Emmanuel Mena-Morcillo ,Lucien Veleva,? ,Mariana Cerda-Zorrilla ,Montserrat Soria-Castro ,Juan C.Castro-Alcántara ,Rosa C.Canul-Puc

1 Applied Physics Department,Center for Research and Advanced Study,(CINVESTAV-IPN),Carr.Ant.a Progreso Km.6,97310,Merida,Yucatan,Mexico

2 Marine Resources Department,Center for Research and Advanced Study,(CINVESTAV-IPN),Carr.Ant.a Progreso Km.6,97310,Merida,Yucatan,Mexico

Abstract This work presents a simple method to functionalise the surface of AZ31 magnesium alloy by applying a duplex MgF2/chitosan coating,which improves its corrosion resistance and provides it with some antibacterial performance.First,the effect of three chitosan solutions with different concentrations on the growth of the bacteria Klebsiella pneumoniae in nutritive medium(TSB)was evaluated by absorbance kinetics experiments,where the chitosan solution at 2%(m/V)was selected for the coating preparation.Before coating application,the AZ31 substrate was pretreated with hydrofluoric acid for 48 hours in order to form a MgF2 conversion layer.Subsequently,the coating was applied to the pretreated substrate through the dry-casting method.Samples of the alloy in each surface condition(bare,pretreated,and pretreated+coated with chitosan)were exposed to simulated body fluid(SBF)for 21 days at 37 °C,with the solution renewed every 24 hours and the wastes stored.The surfaces were characterised by SEM-EDS,and XPS after the immersion tests,whereas the stored solutions were employed to measure the change in the Mg-ions concentration.Electrochemical impedance spectroscopy and potentiodynamic polarisation were performed in each surface condition to compare their corrosion resistance in SBF.The antibacterial activity of the functionalised surfaces was evaluated by the plate counting method and compared with bare samples.All results were correlated and demonstrate that the modified surface of AZ31 achieved a higher corrosion resistance when it was exposed to SBF,as well as a reduction of the bacterial growth during in vitro tests.

Keywords: AZ31;Chitosan coating;Simulated body fluid;Corrosion;Antibacterial activity.

1.Introduction

Magnesium alloys display several features that make them suitable candidates for degradable implant materials,as their susceptibility to corrode in aqueous media,good biocompatibility,and mechanical properties [1–3].However,their high chemical reactivity and fast corrosion rates in physiological environments still hinder their clinical application [4,5].The main drawbacks about implanting Mg-based orthopaedical devices into a living body are the rapid decrease of their mechanical stability [6],and the intense hydrogen gas generation that is coupled to the Mg anodic dissolution process [7,8].Furthermore,Mg corrosion may cause an alkaline pH shift in the vicinity of the implant surface [8–10],thus diminishing the tissue-healing rate.One strategy to tailor the degradation rate of Mg alloys under relevant physiological conditions is by coating their surface [11–14],which could also enhance other properties,such as biocompatibility and antibacterial performance [15,16].Among antibacterial coatings applied to Mg alloys,those based on chitosan(CS)are an attractive option owing to their excellent biocompatibility and resorption of their degradation products [17–23],and also because they are non-antigenic,non-toxic,and bio-functional [24–26].Chitosan(C6H11NO4)nis a cationic polysaccharide derived from the N-deacetylation of chitin that presents amino groups at the C-2 positions [27,28].These functional groups make CS reactive and thus advantageous to adhere well on either Mg alloys [17,22,23],Mg composites [19,20],or micro-arc oxidation layers [18,21].

The preparation of CS-based coatings usually requires the CS to be dissolved in acetic acid solution.Therefore,it is necessary to apply a pretreatment on the Mg alloy surface to avoid inducing strong corrosion during the coating process itself and achieve a uniform coating.Fluoride-based acid pickling has been applied on Mg to form a conversion layer containing MgF2on the surface [29–34].Such fluoride coatings display excellent biocompatibility [35–37],as well as enhanced cell adhesion and proliferation [33,38,39].However,this conversion layer alone does not provide a prolonged protective effect on Mg alloys [15,23].

Although there are some studies in the literature that focus on applying either MgF2[30–34]or chitosan as a coating on Mg alloys [17–22]and even comparing the difference between them [23],the development of a duplex MgF2/CS coating has not been explored before.The aim of this work was to develop a simple method to apply a chitosan(CS)coating on a pretreated AZ31 Mg-alloy surface,characterise its corrosion behaviour in simulated body fluid(SBF),and evaluate its effect on the bacterial growth ofKlebsiella pneumoniae.

2.Materials and methods

2.1.Reagents

Hydrofluoric acid(HF,48-51% ACS grade,Meyer,Mexico),acetic acid(CH3CO2H,99.8% purity,CTR Scientific,Mexico),and chitosan(powder,110 kDa,ALZOR,Mexico)were employed for the surface modification of AZ31.Simulated body fluid was prepared as described in the literature[40],with ultrapure deionised water(18.2 MΩ· cm),and the following reagents in analytical grade:NaCl,NaHCO3,Na2CO3,KCl,K2HPO4· 3H2O,MgCl2· 6H2O,HEPES,CaCl2,Na2SO4,NaOH.

2.2.CS solutions preparation and antibacterial susceptibility experiments

Chitosan(CS)solutions were prepared with three different concentrations(m/V):2%,3% and 4%.In brief,CS was stirred into an acetic acid solution(0.5% V/V)until it was completely dissolved.The effect of CS on bacterial growth was examined by means of absorbance kinetics performed with a UV-Vis spectrophotometer(UV-1800,Shimadzu Inc.,Japan).The bacteria selected for this study was the GramnegativeK.pneumoniaeowing to its significance in the clinical setting because it is related to surgical wound infections[41,42].The bacteria were cultured in 10 mL of nutrient broth(TSB,Becton-Dickinson,Sparks,MD,USA),adjusted to 5 McFarland(5×108CFU mL?1),and further placed in sterile test tubes as follows:one was kept as control,and the other three were combined with the CS solution in each concentration(performed in triplicate).Samples were further incubated at 35 °C,and the absorbance values were measured at 0,1,2,3,and 24 h.

2.3.Evaluation of the effect of CS on the morphology of bacteria

The cellular damage ofK.pneumoniaecaused by the presence of CS was assessed by employing a field-emission scanning electron microscope coupled with energy dispersive X-ray spectrometry(FESEM-EDS,Jeol JSM-7600F,Japan).The SEM images were acquired after the absorbance kinetics experiments of a control sample(only bacteria)and another sample with bacteria+2%(m/V)of CS.The sample preparation for SEM observations is briefly described below.First,the bacteria were chemically fixed with glutaraldehyde at 2.5% for 60 minutes,then bacterial cells were dehydrated with increasing concentrations of ethanol(30%,50%,70%,90%,and 100%).The bacteria cells were dried at the critical point(31 °C and 110 Psi),using liquid carbon dioxide(CO2)to maintain the typical morphology of the bacteria.Finally,the bacteria were covered by cathodic spraying(sputtering)of a thin gold-palladium film for 40 s with a Quorum Q150RES sputter coater(Quorum Technologies,UK),and stored in a desiccator pending SEM observations.

2.4.Measurement of ζ-potential

Theζ-potential value of CS solution at 2%(m/V)was measured by means of the dynamic light scattering(DLS)method,which was carried out with the instrument Zetasizer Nano ZS(Malvern Instruments,Malvern,UK).The data was recorded at 21±5 °C andλ=633 nm.To ensure the accuracy of the value,the measurements were repeated 5 times.

2.5.Preparation of AZ31 substrates

The AZ31 magnesium alloy(composition in wt.%:Al 3%;Zn 1%,Mn 0.2%;and the balance Mg;Alfa Aesar,Ward Hill,MA,USA)was cut into rectangular specimens(1.5 cm x 3.0 cm x 0.2 cm),which were mechanically sanded with SiC papers from 400 to 1200 grit,polished with an Al2O3(0.3 μm particle size)suspension,rinsed and sonicated in ethanol,and dried in warm air.The AZ31 specimens were immersed in HF acid(48-51%)for 48 h prior to the chitosan coating application.

2.6.CS coating application on AZ31 substrates

The CS solution(2% m/V)was homogeneously applied to the pretreated surface of AZ31 through the dry-casting method.This method is one of the major phase inversion techniques in which a homogeneous polymer solution consisting of solvent(s)and nonsolvent(s)is cast on a support and then evaporation of the casting solution takes place under convective conditions [43].The CS coating was prepared by evaporating the remaining solvent(acetic acid)at a constant temperature of 45 °C.Subsequently,the CS film was neutralised with an alkaline solution(pH=9),and the resulting coating was rinsed until it reached a neutral pH.

2.7.SEM-EDS analysis of the modified surfaces

The surface of AZ31 was evaluated by SEM-EDS in three different conditions:bare,pretreated with HF acid(HFPT),and pretreated+chitosan-coated(CS-CT).In addition,an SEM image of the cross-section of a CS-CT specimen(mechanically sawed)was acquired to measure the thickness of each layer.A stylus profiler(Alpha Step D120,KLA-Tencor,Milpitas,CA,USA)was employed in a similar CS-CT specimen to confirm the total thickness of the surface modification.

2.8.Electrochemical characterisation

The surface of AZ31 in each condition(bare,HF-PT,and CS-CT)was studied through electrochemical impedance spectroscopy(EIS)and potentiodynamic polarisation(PDP)after 2 h of immersion in SBF.The experiments were carried out with an Interface-1000E potentiostat(Gamry Instruments,Inc.,Philadelphia,PA,USA)in a three-electrode cell configuration:the AZ31 specimens as working electrodes(1 cm2exposed),a platinum mesh(Alfa Aesar,Ward Hill,MA,USA)as a counter electrode,and a saturated Ag/AgCl/KCl reference electrode(CH Instruments Inc.,Austin,TX,USA).The EIS spectra were acquired employing a perturbation amplitude of±10 mV(vs OCP),and a frequency interval from 104Hz to 10?2Hz.The PDP experiments were performed scanning from -0.3 V to +0.5 V(vsOCP)at a scan rate of 1 mV s?1.The obtained electrochemical data from EIS and PDP were analysed with Gamry Echem Analyst software(Gamry Instruments,Inc.).

2.9.Immersion tests and analysis of exposed surfaces

Specimens of AZ31 in the three different conditions(bare,HF-PT,and CS-CT)were exposed to SBF at 37 °C for 21 days,maintaining a volume of 10 mL per each cm2of the exposed area as recommended by ISO 16428 standard [44].Besides,a portion of the electrolyte was renewed daily to mimic the 46% urinary excretion [45].The waste solution of each specimen was stored in separate containers to measure the final concentration of the Mg-ion,using photometry(HI83200,Hanna Instruments,Woonsocket,RI,USA).

SEM-EDS and X-ray photoelectron spectroscopy(XPS,KAlpha,Thermo Scientific,Waltham,MA,USA)were used to characterize the surface on each condition after immersion tests.

2.10.Antibacterial activity test of the CS coated specimens

The bacterial growth on the CS-CT(2% m/V)specimens was evaluated through the plate counting method.Coated and bare substrates were sterilised by UV light and placed in sterile Petri dishes prior to the application of 100 μL of medium culture withK.pneumoniae(5×105CFU/mL)on their surface.These specimens were incubated for 1 h at 37 °C,and their surface was washed with a saline solution to collect the grown bacteria.The collected suspensions were diluted four times and further placed in agar-agar(TSA,Becton-Dickinson,Sparks,MD,USA)plates,which were incubated for 24 h at 37°C.At the end of incubation,a light microscope was employed to count colonies formed on each plate,and thus verify the antibacterial effect of CS onK.pneumoniae.

3.Results and discussion

3.1.Absorbance kinetics

Fig.1 shows the absorbance kinetics values obtained at 0,1,2,3,and 24 h of incubation of theK.pneumoniaebacteria at 37 °C.For this purpose,mixtures of 10 mL of nutritive medium(TSB)+1 mL of each of the CS solutions(at 2%,3%,and 4% m/V)were employed.As the bacterial cell concentration increases,the culture becomes turbid and the amount of transmitted light reached by the photoelectric cell is reduced.

Fig.1.Absorbance values against the growth time of K.pneumoniae bacteria(control)and the bacteria with the addition of chitosan at different concentrations(2%,3%,and 4% m/V).

Fig.2.SEM images of K.pneumoniae bacteria incubated for 24 h at 37 °C:(A)control;(B)with the addition of 2% chitosan(m/V).

The values of absorbance(related to bacterial growth)decrease when CS concentration increases during the first 3 h,but they all reach a stability value of about 0.76 after 24 h,while for the control(only bacteria),the absorbance value was approximately 1.93.These results indicate that the presence of CS reduces the growth of the bacteria approximately 2.5 times(60% less)after 24 h,regardless of the concentration of chitosan used.The main reason the reduction of bacterial growth was not influenced by the chitosan concentrations is that chitosan’s activity is mostly growth-inhibitory [46].Therefore,the chitosan solution at 2%(m/V)was selected for the coating preparation,in order to employ the lowest amount of reagent and optimise the procedure.

3.2.Effect of CS on the morphology of K.pneumoniae bacteria observed by SEM

Fig.2 shows the bacteria morphology after the absorbance kinetics experiments.

Bacterial cells of the control sample display the typical structure ofK.pneumoniae,in the form of a bacillus(or rod).In the case of bacteria exposed to CS,it can be noted that there were changes in their morphology,exhibiting a rough surface with discrete protrusions.Besides,they showed an increase in their size(both transverse and longitudinal),probably generated by the cellular stress caused by the CS.

3.3.ζ-potential values

Theζ-potential is a function of the charged surface of a particle,in the medium in which the particle is suspended.The obtained value of theζ-potential of the 2%(m/V)CS solution was 25.6±1.62 mV,whereas the reported value forK.pneumoniaebacterium is -24.1±0.96 mV [47].According to both values of theζ-potential,it may be suggested that the mechanism of chitosan action on the bacteria is directly related to electrostatic effects:the positive potential of CS acts with the negative potential of the membrane of the bacterium and exerts mechanical stress on it,resulting in the depolarisation of the membrane and its morphological transformation.In fact,it is generally assumed that the polycationic nature of chitosan,might be a fundamental factor contributing to its interaction with negatively charged surface components of many fungi and bacteria,causing extensive cell surface alterations,leakage of intracellular substances,and ultimately resulting in impairment of vital bacterial activities [48].

Fig.3.Diagram of the surface modification process carried out in this work.

3.4.Analysis of AZ31 surface modification

The surface modification process done in this work is summarized in Fig.3.First,the bare surface of AZ31 was mechanically abraded and polished to a mirror-like finish.After that,the surface was treated in HF to form a MgF2layer.Finally,the CS was applied onto this first layer to form a duplex MgF2/CS coating.

Fig.4 presents the morphology of AZ31 alloy at different stages of its surface modification.The bare surface(Fig.4A)displayed a microstructure corresponding to theα-matrix of the alloy(Site 1),and white intermetallic particles of different size(Site 2).The EDS quantification results(Table 1)confirmed that the composition of Site 1 corresponds to the AZ31 matrix owing to the percentages of Mg,Al,and Zn elements.The values of Al and Mn of the white particles(Site 2)allowed their correlation with Al-Mn intermetallic particles(Al8Mn5),as reported elsewhere [49].

Table 1 EDS elemental composition(wt.%)of AZ31 surface in different conditions.

Fig.4.SEM images of AZ31 surface in different conditions:(A)bare;(B)HF-PT;and(C)CS-CT.

After the pretreatment in HF,the morphology of AZ31(Fig.4B)changed by presenting considerable roughness,in the absence of intermetallic particles,which is related to the formation of a conversion layer on the surface.The composition of this layer(Site 3)was associated with the MgF2phase,owing to the acquired results by EDS analysis.The SEM image presented in Fig.4C shows that the pretreated surface is completely covered by a homogeneous and smooth film with many nanospheres(around 80 nm diameter)distributed over its surface.Some of these nanospheres formed linear-shaped colonies(agglomerations)in certain regions.The elemental quantification of both the film and nanospheres(Site 4)coincides with the expected composition of chitosan(C6H11NO4)nwith high percentages of C,O,and N(Table 1).Chitosan nanoparticles are usually obtained by ionotropic gelation,which occurs due to the inter/intramolecular cross-linking of polycationic chitosan by an anionic cross-linker [50,51].Therefore,it may be suggested that the formation of the nanospheres could be promoted by fluoride anions arising from the MgF2conversion layer.However,the exact mechanism of formations of these CS nanospheres is still unclear,so additional work should be done on this topic.

The cross-section of a CS-CT specimen is shown in Fig.5A,and the corresponding EDS analysis of each layer is displayed in Table 2.It can be observed that the pretreatment layer is about 5 times thinner than the CS coating layer.

The EDS quantification(Table 2)confirmed the composition of each layer,given the weight percentage of the relevant elements.The percentage of Mg,Al,and Zn elements agrees with the AZ31 matrix composition.In the case of the pretreatment layer,Mg and F percentages were related to the MgF2film formed after the HF acid pickling.The C,O,and N percentages of the CS coating layer correspond to the chitosan compound.The presence of additional elements was attributed to the disturbance caused by the sawing process of the specimens.

Fig.5.(A)SEM image of the cross-section of a CS-CT sample,and(B)profilometry of a similar CS-CT sample(blue line)compared with the thickness value measured in A(dashed line).

Fig.6.EIS Nyquist plots of the surface of AZ31 in its different conditions:(A)bare,(B)HF-PT,and(C)CS-CT,after their immersion in SBF for 2 h.(D)Equivalent electric circuit employed to fit the EIS data.

Table 2 EDS analysis(wt.%)of the layers in the cross-section(Fig.4A)of the modified AZ31 surface.

It can also be noted that the total thickness of the surface modification(pretreatment layer and CS coating)was approximately 9 μm.Fig.5B shows the profilometry of a similar CS-CT specimen,which matches with the total thickness measured by SEM.

3.5.Electrochemical tests results

Fig.6 shows the EIS Nyquist diagrams obtained from AZ31 specimens in different surface conditions.It can be noted that the Nyquist diagrams present capacitive loops,which are usually related to the charge transfer,film effects,or mass transport [52].The proposed equivalent circuit was based in literature [53–58],whereRsis associated with the resistance of the solution,Rporwith the resistance caused by different thicknesses at different places(or pores)in the surface layer,andRctwith the charge transfer resistance.Constant phase elements(CPE)were employed instead of capacitors,owing to dispersion effects that may be produced by the microscopic roughness of the studied surface [59].TheCPEdlelement represents the capacitance formed by the double layer at the surface/electrolyte interface.On the other hand,theCPEcoatelement is related to the capacitance of either the corrosion product film formed in the bare surface,the MgF2conversion layer in the HF-PT surface,or the chitosan coating in the CS-CT surface.Table 3 presents the fitting parameters extracted from the EIS data in Fig.6.The polarisation resistance(Rp)value was obtained by calculating the equivalent resistance of the proposed electrical circuit(Rp=Rpor+Rct).It can be seen that the polarisation resistance(Rp)values show an increasing trend for each step of the surface modification,revealing a significant difference of 3 orders of magnitude between the bare surface of AZ31 and the CS-CT surface.

Table 3 Fitting parameters of the equivalent electrical circuits obtained from the EIS spectra in Fig.5.

Fig.7 presents the potentiodynamic polarisation(PDP)curves acquired for the bare,HF-PT,and CS-CT surfaces,after their immersion in SBF for 2 h.The corrosion mechanism for AZ31 at initial stages is governed by the anodicdissolution of the alloy,which is balanced by hydrogen evolution in the cathodic areas [60,61].

Fig.7.Potentiodynamic polarisation(PDP)curves of AZ31 surface in different conditions:bare,HF-PT,and CS-CT,after their immersion in SBF for 2 h.

According to the PDP curves,cathodic branches display activation-controlled kinetics,whereas anodic branches tend to reach a limiting-current value related to a barrier effect of the alloy surface.Corrosion product formation occurs when Mg2+ions migrate towards the cathodic areas,while anions(Cl?,OH?)do it towards dissolution sites.Hence,Mg(OH)2is usually formed close to the anodic sites [60,61].

Further,a sharp increase in the anodic current is observed,which is associated with the dissolution of the oxide/conversion layer or dissolution of the coating at more anodic potentials.The above is attributed to the chloride ions in SBF,which transform the insoluble Mg(OH)2into highly soluble MgCl2[62].

The Tafel extrapolation method was applied to extract the cathodic(βc)and anodic(βa)Tafel slopes from PDP data,as well as the corrosion potential(Ecorr).On the other hand,the corrosion current density was calculated by employing the Stern-Geary relationship Eq.(1),as stated in the ASTM G102-89 standard [63]:

whereRpwas taken from EIS results,and the constantBwas calculated from the Tafel slopes,as follows in Eq.(2)[63]:

Furthermore,the corrosion rate(CR)and mass loss rate(MR)were calculated through Eqs.(3)and(4)[63]:

whereK1andK2are constants(K1=3.27×10?3mm g μA?1cm?1yr?1;K2=8.954×10?3g cm2μA?1m?2d?1),ρis the density of AZ31(≈1.74 g cm?3),EWis the equivalent weight of AZ31(≈12.2),andjcorris the corrosion current density calculated from Eq.(1).

Table 4 shows the corrosion parameters obtained from electrochemical measurements.The corrosion rate of the bare surface is within the range of values found in the literature for the same AZ31/SBF system at short exposure times[64–68].Results indicate that the degradation rate of bare AZ31 decreased by 3 orders of magnitude when its surface was modified.Similarly,Ecorrshifted by about 150 mV to a less negative value,confirming the lower tendency to corrosion of the CS-CT surface.According to theMRvalues,the increase in corrosion resistance was mainly attributed to the MgF2layer formed during the pretreatment.This conversion layer acted as a physical barrier to avoid direct contact of the AZ31 surface with SBF electrolyte.The application of chitosan onto the MgF2layer further diminished the degradation of the alloy by enhancing the barrier effect and forming a more stable duplex coating.

Table 4 Corrosion parameters of each surface condition,obtained from electrochemical measurements.

Table 5 EDS analysis(wt.%)of the corrosion products layers displayed in Fig.4.

Fig.8.SEM images of samples after their exposure to SBF for 21 days at 37 °C:(A)bare;(B)HF-PT;and(C)CS-CT.

3.6.Analysis of the surface layers after immersion tests

Fig.8 presents SEM images of the surface morphology of AZ31,in its different conditions(bare,HF-PT,and CS-CT),after 21 days of immersion in SBF solution at 37 °C.

Table 5 shows the EDS analysis of the areas presented in the said SEM images.The bare AZ31 alloy surface(Fig.8A)was damaged and presented a cracked corrosion products layer,which contains significant percentages of O(44.5%),Ca(21.8%),and P(17.4%).On the other hand,the surface of the HF-PT specimen(Fig.8B)did not present cracks,and its corrosion products layer mainly contained Mg(34.2%),F(31.3%),C(18.5%),and O(7.6%).Meanwhile,the morphology of the CS-CT specimen(Fig.8C)changed,losing the chitosan nanospheres,but it was still homogeneous and presented considerable percentages of C(57.4%),O(23.4%),N(9.6%),followed by Ca(4.0%)and P(3.0%)elements,both as a part of the SBF solution.

In order to correlate the elemental quantification analysis(Table 5)with the composition of the phases formed on each AZ31 magnesium surface,XPS analysis was carried out.The high-resolution spectra of Mg,Ca,P,C,N,O,and F elements are displayed in Fig.9.

Table 6 Increase of Mg2+ concentration after exposing the bare and modified surfaces of AZ31 to SBF solution for 21 days at 37 °C(initial [Mg2+]in SBF is 30 mg L?1)

The spectrum for Mg showed one peak centered:at 50.8 eV for the bare surface,which was associated with magnesium hydroxide [69];and at 51.8 eV for the HF-PT surface,related to MgCO3phase [69].The spectrum for Ca split intotwo peaks centered at 347.8 eV and 351.4 eV as a result of spin-orbit splitting,whereas the spectrum for P exhibit one peak at 133.9 eV originated from bonding between(PO4)3?and Ca in the form of calcium phosphates [62].The peak of the spectrum for C displayed three components:at 284.8 eV,characteristic of carbon only bound to carbon and hydrogen[C–(C,H)][70];near 286.6 eV,typical of carbon making a single bond with oxygen or nitrogen [C–(O,N)][70];at 288.5 eV typical of acetal and amide groups [O–C–O,N–C=O][70].The spectrum for N showed one peak at 399.4 eV,which was related to the non-protonated amine or amide[70,71].The peak of the spectrum of O exhibited two components:one at 532.8 eV,related to the polysaccharide backbone(C6H10O5),and the second component at 531.0 eV may be associated with amide or acetal groups [70].The spectrum of F exhibited a peak position at 685.4 eV,which has been attributed to the MgF2phase [72].

Therefore,according to both EDS and XPS results,the layer formed on the bare AZ31 alloy(Fig.8A)contains a mixture of MgO/Mg(OH)2[68],and calcium phosphates,such as Ca10(PO4)6(OH)2,Ca3(PO4)2,or Ca3Mg3(PO4)4as previously reported [73–75].

On the HF-PT surface(Fig.8B),the phases detected were MgF2and MgCO3.The CS-CT surface(Fig.8C)was still composed of chitosan(C6H11NO4)n,but with the presence of the calcium phosphate phases at a low content.

3.7.Mg-ion concentration after immersion tests

Table 6 presents the change in Mg-ion concentration,measured from the stored solutions at the end of the immersion tests.The increase in [Mg2+]is related to both magnesium oxidation,and the dissolution of magnesium hydroxide caused by Cl?ions [76].Results show that the increase in Mg-ion concentration was higher for the bare AZ31 magnesium alloy than for both HF-PT and CS-CT specimens.These values are related to the degradation rate of the specimens.The?[Mg2+]was about 10 to 11 times lower for the CS-CT surface,compared with that of the bare alloy.These results are in good agreement with those obtained by electrochemical tests.

Fig.9.High-resolution XPS spectra of AZ31 in each surface condition:bare(red line),HF-PT(green line),and CS-CT(blue line)after the exposure to SBF for 21 days at 37 °C.

Fig.10.Comparison of bacterial growth on the bare substrate and the CS-CT one,evaluated by the plate-counting method.

3.8.Antibacterial activity assessed by the plate counting method

Results obtained by the plate counting method after 24 h of incubation at 37°C were as follows:the bare AZ31 surface displayed a bacterial concentration of about 2.034×106CFU mL?1,whereas the CS-CT surface of AZ31 showed a bacterial concentration of about 0.797×106CFU mL?1.Typical photos of colonization by the bacteria in the plate are presented in Fig.10.These results showed that chitosan reduces the growth of the bacteria by approximately 60% after 24 h of incubation at 37 °C,which agrees with the results obtained with the absorbance kinetics experiments(Fig.1).Thus,the modified surfaces are capable of inhibiting bacterial growth only through direct contact.

4.Conclusions

A simple method of functionalising the surface of AZ31 was developed by applying a duplex MgF2/CS coating,which enhanced the corrosion resistance and antibacterial performance of the alloy.Results from absorbance kinetics experiments and the plate counting method demonstrated that the growth ofKlebsiella pneumoniaebacteria decreased by about 60% when the coating was applied on the alloy.Furthermore,according to results obtained by SEM imaging andζ-potential measurements,this inhibition effect was attributed to cellular stress caused by the electrostatic interaction between the CS and the bacterial cell membrane,which may cause depolarisation on the membrane’s surface.

SEM analysis revealed that the CS coating layer was smooth,homogeneous,and presented many nanospheres with a diameter of about 80 nm.After its exposure to SBF for 21 days at 37 °C,the CS coated surface released/lost these nanospheres but maintained its homogeneity with the presence of calcium phosphates at a low content.Results from electrochemical tests and measurement of Mg-ions concentration indicated that the corrosion resistance of AZ31 was significantly improved after the surface modification.TheRpand mass loss rate(MR)values,calculated from EIS and PDP data,displayed a difference of 3 orders of magnitude between the bare and coated surface.Results from this work are expected to be useful for further research,including the evaluation in-depth of the formation mechanism of the CS nanospheres and their possible applications,as well as the application of the MgF2/CS coating onin vivoexperiments.

Acknowledgments

Emmanuel Mena-Morcillo gratefully thanks CONACYT for his scholarship as a Ph.D.student at CINVESTAV-IPN.The authors acknowledge LANNBIO-CINVESTAV for permitting the use of their facilities,as well as Dora A.Huerta-Quintanilla,Wilian J.Cauich-Ruiz,and Oswaldo J.Gónzalez-Yam for their technical assistance.