cheng, Y.; Zhong, S.; Xi, T. In VitroCorrosion and Biocompatibility of Binary Magnesium Alloys.Biomaterials 2009, 30, 484−498.(14) Zeng, R. C.; Sun, L.; Zheng, Y. F.; Cui, H. Z.; Han, E. H.Corrosion and Characterisation of Dual Phase Mg-Li-Ca alloy inHank’s Solution: The influence of Microstructural Features. Corros. Sci.2014, 79, 69−82.(15) Gu, X. N.; Li, S. S.; Li, X. M.; Fan, Y. B. Magnesium BasedDegradable Biomaterials: A Review. Front. Mater. Sci. 2014, 8, 200−218.(16) Zeng, R. C.; Lan, Z. D.; Kong, L.; Huang, Y.; Cui, H. Z.Characterization of Calcium-modified Zinc Phosphate ConversionCoatings and Their Influences on Corrosion Resistance of AZ31 Alloy.Surf. Coat. Technol. 2011, 205, 3347−3355.(17) Zeng, R. C.; Zhang, F.; Lan, Z. D.; Cui, H. Z.; Han, E. H.Corrosion Resistance of Calcium-modified Zinc Phosphate Con-version Coatings on Magnesium-aluminium Alloys. Corros. Sci. 2014,88, 452−459.(18) Gu, X. N.; Li, N.; Zhou, W. R.; Zheng, Y. F.; Zhao, X.; Cai, Q.Z.; Ruan, L. Corrosion Resistance and Surface Biocompatibility of aMicroarc Oxidation Coating on a Mg-Ca Alloy. Acta Biomater. 2011, 7,1880−1889.(19) Liang, J.; Zhang, R. H.; Peng, Z. J.; Liu, B. X. One-stepElectrochemical Fabrication of Bilayered MgO/Polymer Coating onMagnesium Alloy. Front. Mater. Sci. 2014, 8, 307−312.(20) Badar, M.; Lunsdorf, H.; Evertz, F.; Rahim, M. I.; Glasmacher,B.; Hauser, H.; Mueller, P. P. The Formation of An Organic Coat andthe Release of Corrosion Microparticles from Metallic MagnesiumImplants. Acta Biomater. 2013, 9, 7580−7589.(21) Abdal-hay, A.; Barakat, N. A. M.; Lim, J. K. Hydroxyapatite-doped Poly(lactic acid) Porous Film Coating for Enhanced Bioactivityand Corrosion Behavior of AZ31 Mg Alloy for OrthopedicApplications. Ceram. Int. 2013, 39, 183−195.(22) Wong, H. M.; Yeung, K. W. K.; Lam, K. O.; Tam, V.; Chu, P. K.;Luk, K. D. K.; Cheung, K. M. C. A Biodegradable Polymer-basedCoating to Control the Performance of Magnesium Alloy OrthopaedicImplants. Biomaterials 2010, 31, 2084−2096.(23) Zeng, R. C.; Liu, L. J.; Pang, T. T.; Zhang, F.; Zhang, W. W.; Li,S. Q.; Cui, H. Z.; Han, E. H. Corrosion Resistance of Silane-modifiedHydroxide Zinc Carbonate Film on AZ31 Magnesium Alloy. ActaMetall. Sin. (Engl. Lett.) 2015, 28, 373−380.(24) Zeng, R. C.; Liu, L. J.; Li, S. Q.; Zou, Y. H.; Zhang, F.; Yang, Y.;Cui, H.; Han, E. H. Self-assembled Silane Film and SilverNanoparticles Coating on Magnesium Alloys for Corrosion Resistanceand Antibacterial Applications. Acta Metall. Sin. (Engl. Lett.) 2013, 26,681−686.(25) Liu, X.; Yue, Z.; Romeo, T.; Weber, J.; Scheuermann, T.;Moulton, S.; Wallace, G. Biofunctionalized Anti-corrosive SilaneCoatings for Magnesium Alloys. Acta Biomater. 2013, 9, 8671−8677.(26) Zeng, R. C.; Liu, Z. G.; Zhang, F.; Li, S. Q.; Cui, H. Z.; Han, E.H. Corrosion of Molybdate Intercalated Hydrotalcite Coating onAZ31 Mg Alloy. J. Mater. Chem. A 2014, 2, 13049−13057.(27) Zhang, F.; Liu, Z. G.; Zeng, R. C.; Li, S. Q.; Cui, H. Z.; Song, L.;Han, E. H. Corrosion Resistance of Mg-Al-LDH Coating onMagnesium Alloy AZ31. Surf. Coat. Technol. 2014, 258, 1152−1158.(28) Ding, Y. F.; Wen, C.; Hodgson, P.; Li, Y. C. Effects of AlloyingElements on the Corrosion Behavior and Biocompatibility ofBiodegradable Magnesium Alloys: A Review. J. Mater. Chem. B2014, 2, 1912−1933.(29) Vogt, C.; Bechstein, K.; Gruhl, S.; Lange, M.; Druecker, H.;Witte, F.; Vogt, J. Investigation of the Degradation of BiodegradableMg Implant Alloys in Vitro and in Vivo by Analytical Methods.

Proceeding of the 8th International Conference on Magnesium Alloys andTheir Applications; Wiley-VCH: Weinheim, 2010; pp 1162−1174.(30) Sankara Narayanan, T. S. N.; Park, I. S.; Lee, M. H. Strategies toImprove the Corrosion Resistance of Microarc Oxidation (MAO)Coated Magnesium Alloys for Degradable Implants: Prospects andChallenges. Prog. Mater. Sci. 2014, 60, 1−71.(31) Gao, Y.; Yerokhin, A.; Matthews, A. DC Plasma ElectrolyticOxidation of Biodegradable Cp-Mg: In-vitro Corrosion Studies. Surf.Coat. Technol. 2013, 234, 132−142.(32) Nemcova, A.; Skeldon, P.; Thompson, G. E.; Pacal, B. Effect ofFluoride on Plasma Electrolytic Oxidation of AZ61 Magnesium Alloy.Surf. Coat. Technol. 2013, 232, 827−838.(33) Zhang, R. F.; Xiong, G. Y.; Hu, C. Y. Comparison of CoatingProperties Obtained by MAO on Magnesium Alloys in Silicate andPhytic Acid Electrolytes. Curr. Appl. Phys. 2010, 10, 255−259.(34) Zhang, R. F.; Zhang, S. F.; Duo, S. W. Influence of Phytic AcidConcentration on Coating Properties Obtained by MAO Treatmenton Magnesium Alloys. Appl. Surf. Sci. 2009, 255, 7893−7897.(35) Fischerauer, S. F.; Kraus, T.; Wu, X.; Tangl, S.; Sorantin, E.;Hanzi, A. C.; Loffler, J. F.; Uggowitzer, P. J.; Weinberg, A. M. In VivoDegradation Performance of Micro-arc Oxidized Magnesium Implants:A Micro-CT Study in Rats. Acta Biomater. 2012, 9, 5411−5420.(36) Han, X. G.; Zhu, F.; Zhu, X. P.; Lei, M. K.; Xu, J. J.Electrochemical Corrosion Behavior of Modified MAO Film onMagnesium Alloy AZ31 Irradiated by High-intensity Pulsed Ion Beam.Surf. Coat. Technol. 2013, 228, S164−S170.(37) Liu, G. Y.; Hu, J.; Ding, Z. K.; Wang, C. Bioactive CalciumPhosphate Coating Formed on Micro-arc Oxidized Magnesium byChemical Deposition. Appl. Surf. Sci. 2011, 257, 2051−2057.(38) Ivanou, D. K.; Starykevich, M.; Lisenkov, A. D.; Zheludkevich,M. L.; Xue, H. B.; Lamaka, S. V.; Ferreira, M. G. S. Plasma AnodizedZE41 Magnesium Alloy Sealed with Hybrid Epoxy-silane Coating.Corros. Sci. 2013, 73, 300−308.(39) Garlotta, D. A. Literature Review of Poly (lactic acid). J. Polym.Environ. 2001, 9, 63−84.(40) Anderson, J. M.; Shive, M. S. Biodegradation and Biocompat-ibility of PLA and PLGA Microspheres. Adv. Drug Delivery Rev. 1997,28, 5−24.(41) Zhou, H.; Lawrence, J. G.; Bhaduri, S. B. Fabrication Aspects ofPLA-CaP/PLGA-CaP Composites for Orthopedic Applications: AReview. Acta Biomater. 2012, 8, 1999−2016.(42) Bergsma, J. E.; De Bruijn, W. C.; Rozema, F. R.; Bos, R. R.;Boering, G. Late Degradation Tissue Response to Poly (L-lactide)Bone Plates and Screws. Biomaterials 1995, 16, 25−31.(43) Lu, L.; Garcia, C. A.; Mikos, A. G. In Vitro Degradation of ThinPoly (DL-lactic-co-glycolic acid) Films. J. Biomed. Mater. Res. 1999, 46,236−244.(44) Van der Elst, M.; Klein, C. P.; De Blieck-Hogervorst, J. M.;Patka, P.; Haarman, H. J. Bone Tissue Response to BiodegradablePolymers Used for Intramedullary Fracture Fixation: A Long-term inVivo Study in Sheep Femora. Biomaterials 1999, 20, 121−128.(45) Abdal-hay, A.; Barakat, N. A. M.; Lim, J. K. Influence ofElectrospinning and Dip-coating Techniques on the Degradation andCytocompatibility of Mg-based Alloy. Colloids Surf., A 2013, 420, 37−45.(46) Xu, L.; Yamamoto, A. In Vitro Degradation of BiodegradablePolymer-coated Magnesium under Cell Culture Condition. Appl. Surf.Sci. 2012, 258, 6353−6358.(47) Xu, L.; Yamamoto, A. Characteristics and Cytocompatibility ofBiodegradable Polymer Film on Magnesium by Spin Coating. ColloidsSurf., B 2012, 93, 67−74.(48) Chen, Y.; Song, Y.; Zhang, S.; Li, J.; Zhao, C.; Zhang, X.Interaction Between A High Purity Magnesium Surface and PCL andPLA Coatings During Dynamic Degradation. Biomed. Mater. 2011, 6,025005.(49) Guo, M.; Cao, L.; Lu, P.; Liu, Y.; Xu, X. Anticorrosion andCytocompatibility Behavior of MAO/PLLA Modified MagnesiumAlloy WE42. J. Mater. Sci.: Mater. Med. 2011, 22, 1735−1740.



10027


(50) Lu, P.; Cao, L.; Liu, Y.; Xu, X.; Wu, X. Evaluation of MagnesiumIons Release, Biocorrosion, and Hemocompatibility of MAO/PLLA-modified Magnesium Alloy WE42. J. Biomed. Mater. Res., Part B 2011,96B, 101−109.(51) Zeng, R. C.; Qi, W. C.; Song, Y. W.; He, Q. K.; Cui, H. Z.; Han,E. H. In Vitro Degradation of MAO/PLA Coating on Mg-1.21Li-1.12Ca-1.0Y Alloy. Front. Mater. Sci. 2014, 8, 343−353.(52) Wei, Z. L.; Tian, P.; Liu, X. Y.; Zhou, B. X. In Vitro Degradation,Hemolysis, and Cytocompatibility of PEO/PLLA Composite Coatingon Biodegradable AZ31 Alloy. J. Biomed. Mater. Res., Part B 2015, 103,342−354.(53) Song, G. L.; Atrens, A. Corrosion Mechanisms of MagnesiumAlloys. Adv. Eng. Mater. 1999, 1, 11−33.(54) Song, Y.; Han, E. H.; Shan, D.; Yim, C. D.; You, B. S. The Effectof Zn Concentration on the Corrosion Behavior of Mg-xZn Alloys.Corros. Sci. 2012, 65, 322−330.(55) Zhang, J.; Dai, C.; Wei, J.; Wen, Z.; Zhang, S.; Chen, C.Degradable Behavior and Bioactivity of Micro-arc Oxidized AZ91DMg Alloy with Calcium Phosphate/Chitosan Composite Coating inM-SBF. Colloids Surf., B 2013, 111, 179−187.(56) Shu, Y.; Ou, G. M.; Wang, L.; Zou, J. C.; Li, Q. L. SurfaceModification of Titanium with Heparin-chitosan Multilayers via Layer-by-layer Self-assembly Technique. J. Nanomater. 2011, 2011, 1−8.(57) Takagishi, Y.; Kawakami, T.; Hara, Y.; Shinkai, M.; Takezawa,T.; Nagamune, T. Bonelike Tissue Formation by Three-dimensionalCulture of MG63 Osteosarcoma Cells in Gelatin Hydrogels UsingCalcium-enriched Medium. Tissue Eng. 2006, 12, 927−937.(58) Gao, J. C.; Qiao, L. Y.; Li, L. C.; Wang, Y. Hemolysis Effect andCalcium-phosphate Precipitation of Heat-organic-film Treated Mag-nesium. Trans. Nonferrous Met. Soc. China 2006, 16, 539−544.(59) Tokiwa, Y.; Calabia, B. P. Biodegradability and Biodegradationof Poly (lactide). Appl. Microbiol. Biotechnol. 2006, 72, 244−251.(60) de Jong, S. J.; Arias, E. R.; Rijkers, D. T. S.; Van Nostrum, C. F.;Kettenes-van den Bosch, J. J.; Hennink, W. E. New Insights into theHydrolytic Degradation of Poly(lactic acid): Participation of theAlcohol Terminus. Polymer 2001, 42, 2795−2802.(61) Van Nostrum, C. F.; Veldhuis, T. F. J.; Bos, G. W.; Hennink, W.E. Hydrolytic Degradation of Oligo(lactic acid): A Kinetic andMechanistic Study. Polymer 2004, 45, 6779−6787.(62) Burkersroda, F.; Schedl, L.; Gopferich, A. Why DegradablePolymers Undergo Surface Erosion or Bulk Erosion. Biomaterials2002, 23, 4221−4231.(63) Piemonte, V.; Gironi, F. Kinetics of Hydrolytic Degradation ofPLA. J. Polym. Environ. 2013, 21, 313−318.(64) Zeng, R. C.; Zhang, J.; Huang, W. J.; Dietzel, W.; Kainer, K. U.;Blawert, C.; Ke, W. Review of Studies on Corrosion of MagnesiumAlloys. Trans. Nonferrous Met. Soc. China 2006, 16, s763−s771.(65) Yuan, X.; Mak, A. F. T.; Kwok, K. W.; Yung, B. K. O.; Yao, K.Characterization of Poly (L-lactic acid) Fibers Produced by MeltSpinning. J. Appl. Polym. Sci. 2001, 81, 251−260.(66) Tomihata, K.; Suzuki, M.; Ikada, Y. The pH Dependence ofMonofilament Sutures on Hydrolytic Degradation. J. Biomed. Mater.Res. 2001, 58, 511−518.



10028


Documents

In Vitro Corrosion and Cytocompatibility of a Microarc