Effect of silver and strontium incorporation route on hydroxyapatite coatings elaborated by rf-SPS

OATAO is an open access repository that collects the work of Toulouse

researchers and makes it freely available over the web where possible

Any correspondence concerning this service should be sent

to the repository administrator:

tech-oatao@listes-diff.inp-toulouse.fr

This is an author’s version published in: http://oatao.univ-toulouse.fr/27198

To cite this version:

Chambard, Marine and Remache, Djamel

and Balcaen,

Yannick

and Dalverny, Olivier

and Alexis, Joël

and Siadous,

Robin and Bareille, Reine and Catros, Sylvain and Fort, Pascal and

Grossin, David and Gitzhofer, François and Bertrand, Ghislaine

Effect of silver and strontium incorporation route on hydroxyapatite

coatings elaborated by rf-SPS. (2020) Materialia, 12. 1-12. ISSN

25891529

Official URL:

https://doi.org/10.1016/j.mtla.2020.100809

Effect

of

silver

and

strontium

incorporation

route

on

hydroxyapatite

coatings

elaborated

by

rf-SPS

Marine

Chambard

_,

_Djamel

_Remache

_,

_Yannick

_Balcaen

_,

_Olivier

_Dalverny

_,

_Joël

_Alexis

_,

Robin

Siadous

_,

_Reine

_Bareille

_,

_Sylvain

_Catros

_,

_Pascal

_Fort

_,

_David

_Grossin

_,

François

Gitzhofer

_,

_Ghislaine

_Bertrand

a CIRIMAT, Université de Toulouse, CNRS, INP- ENSIACET 4 allée Emile Monso - BP44362, 31030 Toulouse cedex 4, France b CREPE, Department of Chemical and Biotechnological Engineering, Université de Sherbrooke, Sherbrooke, QC, Canada c _{LGP, Université de Toulouse, INP/ENIT, 47, avenue d’Azereix, Tarbes, F-65016, France}

d _{Inserm, U1026, Tissue Bioengineering, Bordeaux University, Bordeaux, France} e _{PROJECTION PLASMA SYSTEME (2PS), ZI du Colombier, 12220 Montbazens, France}

Keywords:

rf - Suspension plasma spraying doped hydroxyapatite coating mechanical properties biological properties bactericidal properties

a

b

s

t

r

a

c

t

Hydroxyapatitecoatingshavebeencurrentlyusedonhipprosthesesfortheirabilitytopromotefaster osseointe-grationandbonegrowth.Nevertheless,post-operativeinfectionsremainarecurringproblem.Toovercomethis issue,dopingwithantibacterialelementshasbecomeanewtrend.Inthiswork,hydroxyapatitecoatings elab-oratedbyradio-frequencysuspensionplasmaspraying(rf-SPS)weredopedwithsilverandstrontium.Several dopingstrategieswereexploredthankstotheversatilityofferedbySPScomparedwithconventionalspraying. Firstway:calciumphosphatedopedpowdersweresynthesizedbycoprecipitationandthendispersedintowater beforeplasmaspraying;secondway:undopedpowderwasdispersedintoaqueousmediuminwhichnitratesor nanoparticlesofthedopant(s)wererespectivelydissolved/dispersed.XRDrevealedahighlevelofcrystallinity ratio(ISO13779)andhydroxyapatiteproportionformostofthecoatings,withthepresenceofAg/Ag2O

nanopar-ticleswhateverthedopingroute.SEM-EDSandSTEMhavedemonstratedamorehomogeneousdistributionofthe strontiumwithinthecoatingmadefromthedopedpowder.Adherenceofthecoatingswasestimatedviaa3-point bendingtest,whilebacteriologicaltestswithS.aureusandproliferationofmesenchymalstemcells(hMSC)were performed.Theresultsindicatedapreferentialincorporationofstrontiumintothesecondaryphases,showed efficientbactericidalproperties,excellentmechanicalpropertiesincomparisonwithanAPSreferencecoating, andnoevidenceofcytotoxiceffect.Thisopensthewayofanewtypeofcoatingswithafinerstructureanda higherhomogeneitythroughabettercontrolofphysicochemicalpropertiesusingasuspensionastheprecursor.

1. Introduction

Theincreaseintheworld’spopulationanditsaginginducesan ever-increasingdemandfororthopedicprostheses[1].However,inabout1% ofthenewlyimplantedprostheses[2],aninfectioncausedbybacterial agentssuchasstaphylococciorenterobacteria istriggered, requiring newsurgery.Suchaninterventionisbothexpensiveandtraumatic. Cal-ciumphosphatecoatingshavebeensuccessfullyusedfor30yearsin or-thopaedicsurgerysincetheysuppressmicro-movements,supportbone in-growthandprovideosseostabilitytouncementedhipimplants[3,4]. Hydroxyapatiteisparticularlyusedascalciumphosphateasitis struc-turallyandchemicallysimilartobonemineral,andlessresorbablethan othercalciumphosphatephases[5]suchastricalciumphosphate(TCP), tetracalciumphosphate(TTCP),amorphouscalciumphosphate(ACP)or

∗_{Correspondingauthorat:CIRIMAT,Université deToulouse,CNRS,INP-ENSIACET4alléeEmileMonso-BP44362,31030Toulousecedex4,France.}

calciumoxide(CaO),whosepresenceislimitedeventhoughtypicalin plasma-sprayedcoating.

Thedevelopmentofhipprostheseswhichcanpreventbacterial infec-tionsisthereforeacurrentchallenge.Themainstrategywouldconsistin addingadopingelementwithantibacterialandanti-inflammatory prop-ertiessuchasmetalelements(Cu,Ag,Zn…)tohydroxyapatitethatis coatedontheimplant.Amongtheavailabledopants,silverisofgreat in-terestbecauseofitsreactivitywiththiolicmoieties(especiallyagainst gram-bacteria)[6]makingitoneofthemosteffectiveantibacterial agentknownatlowconcentration.Eventhoughfirstclinicalresults im-plyingsilverdopedhydroxyapatitecoatingarepromising[7],ahigh concentrationofsilvercanbetoxicandreportedlycausesargyria, hep-atopathy,andnephropathy[8,9].Albersetal.[10]conductedastudy inwhichtheydemonstratedthatthecytotoxicitythresholdofsilverfor osteoblastsandosteoclastswas2to4timeslowerthanitsantibacterial

efficacyagainstS.epidermidis. Thissamestudyhasalsoputforwarda

significantincreaseinthecytotoxicityofsilverparticleswhentheirsize isreducedfromthemicro(3𝜇m)tothenanometer(50nm).In

paral-lel,otherstudieshavedemonstratedthatsilverbecomescytotoxicwhen theconcentrationinthebloodexceeds2𝜇gmL−1_{, buthasnoadverse}

effectonhMSCaslongastheionicconcentrationintheapatitedoesnot exceed0.7wt%[11,12].Inconclusion,manyparametersmustbe con-sidered,anditisnecessarytocontrolboththesilverconcentrationin thecoatinganditsreleaseoncetheprosthesisisimplanted.Thisway, itshouldbepossibletocounteracttheharmfulnessofsilver.Fielding etal.showedthatstrontium,whenusedinreasonableproportion(3– 7 at%[13]),canbalancethecytotoxicityofsilverviaosteoblast stimula-tionandosteoclastinhibition[3,14],andotherwouldevenlenditsome antibacterialproperties[15]. Inaddition,Ratnayakeetal.[16]reports thatsilverandstrontium-substitutedhydroxyapatitehasitsmechanical propertiesreinforced,eventhoughitssolubilitywithinbodyfluids in-creases.Othersworksconfirmedtheefficiencyandinterestofsilverand strontiumco-doping[17,18].Thisco-dopingseemsthereforepromising forpotentialclinicalapplications,eventhoughaspecialattentionshall begiventoitsimpactontheresorptionofHAin biologicalmedium, whichhasbeenshowntobeweakenedaboveacertainsubstitution de-gree[16]anddeterminesthesustainabilityof thedevicewithin the body.

Currently,conventionalpowderplasmaspraying(APS)isthe tech-niqueused bymost of themanufacturerstoproduce coateddevices whichmeettheISOstandardsandtheFDArequirements[19,20].But thistechniquehasitslimits.Firstly,becauseoftheuseoflarge pow-ders,itdoesnotallowtoreachamicrostructureasfineasthatofthe mineralbone,eventhoughresearchhashighlightedthestimulationof osteoblastsseededonnanostructuredcalciumphosphates[21]. And sec-ondly,thistechniqueoffersonlytwowaysofintegratingthesedopants intothecoatings:eitherviathesynthesisofadopedpowderorby mix-ingpowdersofhydroxyapatiteandofdopantsinoxidizedormetallic form[3,22,23].Thelattersolutionleadstoaratherpoorstructuraland chemicalhomogeneityofthecoating,andtheformerhaslimitationin thechemicalcompositionofthedopedpowder.Indeed,thesynthesisof thesilver-dopedpowderwhilepreservingthestoichiometryofthe hy-droxyapatiteischallengingduetothedifferenceinchargeandsizeof thesilverionAg+_{. Typically,silverisonlyincorporatedatamaximum}

of10wt.%[6]. Usingasuspensioninsteadofapowderoffersan in-terestingalternativesinceitallowsononehandsprayingsmaller(even nanometric)particlesleadingtofinerstructures[24], improvingthe ho-mogeneityofboththecoatingandthedoping,andontheotherhand itenrichesthedopingwaysbyinvolvingtheuseofaliquiddispersive medium.Inaddition,bydopingthedispersivemedium,abettercontrol oftheincorporatedamountofdopantswhileimprovingtheir distribu-tionwithinthecoatingisexpected.Thequestionis:towhatextentdoes thedopingprotocolusedinrf-SPSaffectthebiologicalandmechanical propertiesofthecoating?

Thispaperaimstoevaluatethemechanicalandbiologicalproperties ofsilverandstrontium-hydroxyapatitecoatingssynthesizedbyrf-SPS, implementingdifferentdopingroutes.Thedopingofthesuspensionwas performedeitherusingapowderdopedbycoprecipitationordissolving silverandstrontiumnitratesoraddingsilvernanoparticlesinthe dis-persivemedium.A3-pointbendingtestwasimplementedtodetermine themechanicaladherenceofthecoatings,whiletheadhesionand pro-liferationofhMSCwasevaluatedtoassesstheirbioactivity,andthe de-velopmentofstaphylococcusaureusbacteriawastestedtoestablishtheir bactericidalactivity.

2. Experimental

2.1. PreparationofunsubstitutedandAg/Sr-substitutedHA

Apurehydroxyapatite(Hap)powderandanotheronesubstituted with Ag and Sr (Ag/Sr-Hap) were synthesized to produce the

sus-pensionsdestinedtorf-SPSplasmaspraying.Calciumnitrate tetrahy-drate (Ca(NO₃)₂, 4 H₂O – Sigma Aldrich), diammonium phosphate ((NH₄)₂HPO₄-Fischer),silvernitrate(Ag(NO₃),Fischer),strontium ni-trate(Sr(NO₃)₂,Sigma-Aldrich)andNH₄OHsolutionat20wt.% (Fis-cher)wereusedaschemicalprecursorsforCa2+_,PO

42−,Ag+,Sr2+and

OH−_{.Theunsubstitutedcalciumphosphatewasobtainedby}

coprecip-itationaccordingtothefollowingbalancedchemicalequation(Eq1):

10 Ca (N O 3 ) 2 + 6 (N H 4 ) 2 HP O 4 + 8 N H 4 OH → C a 10(P O 4 ) 6 O H 2 + 6 H 2 O + 20 N H 4 N O 3 (1) Thepowdersynthesisprotocolswereadaptedfromthosedescribed byKannan[25].Asolutionofdiammoniumphosphatewasaddedtoa solutionofnitrate(s)suchastheCa/Pand(Ca+Sr+Ag)/Pratiosofthe powdersHapandAg/Sr-Haprespectivelywerebothsetto1.667to pro-motetheformationofstoichiometrichydroxyapatite.ThepHwas main-tainedat9allalongthesynthesisbyadding20wt.%NH₄OHtoobtain theappropriateacidityofthephosphateion.Theinitialmolar concen-trationofthedopantswassettomol.%Ag=Ag/(Ca+Sr+Ag)=10.0and mol.%Sr=Sr/(Ca+Sr+Ag)=5.0.Afterthematurationtime(24h),the powderswerefilteredandwashedwith10timesthereactionmedium volumeofwaterbeforefreeze-drying(pressureof0.10mbar,−80°Cfor 3days).Themol%ofthedopantwithinthesynthesizedpowderwas measuredat1.41mol.%forAgand5.27mol.%forSr.Twopowders weresynthesizedfromthisprotocol,namelyHapandAg/Sr-Hap.

2.2. Suspensionpreparation

Priortospraying,thepreviouslydescribedpowders(Hapor Ag/Sr-Hap)weredispersedintowateratasolidloadof13wt.%,andan ul-trasonicprobewasplungedintothesuspensionfor5minat30Wto breakuptheaggregates,leadinguptotwosuspensions,namedafter thepowders:HapandAg/Sr-Hap.

Then, two more Ag and Sr doped suspensions were produced, where the liquid medium was doped instead of the powder: one with silver nanoparticles (AgNP_{), the other one with silver nitrates}

(Ag+_{),andbothwithstrontiumnitrates(Sr}2+_{).UndopedHapwasthen}

added tothis dopedmedium,giving risetotwo suspensionsnamed Hap+[Ag+_]/[Sr2+_]andHap+AgNP_+[Sr2+_].

In summary, four suspensions were prepared: with the undoped (Hap),with thedopedpowder(Ag/Sr-Hap),withtheundoped pow-der and nitrates (Hap+[Ag+_]/[Sr2+_{]) and with the undoped}

pow-der and a mixture of strontium nitrate and silver nanoparticles (Hap+AgNP+[Sr2+_{]),whichleadtothesynthesisofcoatingsidentified}

withthesamereferenceastheirprecursor.

2.3. Coatings 2.3.1. Substrates

TitaniumalloyTi-6Al-4V(grade5)wasselectedasthesubstrate ma-terialforitssufficientmechanicalstrength,anti-corrosionandbioinert properties[26,27].Platesof50×55×1.6mm3adaptedtotheprocess wereused.The3-pointbendingsamples(50×10×1.6mm3)werelaser pre-cutexceptattwoslotsinthetitaniumplates,aswellascylindersof 10mmindiameterusedforbiologicalandbacteriologicaltests.

Prior toplasma spraying, the plates were cleaned with solvents (ethanolandacetoneinultrasonicbath)andgrit-blastedwithalumina (F120)atanincidenceangleof90°,a120mmblastingdistanceand apressureof4.5bar.Thearithmeticroughnessofthesurface(Sa)was determinedwithaprofilometertobeapproximately1.5±0.2µm.

2.3.2. Plasmasprayingparameters

Rf-SPS coatings were produced at the Université de Sherbrooke (Québec)withaplasmageneratedbya50-kWRFplasmatorch operat-ingat3MHz(PL-50modelfromTeknaPlasmaSystemInc.)equipped

withasupersonicnozzle.Thesuspensionwasinjectedbyaperistaltic pumpatafeedingrateof12.5mLmin−1_{andthroughasupersonic}

noz-zlewitha45mmoutputdiameter[28]. Theliquidatomisationprobe washome-designed,andspraysofwaterdropletswithamean diame-terD₅₀of11µmwereproduced.Thecompositionoftheplasmawasa mixtureofargonanddioxygenbasedonpreviousworkperformedby Loszachetal.[29], andthegeneratedpowerwas35kW,withapressure withinthechamberadjustedto90torr.Theworkingdistancefromthe torchtothesampleholderwassetto185mm,andthedistanceswept bythesampleholderundertheplasmajetwassetto80mm.8 preheat-ingpassesfollowedby30sprayingpasseswerenecessarytoachievea thicknessofabout100µm.

2.4. Characterizationofpowdersandcoatings

AllthefollowingcharacterizationswereperformedinlinewithISO standard(i)ISO13,779–2:2008and(ii)ISO13,779–3:2008[19,20].

TheparticlesizedistributionwasmeasuredwithaMalvern Master-sizerHydro2000Sgranulometer(UK)inaqueousmedia,toreproduce thestateofthesuspensionbeforeplasmaspraying.Themeasuringcell waskeptunderagitationat1750rpmforalltrialstoprevent sedimen-tation.

Theelementalcompositionsofpowdersandcoatingsweremeasured byaninductivelycoupledplasmaopticalemissionspectrometerJobin YvonUltima2(Horiba,Japan).Theroughnessofthesubstratesandthe coatingswasevaluatedusing aSensofarS-Neoxconfocal microscope (Barcelona,Spain)atx10magnification.Theaveragevaluewas deter-minedfrom5scanningzonesof1750× 1300µmat0,90°,180° and 270°.Thecrystallinephaseanalysisof thepowdersandthecoatings wereinvestigatedusing D8diffractometer(Bruker AXSGmbH, Karl-sruhe,Germany)withCuK𝛼 radiation(𝜆 =1.5406Å)producedat40kV

and50 mA.Datawererecordedin the20 ° - 60 ° 2𝜃 rangewith a

stepsizeof0.02° and38sperstep.Thepatternswereidentified us-ingMatch!softwareandsemi-quantitativeanalyseswererealizedusing theReferenceIntensityRatiomethod[30], usingthefollowingJCPDS data(HA,hydroxyapatite,JCPDSno.00–009–0432;𝛼-TCP,tricalcium

phosphateCa₃(PO₄)₂, JCPDSno.04–018–9895;𝛽-TCP,tricalcium

phos-phateCa3(PO4)2, JCPDSno.00–070–2065;TTCP,tetracalcium

phos-phate,Ca₄P₂O₉, JCPDSno.00–025–1137;CaO,calciumoxide,JCPDS no.04–011–9020).ThecrystallinityratioR₁(Eq2)wascalculated ac-cordingtotheformulaproposedinthestandardISO13,779–3,involving theuseofafullycrystallisedstandard(hydroxyapatitepowdercalcined at1000°Cfor15h):

R1=

Integratedintensityof10linesofthesample

Integratedintensityof10linesofthestandard× 100 (2) FourierTransformInfrared FTIR analysiswas carried out with a PerkinElmer 1700spectrometer(USA)onscraped coatingsdispersed inKBrpelletsinaratioof0.3percentmass,ataresolutionof4cm−1

inthe400-4000cm−1_{range.Chemicalstateofdopantswasevaluated}

byX-rayphotoelectronspectroscopy(XPS)withaVGESCALAB220i-XL (ThermoFischerScientific,US)spectrometer,witha150µm-lateral reso-lution.Theobservationofsurfaceandmicrostructurewasconductedby SEMwithaFEIQuanta450(Bruker,Germany),usingtheback-scattering modeat12.5kVand10mmworkingdistance,andbySTEMwitha JEM-ARM200FColdFEG(Jeol,Japan).Todeterminetheporosityfromthe cross-sectionimages,thesampleswerecut,thenmountedandvacuum impregnatedwithcoldepoxyresin.Theywerethengroundandpolished inaccordancewiththeproceduredescribedbyGeelsetal.[31].Image analysiswereperformedwithImageJon10differentviews, implement-ingmanualthresholdinganddespeckling.Contactanglemeasurements werecarriedoutwithaDigidropMCATgoniometer(GBX,Ireland).An equalvolumeofmilliQwater(2 µL)wasplacedon everysampleby meansofamicropipetteformingadroponthesurface,andthecontact anglewasrecorded10safterdropdeposition.

2.5. Calcium,silverandstrontiumrelease

The evolution of the dopedcoatings in physiological media and specificallyinaggressivemediummimickingthelocalinflammatory sit-uationappearingafterimplantationwasassessedbyimplementing dis-solutiontestsinastrongacidicmedium.Toperformthedissolutiontest, acylinderofeachspecimenexceptfortheHap+AgNP+[Sr2+_]condition

wasimmersedin50mLofHNO₃acidifiedwaterat37°C,pH3.5and subjectedtoconstantshakinginaDissolutestSotaxCE6unit (Switzer-land)ata9mLmin−1_{flowrateinclosedcirculation.Anextracoating,}

synthesizedinsimilarplasmaconditions,fromasuspensioncontaining astrontium-dopedhydroxyapatitepowder(3.39±0.04wt.%)and sil-verunderionicandmetallicform,wasannealedat600°Cfor2hand testedwiththeotherstohighlighttheeffectofcrystallinityonthe dis-solutionbehaviorofthedopedcoatings.Analiquotof10mLofmedium wascollectedateachtimepointstartingat30min,with10mLoffresh solutionaddedateachtimepoint,for5days.ThepHwasmeasured foreachaliquot.ThealiquotcontainingthereleasedAg+_andSr2+_was

analysedviaatomicabsorptionspectrophotometer(ContrAA300 Ana-lytikJena,Germany)forthestrontiumandcalcium,andinductively coupledplasmamassspectrometer(ICP-MS7500ce,Agilent Technolo-gies,USA)forthesilver.

2.6. Mechanicalproperties

Theadherenceofacoatingisoneofthecriticalcharacteristicsthatis requiredwhendevelopingnewsolutions.Thetestselectedinthepresent studyisathree-pointbendingtest,elaboratedbyRocheetal.[32]and isdescribedintheISO14679:2001standard[33].Apolyepoxide stiff-enerof 25 ×4×5mm3 is placedandadheredin thecenterofthe surfaceofthecoatedtestpiece,whosesurfaceareais50×10mm2.The wholepieceisthenplacedinanInstrontestingmachineandisloadedin bendingatacrossheadspeedof0.50mm.min−1_{,untilfailure.Adhesion}

energy(mJ)isthencalculatedfromthefailureforceandthedeflection. Then,theresidualcoatingpercentageisevaluatedwithEDS cartogra-phy.Ifthepercentageofresidualcoatingisbeyond65%,therupture isdesignedascohesive,below35%meansadhesive, andin between standsforamixedrupturetype.Inordertoensurereproducibilityof theresults,fourreplicateswereused.

2.7. Resazurinassay,cellproliferation

Humanmesenchymal stem cells wereisolated from humanbone marrowaspiratesobtainedfromtheUniversityHospitalofBordeaux af-terapprovalbythelocalethicalcommissionandwrittenconsentofthe patient.TheAlamarBlue○R _{cellviabilityreagent(Thermofischer,USA)} wasusedtoevaluatecellproliferation.Thesampleswerefirstplaced intoanalphaMinimumEssentialMedium(𝛼-MEM)(Sigma,USA)fora

soakingpurposein48-wellplates.1mLof𝛼-MEM+10%Fetalbovine serum(FBS)holdingadensityof50×103cellsml−1_{werethenseeded}

ontothesamples.Ateachtimepoint,themediumwaswithdrawnfrom thewelland1mLof𝛼-MEM+10%serum+10%alamarBluesolution at3.5mgmL−1_{wereadded.Afterincubation,200µloftheresulting}

su-pernatantwastransferredtoa96-wellplateandreadbyaplatereader at590nm.Cellproliferationwasevaluatedat4,7and14 days,and theviabilityofthecellsCV(Eq3)wasestimatedthroughthefollowing formula:

CV=%proliferationoftestedgroup

%proliferationofreference × 100 (3) Foranygivenexperiment,eachdatapointrepresentsthemean± standarderroroffiveindividualcultures.Datawereprocessedby anal-ysisof variancewithtwo-wayANOVA testandBonferroni’sposttest tocompareallmaterialswitheachotherateachtime.Pvalues<0.05

Table1

Physicochemicalcharacteristicsofprecursors.

D 10 D 50 D 90 wt.% Ag wt.% Sr Ca/P (Ca + Sr + Ag)/P

Hap 2.8 ± 0.1 9.3 ± 0.3 24 ± 3 – – 1.68 ± 0.03 –

Ag/Sr-Hap 2.2 ± 0.1 4.6 ± 0.1 9.5 ± 0.2 1.40 ± 0.02 4.30 ± 0.07 1.49 ± 0.03 1.59 ± 0.03 Hap + Ag NP + [Sr 2+ ] 2.8 ± 0.1 9.3 ± 0.3 24 ± 3 _{0.40 ± 0.01 4.37 ± 0.05 1.68 ± 0.03 1.75 ± 0.03}

Hap + [A g + ]/[Sr 2+ ] 2.8 ± 0.1 9.3 ± 0.3 24 ± 3 _{1.54 ± 0.01 4.32 ± 0.02 1.68 ± 0.03 1.77 ± 0.03}

2.8. Antimicrobialactivity

TheantimicrobialactivityofdopedHA-coatedsampleswas deter-minedbychallengingwithS.Aureus(ATCC33,591,MRSA).The bac-terialstocksolutionwaspreparedextemporaneouslyinsteriledistilled waterbeforedilutioninaculturemedium.Thepreparedsampleswere sterilizedbyautoclavingat121°Cfor15minandplacedinto24-well plates,coveredwith2mLofbacterialsuspension103_UFCmL−1_and

incubatedbyrotation for24hat 37°C.Each samplewastakenout at6,24and48handrinsedwith4mLsterilewaterandplacedinto 3mLofthesamesolutionandshakenbyvortexfollowedby1minof ultrasounds.Afterdilution,1mLofthesupernatantwasseededin Tryp-caseSoyAgar(TSA)andthenumberofvisiblecellswasdeterminedby quantifyingCFUs.APScoatingwasusedasacontrolgroup;eachtest wasperformedintriplicate.Thedegreeofthemicrobialcellreduction

BR(Eq4)wasestimatedbythefollowingformula:

BR= ∑

𝑛 =3Log(Nbbacteriainstandardgroup)−

∑

𝑛 =3Log(Nbbacteriaintestgroup)

3

(4) 3. Results

Inordertoassessthequalityofthecoatingsinrespectwiththe com-mercialstandard,the mechanicalandbiologicaltests werealso per-formedwithanatmosphericplasmaspray(APS)hydroxyapatite stan-dardcoating,servingasReference.

3.1. Physico-chemicalanalysis 3.1.1. Suspension

The suspensions showed very similar characteristics, reported in

Table1.ThemeanD₅₀iscomprisedbetween4.6±0.1µmforthedoped powderand9.3±0.9µmfortheundopedHap.Quitesimilar stron-tiumconcentrationsofabout4.35wt.%weredeterminedwhateverthe dopingstrategy.1.40and1.54wt.%wererespectively measuredfor silverinthecaseofthesuspensionsformulatedeitherfromthedoped Ag/Sr-HappowderorfromtheundopedHappowderaddedwith ni-trates (Hap+[Ag+_]/[Sr2+_{]).Alowercontentof 0.40wt.%wasfound}

forthesuspensionproducedfromtheundopedpowderHapmixedwith AgNP _{nanoparticles(Hap+Ag}NP_+Sr2+_{).Thecation/Pratiosvary from}

1.68±0.03fortheundopedpowderto1.59±0.02fortheAg/Sr-doped powderandto1.76± 0.03forthedopedsuspensions.XRDpatterns ofundopedanddopedsynthesizedpowders(Fig.1)showthetypical crystallographicapatiticstructureofCa₁₀(PO₄)₆(OH)₂(JCPDS00–009– 432),wherenosecondarycrystallinephaseisdetected.Despitean ob-viousdiminutionofthecrystallitesize,thecrystallinityratioisslightly betterforthedopedpowderwithavalueof97±3%forAg/Sr-Hap, comparedwiththevalueof92±3%forHappowder.

3.1.2. Coatings

ThequantitativeXRDanalysis(Table2)establishesthatHAisthe majorphaseforallcoatings(Fig.2).86wt.%ofHAtogetherwitha crys-tallinityratioof70%wereassessedforthecoatingselaboratedfromHap undopedpowder,withorwithoutdopingagentsinthesuspension.The secondaryphasesinthedopedcoatingswereTTCPandCaOinquantities twiceaslargeasinthecoatingproducedfromtheundopedpowderthat

alsocontains𝛼-TCP.Incontrast,forthecoatingmadewiththedoped

powderAg/Sr-Hap,anamountof62wt.%ofHAphasewith33wt.%of

𝛼-TCPweremeasured,aswellasalowcrystallinityratioof57%in

re-gardtothevaluerecommendedintheISO13779–2standardthatmust beover45%.Asmallpeakwasnoticeableat2𝜃=36° inAg/Sr-Hapand Hap+[Ag+_]/[Sr2+_{]patterns, whichcouldbe attributedtoAg (JCPDS}

No.4–0783)orAg₂O(JCPDSno.19−1155),indicatingthatthesilver ionswerepossiblyturnedintometallicoroxidizedsilverduringplasma spraying.FTIRanalysisvalidatedtheapatitestructureofthecoatings sinceallcharacteristicbands(𝜈₁=938cm−1_,𝜈

2=420cm−1,𝜈₃=1017 cm−1_,and𝜈

4=567cm−1_{)arepresent,butalsoevidencedadiminution}

ofthehydroxylationdegree(𝜈_LOH−=631and𝜈_SOH−=3570cm−1)for alldopedcoatings(Fig.3)[34].Spectradisplayedpoorlydefinedbands thatconfirmedthepoorcrystallinityoftheAg/Sr-Hapcoating.Aweak relativeintensityofthebandsassignedtoOH−_{groupevidencesalow}

hydroxylationdegreeforbothAg/Sr-HapandHap+[Ag+_]/[Sr2+_]

coat-ingsincomparisonwiththeundopedHapandHap+AgNP+Sr2+

coat-ings.Thisindicatesthatsilvernanoparticlespromotethemaintenance ofapatitehydroxylationduringplasmaspraying,probablybecausethey arelikelytoconsumeheatinalargerextentthroughtheirmeltingand vaporizationsteps.

Thecrosssectionsofthecoatings(Fig.4)observedbySEMshowsa goodqualityofinterfaceandauniformmicrostructure,withsome het-erogeneouszonesandcracks, characteristicofplasmasprayingas al-readymentionedinliterature[27].Theporosityevolvesfromabout3% forHap+[Ag+_]/[Sr2+_]andHap+AgNP_+Sr2+_{coatingsto5%for}

Ag/Sr-Hapand7%fortheundopedcoatingHap.Itseemsthatthepresence of nitrateswithin thesuspensiondiminishestheporosity ofthe pro-ducedcoatings.Asimilartrendcan benoticedfortheroughness. In-deed,Sa parametersequalto4.4±0.1µmweremeasured forboth Hap+[Ag+_]/[Sr2+_]andHap+AgNP_+Sr2+_{coatingsandhighervaluesof}

about5.4±0.4µmwererecordedforbothHapandAg/Sr-Hap coat-ings.Thepresence ofsilveror silveroxideparticleswas clearly evi-dencedintobothAg/Sr-HapandHap+[Ag+]/[Sr2+]microstructures us-ingSEMimagesathighmagnificationinbackscatteringmode(Fig.5) Thesenanoparticlesweremainlylocatedatthesplats-poresbordersas showninFig.5b,supportingthehypothesisthatsilverdoesnot favor-ablysubstituteintheapatiticstructure.TheSEMimageanalysisofthese particlesresultedinameandiameterof230±95nmwhenthedoped powderisimplementedinthespraysystemand115±91nmwhen sil-verisintroducedasnitrateinthesuspension.Butconsideringthelimits oftheSEM,evensmallerparticlesizesaround1.5µmindiametercould bedetectedbySTEM.Finally,EDScartography(Fig.6)performedon theHap+[Ag+_]/[Sr2+_{]confirmedtheincorporationofstrontiumtoan}

extentdependingonthemeltingdegreeoftheparticleatimpact.Indeed, splatsshowedlowvariationsofstrontiumcontentfromonetotheother, probably duetotheimplementationprocesswhich ledtoavariable concentrationofstrontiumintohydroxyapatiteparticles,dependingon theirsizeandtheirmeltingdegree.TheEDSmappingperformedonthe Ag/Sr-Hapcoatingshowedontheotherhandahomogenous distribu-tionofstrontium(Fig.6cand6g)alloverthemicrostructure,confirming itssubstitutionintheapatite.

Elementalquantificationofthescrapedcoatingsrevealsalossofthe silvercontentduringplasmaspraying:74%lossoftheinitialsilver pro-portionfortheAg/Sr-Hapcoating,80%forHap+[Ag+_]/[Sr2+_]coating

Fig.1.X-Raydiffractogramsofundopedand doped synthesizedpowderscomparedwitha commercialfullycrystallisedhydroxyapatite.

Table2

CrystallographiccompositionobtainedbyRiRmethodandcrystallinityratioofcoatings(ISO 13,779-3[20]).

Hap Ag/Sr-Hap Hap + [A g + ]/[Sr 2+ ] _{Hap + Ag} NP + [Sr 2+ ]

Annealed doped coating HA (wt%) 86 ± 4 63 ± 13 83 ± 2 86 ± 5 98 𝛽-TCP (wt%) – – – – 1.4 𝛼-TCP (wt%) 5.9 ± 1.9 33 ± 11 – – – TTCP (wt%) 7.4 ± 1.6 4.0 ± 2.2 15 ± 2 12 ± 4 – CaO (wt%) 0.7 ± 0.3 0.1 ± 0.2 1.9 ± 0.6 1.6 ± 0.6 1.0 R 1 (%) 71 ± 5 57 ± 11 68 ± 5 70 ± 4 90 Rp (%) 5.2 4.6 5.6 5.4 3.2

Fig.2. X-Raydiffractogramsofdopedand un-dopedrf-suspensionplasmasprayedcoatings.

ratioishighlyover1.667foreverycoatingentailingtheundoped pow-der(Hap,Hap+AgNP+Sr2+_,Hap+[Ag+_]/[Sr2+_{]).Theoneproducedfrom}

thedopedpowderisveryclosetostoichiometrywhenconsideringthe (Ca+Sr+Ag)/PratioinsteadoftheCa/Pratio.Strontiumdidnotsuffer anysignificantloss,sinceitsconcentrationrangesfrom4.24±0.08to 4.62±0.07wt.%inthecoatings.

Thecontactanglesvarysignificantlyfromonecoatingtotheother, asdisplayedinTable3.Butoverall,theyallstandbelow60° whichis consideredastheupperthresholdtofavorizecellattachment accord-ingtoYeungetal.[35].Almosthydrophobicbehaviorcanbeassigned tocoatingsproducedfrom suspensionscontaining nitrates: 47.6° for

Hap+[Ag+_]/[Sr2+_{]and38.8° forHap+Ag}NP_+Sr2+_{.Incontrastcoatings}

producedwiththeHapandAg/Sr-Happowdersaremorehydrophilic (17.8° and22.1° respectively).

3.2. Cationsrelease

ThekineticsofreleaseofCaandSrarereportedinFig.7aand7b respectively.BothfiguresrevealedthatalargereleaseofCaandSrin theacidicmediumwasnoticedduringthefirst500minofimmersion. After10htheconcentrationsofcalciumandstrontiumleveledoff to evolve very slowlyup to4days. At theend of theassay the

disso-Fig.3. FTIRspectraofthecoatings,with phos-phatePO43-bandsrepresentedwiththe𝜈1,𝜈2,

𝜈₃ and𝜈₄ signsandOH− _{bandsrepresented}

with𝜈sOH−and𝜈LOH−.

Fig.4. SEMimagesofrespectivelycrosssectionsandtopsurfacesofHap(a,e),Ag/Sr-Hap(b,f),Hap+[Ag+_]/[Sr2+_{](c,g),Hap+Ag}NP_+[Sr2+_{](d,h)coatingsat}

x500magnification.

Fig.6.STEMmicrographofHap+[Ag+_]/[Sr2+_{](a)andAg/Sr-Hap(e)coatingcross-sectionandrespectiveEDSmappingofCa(b-f),Sr(c-g)andAg(d-h).}

Table3

Physicalandelementalcharacteristicsofcoatings.

Sa (µm) Porosity (%) wt.% Ag wt.% Sr (Ca + Sr) /P Ø Ag np.(nm) Contact angle (°)

Hap 5.4 ± 0.4 6.7 ± 3.1 0 0 – 0 18 ± 4

Ag/Sr-Hap 5.4 ± 0.3 5.0 ± 2.3 0.394 ± 0.004 4.62 ± 0.07 1.69 ± 0.02 231 ± 95 22 ± 13 Hap + Ag NP + [Sr 2+ ] 4.4 ± 0.1 3.0 ± 1.5 _{0.064 ± 0.002 4.23 ± 0.08 1.79 ± 0.02 810 ± 646 39 ± 12}

Hap + [A g + ]/[Sr 2+ ] 4.4 ± 0.1 2.5 ± 1.6 _{0.342 ± 0.005 4.55 ± 0.06 1.80 ± 0.02 115 ± 91 48 ± 7}

Table4

Mechanicalandbiologicalpropertiesofcoatings.

Adhesion energy (mJ) Residual coating proportion (%) Cell viability CV after 14 days (%) Antibacterial reduction BR at 24 h (%) APS Reference 0.9 ± 0.1 93 ± 3 100 – Hap 4.2 ± 1.1 22 ± 3 105.4 – Ag/Sr-Hap 5.5 ± 2.9 91 ± 13 99.42 2.9 ± 1.2 Hap + Ag NP + [Sr 2+ ] _{7.6 ± 4.2 79 ± 2 38.06 1.2 ± 0.9} Hap + [A g + ]/[Sr 2+ ] _{4.3 ± 1.9 61 ± 2 99.97 5.0 ± 1.2}

lutionofAg/Sr-Hapcoatinghasreleasedabout4.5timesmore stron-tiumand1.5timesmorecalciumcomparedwiththecoatingelaborated fromHap+[Ag+_]/[Sr2+_{].TheproportionofSr}2+_{averages5at.%ofthe}

cations(Ca2+_,Sr2+_andAg+_{)intheAg/Sr-HapandHap+[Ag}+_]/[Sr2+_]

coatings.Butthis ratioreaches24 at.%and9.4 at.%inthe dissolu-tionmediumforAg/Sr-HapandHap+[Ag+_]/[Sr2+_{]respectively,which}

meansthatstrontiumispreferentiallyreleasedinthecaseofthe coat-ingmadefromthedopedpowder.AsAg/Sr-Haphasalowcrystallinity andahighproportionofsecondaryphasesevidencedbyXRDandFTIR analyses,itcanbeassumedthatstrontiumislikelyincorporatedinthe secondaryphases.Thisconclusionisreinforcedbythefactthatthese phasesareknowntobemoresolublethanhydroxyapatiteina biolog-icalmedium.Somestudiesalsosuggestapreferentialincorporation,at

highconcentration,ofstrontiumintotheTCPstructure[36,37],thatisa crystallinephasepresentinquitehighproportionintheAg/Sr-Hap coat-ing.Finally,X-raydiffractogramsofthecoatingsurfacesrecordedafter dissolutionthatrevealedthedisappearanceofthesecondaryphasesfor allthecoatings,tendtoconfirmthisinterpretation.ThepHevolution curvesofthemediumincontactwiththedopedsamplesshowthesame trendwithaninitialburstfollowedbyasignificantdecreaseduringthe first500min,followedbyare-increaseandastabilizationatan inter-mediatelevelafter2000min.Theendvaluesdiffer,with6.07forthe annealedsample,6.85forHap+[Ag+_]/[Sr2+_{]and7.02forAg/Sr-Hap.}

ThekineticofreleaseofAginthemediumisverydifferentfromthe onesofstrontiumorcalciumandcanbecomparedwiththepH evolu-tioncurve.SilverstronglydissolveswhenthepHisthelowest,andas

Fig.7. Releasedcumulatedquantityofcalcium(a),strontium(b)andsilver(c)inthedopedcoatingandpHevolutionwithtime(d).

pHstartsrisingdue tothedissolutionof thephosphocalcicmatrix,it stopsdissolvingandseemstoprecipitate,ashighlightedbybothcurves (Fig.7cand7d).

3.3. Mechanicalproperties

Theresultsofthe3-pointbendingtestappliedonthesampleswere evaluatedincomparisonwiththecommercialAPSplasma-sprayed hy-droxyapatitecoating(Table4).Allrf-SPScoatingsshowadhesion ener-gies4.6to8.4timesbetterthantheone(0.9mJ)ofthecommercial coat-ing,withthemaxvalue(7.6mJ)measuredforHap+AgNP+Sr2+_coating.

Nospecifictrendcanbedrawnexceptthatthestandarddeviation in-creaseswith the value itself. Commercialcoating as well as Ag/Sr-HapandHap+AgNP+Sr2+_{coatingsexhibitedacohesiverupture(over}

60%ofthecoatingisstillattachedtothesubstrateafterthetest).For Hap+Ag+_/Sr2+_{amixedmode(61%)rupturewasidentifiedandforthe}

undopedcoatingHapitwasanadhesive(22%)rupture.Nodirect cor-relationbetweentheadherenceenergyvaluesandthefailuredynamics wereestablished.

3.4. Invitrocell-materialsinteraction

Cellproliferationresultsonthesamplesafter4,7and14daysof incubationareshowninFig.8,TabIV.Ateachtimeinterval,thenumber

ofcellsonundoped(HapandReference)coatingsanddoped(Ag/Sr-Hap andHap+[Ag+_]/[Sr2+_{])coatingswasquiteidenticalandconsiderably}

greaterthanthatonHap+AgNP+[Sr2+_{]coating.Theproliferationrates}

reachedmorethan500%oftheirinitialpopulationafter14days,with aslightlybetterresultforHapcoatingthanforthereference.Atday7, Hap+[Ag+_]/[Sr2+_{]showedapromisingburst(340%)sinceitovertook}

bothundoped coatings(290%),butthis resultwasnot confirmedat 14days.Theadditionofsilverasnanoparticlesinthesuspensionhad asignificantnegativeimpactoncellproliferation,diminishingthecell viabilitybelow70%.

3.5. Antimicrobialactivity

Bactericidal potential of our coatings toward S. aureus was ex-plored, anddifferent behaviours emerged depending on the doping route(Fig.9).Theundopedcoatings(ReferenceandHap), as antici-pated,didnotshowanyantibacterialeffectivenessandnosignificant differencecouldbedetectedbetweenthem,implyingthatthechange ofmicrostructuralscaledoesnotimpactthebacterialgrowth.If Ag/Sr-Hapgavetroublingresultsafter6hwithasignificantproliferationof bacteria,itfinallymettheantibacterialexpectationswithareduction

BRof2.9after24h(Table4).Hap+[Ag+_]/[Sr2+_{]coatingproveditself}

veryefficientsinceitreducedthenumberofviablecellsbyalmost5log after24hofinoculation.Ontheotherhand,Hap+AgNP+[Sr2+_]coating

Fig. 8. Proliferation of hMSC with time on plasma-sprayedcoatings(n=5),∗_P≤0.05∗∗∗

P≤0.001.

Fig.9. BacterialproliferationofS.aureuson dopedandundopedcoatings(n=3),∗_P≤0.05 ∗∗_P≤0.01.

showedareductionBRof1.2only,whichisinsufficientfroma bacteri-cidalpointofview.

4. Discussion

4.1. Influenceofthesuspensionformulationonthecoatingcomposition

Thechemical stabilityof hydroxyapatite coatings in a biological mediumisessentialtoguaranteetheirsustainability.Amongthe char-acteristicsintimatelyrelatedtothisstability,thecrystallinityandthe phasepurityofthecoatingarepredominantsincetheydetermineboth itsresorbabilityanditsmechanicalproperties,andthose characteris-ticsaremainlydeterminedbythesprayingconditions.The decompo-sitionofstoichiometrichydroxyapatiteandtheformationofsecondary includingamorphousphasesareacommonoccurrenceinconventional plasmaspray.Thehightemperatureoftheplasmaarc andtherapid coolingatthesubstrateimpactleadtotheformationofdecomposition phases(TCP,TTCP,ACPandCaO),whichweakenthecoatingby

re-sorbing significantlyfasterin thebody thanhydroxyapatite[38,39]. Consequently,theISO13779:2018standardrequiresapercentageof thesecondaryphaseslowerthan30wt.%,andacoatingwitha crys-tallinityratiogreaterthan45%.TheX-raydiffractometrydatashownin

Fig.2andTable2indicatethatcoatingsmadewithasuspension contain-ingthedopingelementsasnitratesareconstitutedofTTCPandCaOas secondaryphasesincontrasttotheequivalentundopedHapcoatingthat contains𝛼-TCP,TTCPandCaOforasimilarconcentrationof

hydroxya-patite.Insuspensionplasmaspraying,thedisturbancecausedbythe ad-ditionofsubstancessuchasnitratesormetalnanoparticlesinthe disper-sivemediummodifiestheplasmathermalcharacteristicsandtherefore theheattreatmentsustainedbyphosphocalcicparticles.Theirdegreeof decompositioncanthusbelargelyaffected:inthepresentcase,nitrates increasethethermalfluxavailableleadingtohightemperaturephase formationwhilemaintaininghighcrystallinityratioandHA%as com-paredtoundopedcoating(Hap).Regardingthemicrostructure, param-eterssuchastheporosityandtheroughnessalsoseemtobeimpacted: thepresenceofnitratesandtheuseofanon-stoichiometricdoped

pow-derleadtoadecreaseofporosity,duetoahighermeltingdegreeinthe formercase,andtoalackofcrystallinitywhichfavorsthecompacityof thematteratimpactinthelattercase.ResearchhasshownthatHAisan idealplatformforcationicsubstitutionand,specifically,silverand stron-tiumionscansubstitutecalciumionintheapatitestructure.Asthe ad-ditionofsuchdopantscanalterthephysico-chemicalcharacteristicsof HA,onewouldhaveexpectedpeakshiftsintheXRDanalysisofthe coat-ingsindicatingsignificantincorporationofthedopingionsintheHA latticestructureduetotheuseofeitherthedopedpowder(Ag/Sr-Hap) orthedopedmedium(Hap+AgNP+[Sr2+_]orHap+[Ag+_]/[Sr 2+_]).While

nopeakshiftswereobservedfordopedcoatings,itremainslikelythatSr wasincorporatedinoneofthecalciumphosphatelatticestructureasno separateSrorSrOpeaksweredetectedinthediffractograms.Incontrast, identificationofAgorAg₂Opeaksinthediffractogramsandthe obser-vationofthecoatingmicrostructureevidencedthatsilverisnot incor-poratedintheHAlatticestructure.HighmagnificationSEMandSTEM observationsclearlyconfirmedthatsilverisincorporatedas nanoparti-clesinthedopedcoatings,whosesizesdependonthedopingmethod. Indeed,itsionicradius14%higherthanthatofcalcium(128ppm ver-sus112ppmrespectivelyforacoordinationnumberCN=8,[40,41]) combinedwiththechargemismatchrevealtheinstabilityoftheAg+

ionintotheapatitestructure,eitherpreventingitsincorporationduring sprayingorleadingtoitsmigrationoutoftheapatiticstructure,which probablyinducesthesignificantdropofcrystallinitynoticedfor Ag/Sr-Hapcoating.Duetothisphenomenon,alargeamountofsilverdopant islostduringspraying(mainlybecauseofthetemperatureandpressure thatfavorvolatilizationofmetallicsilver)andtheexactamountpresent intothecoatingishardlycontrollable.Toconfirmthechemicalstateof silverinHap+AgNP+[Sr2+_]andHap+[Ag+_]/[Sr2+_{]coatings,XPS}

anal-ysiswasperformedoncoatingswitha10timeshighersilver concen-trationinordertodetectsilveranddeconvolutethedoublet3d_3/2 et 3d_5/2 (367,75 eVfor3d_5/2 et373,75eVpour3d_3/2).Itshowedthat silvernanoparticlesintheHap+[Ag+_]/[Sr2+_{]coatingare90%metallic}

and10%oxidized,whiletheyare65%metallicand35%oxidizedinthe Hap+AgNP+[Sr2+_{]coating.AccordingtotheEllingham’sdiagram,fora}

partialoxygenpressureof0.062bar,silveroxideisreducedto metal-licsilveratatemperatureabove398K(approximately125°C).During plasma-spraying,mostparticlesreachmuchhighertemperaturesonthe probe-substratepath.Butoutofthermodynamicequilibrium,the ther-malenergytransmittedtoreducethesilveroxidetometallicsilveris limitedbydynamicconsideration,duetoshortresidencetimeofthe droplets/particles.ForsuspensionscontainingAgNP_{silvernanoparticles}

andconsideringthatthesenanoparticlesarealreadypartlyoxidized be-foreplasma-spraying,theirconversiontofullymetallic nanoparticles isfurther limitedbecauseof theirlowersurface/volumeratiowhich doesnotfavorthethermaltransfer.ForthedopingmodeinvolvingAg+

ions,andtakingintoaccounttheinstabilityofsilveroxideover metal-licsilverin thesprayingconditions, itcan be assumedthat mostof theionsareinstantlyreducedin metallicsilvereitherwithoutgoing throughanoxidizedstate,or undergoingamoreextensivereduction ofoxidizedparticlesduetotheirsmallersize.Becausethesilverpeak wastooweaktoperformaqualitydeconvolutionoftheXPSspectrum recordedforAg/Sr-Hapcoating,andsincetheconcentrationofsilver couldnotbesufficientlyincreasedintheAg/Sr-Happrecursorpowder, thechemicalstateofthesilverparticlescouldnotbeevaluated.Thanks totheelementalanalysisperformedbySTEM(Fig.6),itwasclearly es-tablishedthatstrontiumishomogeneouslyincorporatedinthecalcium phosphatelatticestructureregardlessofitsionicradius(126ppm),even thoughsomesplatsofdopedcoatingsproducedfromnitratesrevealed concentrationheterogeneitieswithstrontium-richerzoneslocatedatthe grainboundariesofthesplats.Fromtheelementalpointofview,the (Ca+Sr)/P(Table3)ratiosofHap+AgNP+[Sr2+_]andHap+[Ag+_]/[Sr2+_]

arethehighestofall(1.79±0.02and1.80±0.02respectivelyagainst 1.73± 0.03fortheundopedcoating),validatingSrincorporationor phosphorusevaporationduringplasmaspraying.SinceAgleftthe

ap-atitic structure during plasma spraying, it was not included in the ratio.

4.2. Microstructureandmechanicalproperties

Thecoatingbondstrengthisascrucialasitschemistryforthe clini-calsuccessofloadbearingimplants.Theadhesionenergy(Table4)that wasdeterminedinthepresentstudyrepresentstheforcerequiredto de-tachthecoatingfromthesubstrate.Dependingonthefailuremode,the adhesionenergyreflectsthetenacityeitheroftheinterface,eitherofthe nearbymicrostructure.FortheundopedSPScoating(Hap)the percent-ageofresidualcoatingonlyreaches22%,correspondingtoanadhesive failure,whilefortheAPSreferenceitisabout93%,correspondingtoa cohesivefailure.Acohesivetypefailure(residualcoating>60%)means

thatthecohesionofthemicrostructureneartheinterfaceisweakerthan thecoating/substrateinterface.Indeed,itisgenerallythemostfragile areabecauseofthehigherproportionofamorphousphases,duetoa lowerlocalcrystallizationofthematerialcausedbyquenching.This lo-calizedamorphizationhasadoubleeffectontheadhesionenergy:on onehanditpromotestheanchoringofthecoatingbylimitingthe gener-ationofresidualstressesusuallygeneratedbythecrystallizationofthe material,butontheotherhanditweakensthecohesionofthe phospho-calcicmaterialitself[42].Thesetwoantagonisticeffectstherefore con-tributetodefinethefailuremodeandadhesionenergy.Theevolutionof thecrystallinityratiowiththefailuremodeandtheadhesionenergyis illustratedinFig.10.Therefore,foralowcrystallinityratioofthe coat-ing(Ag/Sr-Hap)theruptureistypicallycohesive,whileforahighratio (undopedHap),theruptureisadhesive.Inthecaseoftheundoped coat-ing,sinceitbrokeataloweradhesionenergythanthedopedAg/Sr-Hap coating,itcouldbeassumedthatthereisahigheramountofresidual stressesatitsinterface,certainlycausedbythehighstabilityofundoped hydroxyapatitethatfavorsafastercrystallization.Whenusingnitrates intheprecursorsuspension(Hap+[Ag+_]/[Sr2+_]andHap+AgNP_+Sr2+_),

thecrystallinityratiosaresimilar tothatoftheundopedcoating(67 and70%),buttheamountofhightemperaturesecondaryphases(TTCP andCaO)ishigher,whichmakestheruptureevolvefromadhesive to-wardsthemixedandcohesivetype(61%and79%residualcoating re-spectively).Itiscertainlythehighmeltingdegreeofthephosphocalcic particlescausedbytheuseofnitratesthatexplainstheimprovement ofthesurfaceofcontactwiththesubstrate,providingastronger inter-face.However,itwasfoundthattheadhesionenergyismuchhigherfor Hap+AgNP+Sr2+_{(7.6±4.2mJ)thanforHap+[Ag}+_]/[Sr2+_](4.3±1.9

mJ).Sinceitsfailuremodeis moreadhesive,thedecrease ofthe ad-hesionenergyforHap+[Ag+_]/[Sr2+_{]canbeattributedtoaweakening}

oftheanchoringofthedeposit.Itis noteworthytomentionthatthe onlydistinguishingcharacteristicsbetweenthesetwocoatingsarethe concentrationandthesizeofthesilvernanoparticles.Itistherefore as-sumedthatthehigherproportionandthefinersizeofthenanoparticles resultingfromthesilvernitratesbothreducethecontactsurfaceand thusanchorageofthecoating.TheAg/Sr-HapandHap+[Ag+_]/[Sr2+_]

coatingsexhibitthesamesilverconcentrations,shape,anddispersion ofthesilvernanoparticlesaswellasstrontiumconcentrations,buttheir crystallinityratiosarefarapart.Ag/Sr-Hapistheleast crystallineof allcoatings,justifyingwhyitsinterfaceisstrongerandwhythefailure modeiscohesive.

4.3. Physicochemicalpropertiesandbiologicalbehavior

Regardingtheeffectof dopingoncellproliferation,itseemsthat cellsproliferatepreferentiallyontheundopedcoatings(578%forHap and 549% for Reference at day 14). Among the doped substances, Hap+[Ag+_]/[Sr2+_{]andAg/Sr-Hapcoatingsarethemostfavorable(549}

and546%respectively),whiletheHap+AgNP+[Sr2+]coatingshowsan intermediatebehavior(209%).Asavalueofcellviabilitybelow70% reportsapotentialcytotoxicmaterial, thiscoatingdoesnotmeetthe

Fig.10. Evolutionofthemechanical proper-tiesofthecoatingswiththecrystallinityratio R1.

requirements.Thismeansthatthecellsproliferatemorefavorablyon thesurfaces wheresilvernanoparticlessizesaresmallanduniformly distributedthanonthosewithlargeragglomerates,withlittleeffectof thesilvercontent.WorkconductedbyAlbersetal.[10]statesthat sil-verharmfulnessismainlycarriedbytheAg+_{ions.Inaddition,ithas}

beenshowninotherstudiesthatthechemicalstabilityofthecoating influencescellproliferation[43].Thisisalsopossiblythecasewith sil-ver,especiallysinceitsdissolutiondependsonthepHofthemedium, whichinturndependsontheamountofOH−_andPO

43−ionsreleased

fromthephosphocalcicmatrix.Indeed,silverisfurtherdissolvedwhen thephosphocalcicmatrixis poorlysolublebecausethepHis main-tainedlongeratlowvalues attheperipheryofthesurface,as demon-stratedwiththeannealeddopedsample.Thiscoatingreleasesless cal-cium,strontium,phosphateandhydroxylionsthantheothersbecause ofitshighcrystallinityandhydroxyapatiteproportion.Butitreleases significantlymore silvereventhough theinitialAg concentrationof thedepositissimilar(0.313±0.007wt.%).Crystallinestatethusfirst determinesthesolubilitydegreeofthecoating,solubilitythatwanbe increasedby alarge exchange surfacewiththe medium,defined by roughnessandporosity. Asevidencedduringthedissolutiontest,the solubilityofdopedcoatingsevolvesinthefollowingorder:Ag/Sr-Hap

>Hap+[Ag+_]/[Sr2+_](≈Hap+AgNP_+[Sr2+_{]).Thisexplainswhy}

Ag/Sr-Hapcoatingdoesnotblockcellproliferation:itslowcrystallinity(57%) combinedwitharelativelyhighporosity(5%)isfavorabletothe disso-lutionofitsmatrixandtotheincreaseofpH,sothatitreleasesonlyvery fewAg+_{ions.ButwhentheCaPmatrixdissolvesquickly,silver}

nanopar-ticlesdonot,graduallyconcentratingthematthesurface.Thefactthat cellularproliferationisnotcompromisedmeanscellsaremoresensitive toAg+_{fluxthantoAgnanoparticles,confirmingtheresultsofAlbers}

etal.[10].FortheHap+[Ag+_]/[Sr2+_{]coating,thesolubilityofwhichis}

intermediate,thereleaseofAg+_{ionsisalittlemorepronounced,butthe}

nanoparticlesofhomogeneousandfinegrainsizedistributionrelease Ag+_{ionsinasufficientlydilutedconcentrationnottoalertthecells.On}

theotherhand,atequalphosphocalcicmatrixdissolutionandbecause ofthelargesizeofthesilvernanoparticlesdespitetheirlow concentra-tion,theAg+_{ionfluxescomingfromthesurfaceofHap+Ag}NP_+[Sr2+_]

arelocallyimportant,alertingthecellsandblockingtheirdivision.The antibacterialefficacyBRishighfortheHap+[Ag+_]/[Sr2+_]sample(5.0

at24h),duetoitshighsilverlevel(0.342wt.%)andfavoredAg+_ion

release.Ag/Sr-HapcoatinghasaBRof2.9withasimilarsilver concen-tration(0.394wt.%)but,becauseofthehighsolubilityofthematrix, itsreleaseisslower.ThefactthatHap+[Ag+_]/[Sr2+_]hasahigher

ef-ficiencythanAg/Sr-HapsuggestsastrongereffectoftheAg+_ionson

bacteriaincomparisonwiththemetallicAg(s)particlespresentonthe surface.Finally,Hap+AgNP+[Sr2+_{]samplehasthelowestvalueofBR}

(1.2),probablyduetoitssilverconcentrationthatis5to6timeslower thanthepreviousones.

5. Conclusions

Inthisstudy,threesilverandstrontium-dopedhydroxyapatite coat-ingsweresynthesizedbyrf-SPS,implementingsuspensionscontaining strontiuminitsionicstate(freeorincorporatedintheHappowder)and silverundermetallicorionicstateanddispersedeitherinthepowder precursororintheliquidmedium.Strontiumandsilverwerechosen inordertopromotemesenchymalcellproliferationwhiledisablingthe developmentofS.aureusresistantbacteria.Suspensionplasmaspraying providesaccesstoaversatilitythatconventionalplasmasprayinglacks. Bychangingtheformulationoftheprecursorasbytheadditionof ni-tratesormetalnanoparticlesinthedispersingmedium,onecandirectly influencecharacteristicssuchasporosity,crystallinityorcomposition. Somebiologicalpropertiesastheantibacterialefficiencyseemstobe mainlydrivenbythesilvercontentandthereleaseextentofthe sil-verions,whilethehMSCaffinity forthesurfaceseemsratherlinked tothechemicalstabilityofthephosphocalcicmatrixandthesizeof thesilvernanoparticles.TheHap+[Ag+_]/[Sr2+_{]coatingistheonethat}

combinesverygoodmechanicalproperties comparedtotheAPS ref-erencecoating,excellentaffinitiesforcells,andbactericidalactivity. Ithas alsocompliantphysicochemicalproperties (crystalline compo-sition,porosity),andapotentialgoodchemical stability.By incorpo-ratingthedopantsinionicformintheprecursor,thefineand homo-geneousdispersionofthedopantsatthemicrometricscalewithinthe apatitestructureisguaranteedanddoesnotcompromisethe mechani-calorbiologicalpropertiesofthephosphocalcicmatrix.Thisstudythus demonstratesthepotentialoftherf-SPStechniqueforthedevelopment ofnanostructured-hydroxyapatitecoatingswithcontrolledantibacterial andanti-inflammatorydoping.

DeclarationofCompetingInterest

Theauthorsdeclarethattheyhavenoknowncompetingfinancial interestsorpersonalrelationshipsthatcouldhaveappearedtoinfluence theworkreportedinthispaper.

Acknowledgements

ThisworkwassupportedbytheAgenceNationaledelaRechercheof FranceunderprojectnamedARCHICAP(ANR-15-CE19–0021)andthe RégionOccitanieinFranceunderprojectnamedREVAMITIC(CLE).

References

[1] ANSM, Etude des facteurs associés aux révisions sur prothèses totales de hanche (PTH) : rôle du mode d ’ancrage (cimentage) et des constituants prothétiques (couple de frottement) dans les révisions chirurgicales, 2015.

[2] Haute Autorité de Santé, Prothèse de hanche ou de genou : diagnostic et prise en charge de l’infection dans le mois suivant l’implantation, (2014) 1–134.

[3] G.A. Fielding, M. Roy, A. Bandyopadhyay, S. Bose, Antibacterial and biological char- acteristics of silver containing and strontium doped plasma sprayed hydroxyapatite coatings, Acta Biomater. 8 (2012) 3144–3152, doi: 10.1016/j.actbio.2012.04.004 . [4] Z. Geng, Z. Cui, Z. Li, S. Zhu, Y. Liang, Y. Liu, X. Li, X. He, X. Yu, R. Wang, X. Yang,

Strontium incorporation to optimize the antibacterial and biological characteristics of silver-substituted hydroxyapatite coating, Mater. Sci. Eng. C 58 (2016) 467–477, doi: 10.1016/j.msec.2015.08.061 .

[5] Elliott, Structure and Chemistry of the Apatites and Other, Elsevier, 1994 .

[6] J. Kolmas, E. Groszyk, D. Kwiatkowska-Rózycka, Substituted hydroxyapatites with antibacterial properties, Biomed. Res. Int. (2014) 2014, doi: 10.1155/2014/178123 . [7] S. Eto, S. Kawano, S. Someya, H. Miyamoto, M. Sonohata, M. Mawatari, First Clinical Experience With Thermal-Sprayed Silver Oxide-Containing Hydroxyapatite Coating Implant, J. Arthroplasty. 31 (2016) 1498–1503, doi: 10.1016/j.arth.2015.12.034 . [8] M. Trop, M. Novak, S. Rodl, B. Hellbom, W. Kroell, W. Goessler, Silver-

coated dressing acticoat caused raised liver enzymes and argyria-like symp- toms in burn patient, J. Trauma Inj. Infect. Crit. Care 60 (2006) 648–652, doi: 10.1097/01.ta.0000208126.22089.b6 .

[9] M.A. Hollinger, Toxicological aspects of topical silver pharmaceuticals, Crit. Rev. Toxicol. 26 (1996) 255–260, doi: 10.3109/10408449609012524 .

[10] C.E. Albers, W. Hofstetter, K.A. Siebenrock, R. Landmann, F.M. Klenke, In vitro cy- totoxicity of silver nanoparticles on osteoblasts and osteoclasts at antibacterial con- centrations, Nanotoxicology 7 (2013) 30–36, doi: 10.3109/17435390.2011.626538 . [11] M. Š upová, Substituted hydroxyapatites for biomedical applications: a review, Ce-

ram. Int. 41 (2015) 9203–9231, doi: 10.1016/j.ceramint.2015.03.316 .

[12]E.S. Thian , P.N. Lim , Z. Shi , B.Y. Tay , K.G. Neoh , Silver-doped apatite as a bioac- tive and an antimicrobial bone material, Key Eng. Mater. 493–494 (2012) 27–30 doi:10.4028/www.scientific.net/KEM.493-494.27 .

[13] C. Capuccini, P. Torricelli, F. Sima, E. Boanini, C. Ristoscu, B. Bracci, G. Socol, Strontium-substituted hydroxyapatite coatings synthesized by pulsed-laser deposi- tion : in vitro osteoblast and osteoclast response, Acta Biomater. 4 (2008) 1885– 1893, doi: 10.1016/j.actbio.2008.05.005 .

[14] C. Lindahl, S. Pujari-Palmer, A. Hoess, M. Ott, H. Engqvist, W. Xia, The influence of Sr content in calcium phosphate coatings, Mater. Sci. Eng. C 53 (2015) 322–330, doi: 10.1016/j.msec.2015.04.015 .

[15] D.S. Brauer, N. Karpukhina, G. Kedia, A. Bhat, R.V. Law, I. Radecka, R.G. Hill, D.S. Brauer, Bactericidal strontium-releasing injectable bone cements based on bioactive glasses, J. R. Soc. Interface (2012) 1–8, doi: 10.1098/rsif.2012.0647 . [16] J.T.B. Ratnayake, M. Mucalo, G.J. Dias, Substituted hydroxyapatites for bone regen-

eration : a review of current trends, J. Biomed. Mater. Res. Part B (2017) 1285–1299, doi: 10.1002/jbm.b.33651 .

[17] H. Qiao, G. Song, Y. Huang, H. Yang, S. Han, Si, Sr, Ag co-doped hydroxyap- atite/TiO2 coating: enhancement of its antibacterial activity and osteoinductivity, RSC Adv. 9 (2019) 13348–13364, doi: 10.1039/c9ra01168d .

[18] Y. Huang, X. Zhang, H. Zhang, H. Qiao, X. Zhang, T. Jia, S. Han, Y. Gao, H. Xiao, H. Yang, Fabrication of silver- and strontium-doped hydroxyapatite/TiO2 nanotube bilayer coatings for enhancing bactericidal effect and osteoinductivity, Ceram. Int. 43 (2016) 992–1007, doi: 10.1016/j.ceramint.2016.10.031 .

[19] International standard ISO_13779, Implants for surgery — Hydroxyapatite - Part 2: Coatings of Hydroxyapatite, (2008).

[20] International standard ISO_13779, Implants for surgery — Hydroxyapatite - Part 3: Chemical Analysis and Characterization of Crystallinity and Phase Purity, (2008). [21]T.J. Webster , C. Ergun , R.H. Doremus , R.W. Siegel , R. Bizios , Enhanced functions of

osteoblasts on nanophase ceramics, Biomaterials 21 (2000) 1803–1810 .

[22] M. Roy, A. Bandyopadhyay, S. Bose, Induction plasma sprayed Sr and Mg doped nano hydroxyapatite coatings on Ti for bone implant, J. Biomed. Mater. Res. Part B Appl. Biomater. 99 (2011) 258–265 B, doi: 10.1002/jbm.b.31893 .

[23] M. Roy, G.A. Fielding, H. Beyenal, A. Bandyopadhyay, S. Bose, Mechani- cal, in vitro antimicrobial, and biological properties of plasma-sprayed silver- doped hydroxyapatite coating, ACS Appl. Mater. Interfaces 4 (2012) 1341–1349, doi: 10.1021/am201610q .

[24] P. Fauchais, Suspension and solution plasma spraying, J. Phys. D Appl. Phys. 46 (2013) 14pp, doi: 10.1088/0022-3727/46/22/224015 .

[25] S. Kannan, J.M.F. Ferreira, Synthesis and thermal stability of hydroxyapatite- 𝛽-tricalcium phosphate composites with cosubstituted sodium, magnesium, and flu- orine, Chem. Mater. 18 (2006) 198–203, doi: 10.1021/cm051966i .

[26]P. Ducheyne , Comprehensive Materials, Elsevier Science„ 2011 .

[27] R.B. Heimann, Plasma-sprayed hydroxylapatite-based coatings: chemical, mechani- cal, microstructural, and biomedical properties, J. Therm. Spray Technol. 25 (2016) 827–850, doi: 10.1007/s11666-016-0421-9 .

[28] M. Roy, A. Bandyopadhyay, S. Bose, Induction plasma sprayed nano hydroxyapatite coatings on titanium for orthopaedic and dental implants, Surf. Coatings Technol. 205 (2011) 2785–2792, doi: 10.1016/j.surfcoat.2010.10.042 .

[29] E. Bouyer, F. Gitzhofer, M.I. Boulos, Suspension plasma spraying for hydroxyapatite powder preparation by RF plasma, IEEE Trans. Plasma Sci. 25 (1997) 1066–1072, doi: 10.1109/27.649627 .

[30] R. d’Haese, L. Pawlowski, M. Bigan, R. Jaworski, M. Martel, Phase evolution of hydroxapatite coatings suspension plasma sprayed using variable parame- ters in simulated body fluid, Surf. Coatings Technol. 204 (2010) 1236–1246, doi: 10.1016/j.surfcoat.2009.10.022 .

[31] K. Geels, D.B. Fowler, W.-.U. Kopp, M. Rückert, Material/Preparation Ta- bles —Methods C-01/T-01 to C-68/T-68, in: Metallographic and Materialographic Specimen Preparation, Light Microscopy, Image Analysis, and Hardness Testing, 2006, pp. 223–226. https://books.google.cl/books?id = oaehZy3Vo1kC .

[32]A.A. Roche , A.K. Behme , J.S. Solomon , A three-point flexure test configuration for improved sensitivity to metal, J. Adhes. Adhes. (1982) 249–254 .

[33] International standard ISO 14679:1997, Adhésifs - Détermination des caractéris- tiques d’adhésion par une méthode de flexion à trois points, (1997).

[34] L. Berzina-Cimdina, N. Borodajenko, Research of calcium phosphates using fourier transform infrared spectroscopy, in: Infrared Spectroscopy - Materials Science, En- gineering and Technology, 2012, pp. 123–148, doi: 10.5772/36942 .

[35] W.K. Yeung, G.C. Reilly, A. Matthews, A. Yerokhin, In vitro biological response of plasma electrolytically oxidized and plasma-sprayed hydroxyapatite coatings on Ti-6Al-4V alloy, J. Biomed. Mater. Res. Part B Appl. Biomater. 101 (2013) 939–949 B, doi: 10.1002/jbm.b.32899 .

[36] H.W. Kim, Y.H. Koh, Y.M. Kong, J.G. Kang, H.E. Kim, Strontium substi- tuted calcium phosphate biphasic ceramics obtained by a powder pre- cipitation method, J. Mater. Sci. Mater. Med. 15 (2004) 1129–1134, doi: 10.1023/B:JMSM.0000046395.76435.60 .

[37] L. Leroux, J.L. Lacout, Preparation of calcium strontium hydroxyapatites by a new route involving calcium phosphate cements, J. Mater. Res. 16 (2001) 171–178, doi: 10.1557/JMR.2001.0028 .

[38]C. Combes , C. Rey , Biomatériaux à base de phosphates de calcium, Tech. l’ingénieur (2013) 1–25 n4950 .

[39] S.V. Dorozhkin, Calcium orthophosphates – Occurrence, properties, biomineraliza- tion, pathological calcification and biomimetic applications, Biomatter 1 (2) (2011) 121–164, doi: 10.4161/biom.18790 .

[40] D.R. Lide, Handbook of Chemistry and Physics, 2004. doi:10.1016/0022-2860(92)85083-s.

[41] A. Fihri, C. Len, R.S. Varma, A. Solhy, Hydroxyapatite: a review of syntheses, struc- ture and applications in heterogeneous catalysis, Coord. Chem. Rev. 347 (2017) 48–76, doi: 10.1016/j.ccr.2017.06.009 .

[42]G. Britain , J. Metauforschung , Fracture mechanics studies of thermal mismatch using a four-point bending specimen, Acta Metall. Mater 40 (1992) S345–S353 .

[43] G. Costa Machado, E. García-Tuñón, R.V. Bell, M. Alini, E. Saiz, M. Peroglio, Calcium phosphate substrates with emulsion-derived roughness: processing, characterisation and interaction with human mesenchymal stem cells, J. Eur. Ceram. Soc. 38 (2018) 949–961, doi: 10.1016/j.jeurceramsoc.2017.06.043 .