ARTICLE IN PRESS

(1)

Availableonlineatwww.sciencedirect.com

j our na l h o me p ag e:w w w . i n t l . e l s e v i e r h e a l t h . c o m / j o u r n a l s / d e m a

Comparative mechanical behavior of dentin enamel and dentin ceramic junctions assessed by speckle

interferometry (SI)

Michel Fages

^a^,∗

, Pierre Slangen

^b

, Jacques Raynal

^a

, Stephane Corn

^b

, Kinga Turzo

^c

, Jacques Margerit

^a

, Frédéric J. Cuisinier

^a

aEA4203,DepartmentofBiologicSciences,UniversityofMontpellierI,Montpellier,France

bEcoledesMinesd’Alès,I.L.O.A.,Alès,France

cDepartmentofProsthodonticandOralBiology,UniversityofSzegedDentalSchool,Szeged,Hungary

a r t i c l e i n f o

Articlehistory:

Received23October2011

Receivedinrevisedform3May2012 Accepted18May2012

Availableonlinexxx

Keywords:

Dentinenameljunction(DEJ) Dentinceramicjunction(DCJ) Speckleinterferometry(SI) Toothcrown

Ceramiccrown

a bs t r a c t

Objective.Thedentin–enameljunction(DEJ)playsacrucialroleindentalbiomechanics;however,littleisknownaboutitsstructureandmechanicalbehavior.Nevertheless,naturalteeth areanecessarymodelforprostheticcrowns.ThemechanicalbehaviorofthenaturalDEJ andthedentinceramicjunction(DCJ)manufacturedwithaCAD-CAMsystemarecompared.

Methods.Thereferencesamplesundergonomodiﬁcation,whiletheexperimentalsamples weredrilledtoreceiveacementedfeldspathicceramiccrown.Longitudinallycutsamples wereusedtoachieveaplanarobjectobservationandtolook“inside”thetooth.Acom- pleteapparatusenablingthestudyofthecompressivemechanicalbehavioroftheinvolved toothbyanon-contactlaserspeckleinterferometry(SI)wasdevelopedtoallownanometric displacementstobetrackedduringthecompressiontest.

Results.ItisobservedthattheDEJactedasacriticalzoneaccommodatingthemovement betweendentinandenamel.Asmoothtransitionoccursbetweendentinandenamel.In themodeledprosthetic,thesamekindofaccommodationeffectsalsooccurs,butwitha steepertransitionslopebetweendentinandceramic.

Signiﬁcance. On thenatural tooth, thestress accommodation arises froma differential behaviorbetweenenamelanddentinfromtheDEJ.Intheceramiccrown,thecemented dentin–ceramicjunctionshouldplaythisrole.Thisstudydemonstratesthepossiblereal- izationofprostheticcrownreconstructionsapproachingbiomechanicalbehaviors.

1. Introduction

Thedentin–enameljunction(DEJ)inteethisthezonebetween twodistinctcalciﬁedtissueswithverydifferentbiomechan- icalproperties:enamel and dentin[1]. Enamelishard and brittleandenvelopsthesofterdentin.Theenamelanddentin worktogetherduringthemanyloadcyclesexperiencedbythe

∗ Correspondingauthorat:11AvenueCelestinArnaud,34110LaPeyrade,France.Tel.:+33684855715;fax:+33467486092.

E-mailaddress:mifages@wanadoo.fr(M.Fages).

toothoveritslifetime.Generally,interfacesbetweenmaterials withdissimilarelasticandmechanicalpropertiesrepresent

“weak links” ina structure; however, the DEJacts to suc- cessfullytransferappliedloads(e.g.,masticatoryorimpact) from the enamelto thedentinand inhibits enamelcracks frompropagatingintothedentinandcausingtoothfracture [2,3].

http://dx.doi.org/10.1016/j.dental.2012.05.006

(2)

Pleasecitethisarticleinpressas:FagesM,etal.Comparativemechanicalbehaviorofdentinenamelanddentinceramicjunctionsassessedby attheDEJ,thusexplainingwhysofewcrackingeventscause

delaminationwhentheyimpingeontheDEJ.Zaslanskyetal.

[8,9] highlighted the importance ofthe DEJ as the binding interfacebetweenenamelanddentin.Theyhaveshownthat adjacenttotheDEJisa200–300mm-thickzoneofdentinofa muchlowerstiffness(compressionelasticmodulus)thanthe bulkofthedentininthetooth.

Restorations that are all ceramic require proper adhesive bonding on the dentin to achieve their required life expectancy. All-ceramic restorations are made with felds- pathicor zirconiaceramics.Thestrongestceramicshave a fracturetoughnessofatleast3.0MPam^1/2[10],whichisrel- ativelyclosetotheenamelfracturetoughnessof1.3MPam^1/2, inadirectionperpendiculartotheenamelrods[8].Neverthe- less,fracturesoftheceramicpartofall-ceramiccrownsare difﬁculttoprevent, andcrack growth isasigniﬁcant prob- lem[11].Thisphenomenoncanbeexplainedbytheabsence ofastressaccommodationzone.Thenaturalstressaccom- modationzoneof200–300␮m-thickdentinhasamuchlower stiffnessthanthebulkofthedentincore[8].

Bondingagentsmust beselectedvery carefullybecause theydeterminenotonlytheadhesionbutalsotheultimate strengthoffull-ceramiccrowns[12–14];therefore,itisimpor- tant to compare the mechanical behavior ofnatural teeth andoftheall-ceramiccrowncementedondentin.Insteadof

“cementjoint”,wewillusetheterm“dentin–ceramicjunction”

(DCJ).

We applied compressive forces representative of those occurringintheoralcavityonnaturalteethandall-ceramic crowns,andwedeterminetherelativemovementofenamel anddentin,orceramiccrownanddentin,respectively.

2. Materials and methods

2.1. Naturalteeth

Intactlowerﬁrstpremolarsfreeofcarieswerestoredinphys- iological serum after havingbeen extracted as part ofthe routineorthodontic treatmentofyounghealthy adolescent patients(aged<18).Fivesetsoftwosampleseach(onenat- uraltoothandoneprosthetictooth)wereamassed.Rightand leftpremolarsfromthesamepatientwereused.Onewaskept intact,andtheotherwaspreparedtoreceivetheprosthetic crown.

prostheticcrowns[15].Thewholecloningprocessispresented inFig.1.

The Vita MarkII^® (Vita Zahnfabrik, Bad Säckingen, Switzerland) ceramic blocks of albite-enriched feldspathic ceramicwereused.Theirabrasioncoefﬁcientisclosetothatof naturaldentalenamel.Aftermilling,theextradoswereglazed (Azkent^®,VitaZahnfabrik,BadSäckingen,Switzerland).

Thecrownswerecementedontothepreparedteethusing RelyxUnicem^®adhesivecement(3MESPEDentalDivision,St.

Paul,MN,USA)followingthestandardclinicalprotocolofillu- minationofeachsideofthecrown for4sat3000mW/cm² withaSwissmasterLight^®lamp(E.M.S.,Nyons,Switzerland).

2.3. Specimenpreparation

Afterextraction,theteethweredisinfectedandstoredinphys- iological serum with traces of chloroform.The teeth were longitudinally cutinthe vestibular–lingualorientation,and oneofthetworesultingpartswasremovedwithadiamond disc.Longitudinalcutshavebeenusedtoallowplanarobser- vation and toappreciatethe differentbehaviors insidethe toothofthenaturalDEJandoftheDCJinterfaces.Thetooth wasthengluedintothesampleholderwithalayerofAraldite^® (HunstmanAdvancedMaterials,TheWoodlands,Texas,USA).

Thesampleholdershavebeencastinchromiumcobaltusing the imprintofaroot. Themechanicalstabilityofthespec- imenholderswasvalidatedbyaspeckleinterferometry(SI) experiment.

2.4. Loadingsystem

The compression test device fulﬁlls the high sensitivityof speckleinterferometryandcopeswiththerigidbodymotions ofthewholesystem.Thesampletoothcementedinthesam- pleholderwasplacedundertheforcetransducer(Model31, HoneywellInternational,Morriston,NJ,USA).Thismid-range precision miniature load cell isslowly translated vertically bythemotor(M-235^®,PI.Karlsruhe,Germany).Thesystem cangenerateaforce-drivendisplacement(C-862MercuryPI, Karlsruhe, Germany) or simply a user’s displacement.The compression apparatus communicates with the computer throughaNIUSB-6251port(NationalInstruments,Austin,TX, USA)andisinterfacedwithanin-houseLabViewprogram.The entiremechanical systemwas boltedonto theholographic tabletop(Newport,Irvine,CA,USA).Verysmalldisplacement steps,assmallas1.6nm,cantheoreticallybeachieved.The

(3)

Fig.1–Cloningprocess:theintactnaturalcrownsample(A),opticalprint(B),shaping(C),opticalprintofthesecondtooth preparedtoreceivetheprostheticclone(D),adaptationoftheshapingonthetoothprepared(E),CADﬁnished,CAMready (F),ceramicblockinthemillingunit(G),theprostheticclonemilled(H),theprostheticclonecementedonthetoothprepared (I),thenaturaltooth(J),verticalcuts(K),prostheticclone(L)naturaltooth.

forcecanbeappliedtothetoothdirectlywiththeforcetrans- ducerorthrougharelayrod.Forcehasalwaysbeenapplied tothesamepartofthelingualcusp.PreliminarytestingbySI showedgoodperformancesofthemechanicalset-upandno spuriousdisplacements.

2.5. SIapparatus

Theopticalset-upwaspreviouslypresentedindetail[16].The frequency-doubledYAGlaseremits50mWat532nmwave- lengthinthegreenrange.Thelaserbeamistheninjectedin aCOTS(commerciallyoff-the-shelf)system(CanadianInstru- ments,Nottingham,UK)offeringinjection,variableintensity couplingintheoutputﬁbers,andphaseshifting.

Therearetwoinputfibers:oneforinjectionandonefor detectionofthereflectedsignalattheoutputfiberinterfaces.

Partsoftheoutputfibersarebaredandwrappedaroundpiezo- electrictransducers.Thephaseshiftisappliedoneitheror bothofthetwooutputfibersbyapplyingavoltageatthePZT andthusgeneratingatinyextensionofthefiber.Thesystem isprotectedfromthermalandmechanicaleffectsbyaplastic boxandiseasilybreadboardable.Phaseshiftswerecalibrated usingcommonprocedures[16].

Asensitiveopticalﬁberin-planeinterferometerhasbeen designedwithsensitivityvectorSv.Horizontalsensitivityis achieved.Two symmetricalbeamsproducean interference signalonthewholeobjectandthenilluminatethesamples.

Theobjectsareopticallyrough-renderedbywhitepowder.An XC70CCDcamera(Sony,Tokyo,Japan)recordsthesamplesur- faceunderloadingforces.Theimagesarethenstoredinlive memoryoronthecomputer’sharddisk.

Theimageprocessingwasperformedusingtheappropri- ate software in LabView, and the results are presented as vectormapsorfalsecolormaps.Isodisplacementmapscan bevisualizedinrealtimeduringtheloadingofthesample.

Thesystemcanhandleareasrangingfrom5mm×5mmto 1m×1mwhenusingtheappropriatetypeoflaserandload- ingsystem.Themeasurementuncertaintyisapproximately 50nm, which iscommon forinterferometricmeasurement inacontrolledenvironment.Thedisplacementresolutionis

approximately10nm,whilethespatialresolutionisdirectly linkedtothemagniﬁcationoftheobjectsceneontheimage sensor(1280×900pixels).

A “4-buckets” phase shifting algorithm leads to phase variationsduringthecompressiontest[17].Duringthetest, the initialphase statewas memory-resident and real-time subtractedfromthe currentstate.Sometimesthereference statewasalsorefreshedbecauseforsomeloadingsteps,the number offringescan betoohigh,and theresultingnoise wouldinterferewiththeinterpretationoftheresultingfringes.

Interferogramsshowthein-planedisplacementsfromphase shiftingspeckleinterferometry.Highestqualityimageswere storedontheharddiskandoverlaidwithupperjawposition andloadvalue(N).

2.6. Typicalexperiments

Specimens were white powdered using Eutest 3Developer Castolin Eutecticpowder (Castolin, Lausanne, Switzerland) togenerateauniformdiffusingsurfaceand toavoiddiffer- entmodulationsbetweendentinandenamelorceramicand dentin.Thesameloadswereappliedtothenaturaltoothand theprostheticcrownsamplestoallowcomparisonsoftheir respectivecompressivebehaviors.

Thestartingloadwasapproximately0N.Thecompression was increasedstepwise,performing discretedisplacements of the transducertip (one stepis approximately 1.6nmof Y-displacement).Therefore,agreaternumberofstepscorre- spondstoahighercompression.Thesamplecanbeloaded orunloaded.Theloadingrangeappliedtothedifferentsam- pleswasbetween5Nand120N.TheCCDcamerarecordsat thesamplesurfacetheinterferencesofthetwoillumination beamscomingfromthetwooutputopticalﬁbers.Livefringes aredisplayedbetweenareferencestateandthecurrentload state.Betweenminimalandmaximalloading,differentphase mapsarerecordedandstoredinthememory.

Themechanicaldeformationsarecomputedfromthedis- placementmapsgeneratedfromthephasedifferencemaps.

Speckle interferometry is a relative displacement mea- surement. The maximum range between two successive

(4)

Pleasecitethisarticleinpressas:FagesM,etal.Comparativemechanicalbehaviorofdentinenamelanddentinceramicjunctionsassessedby measurementsisapproximately20␮m.Inourexperiments,

weneedtorecord smallerstepsassomemechanicalnoise appears.Therefore, newdisplacement references(zerodis- placementreset)arerecordedduringthetest.

2.7. Displacementcalculations

Different operation modes of SI are commonly used, e.g., subtraction-mode, time-averaged SI, and double-pulsed SI [16].Inthiswork,wefocusonsubtraction-modeSI,ormore speciﬁcally, onphase-shifting SI, which is mainlyused for staticdeformationmeasurements.

Combining the primary interference pattern phase changes between the recordings yields new secondary interferencefringes(alsocalledcorrelationfringes).

Thevariableϕsdenotesthestartphase(alsocalledspeckle phase)attheinitialstateoftheobject.Thevariableϕrepre- sentsthephasechangebetweentwostates.

Thesespeckleinterferogramscanbesubtractedandlead tothefollowingequationforthesecondaryinterferencefringe pattern,assumingperfectspatialcorrelationbetweenthetwo primaryspecklepatterns:

I1−I2=2

IrIo(cos(ϕs+ϕ)−cos(ϕs))

Currently,noiselimitstheaccuracyofintensitysubtraction SItoapproximately15nm.Regardingthein-planesensitivity ofthesetup,theangleoftheimpingingbeamisapproxi- mately30^◦anddeterminesthecorrespondencebetweenthe phasegraylevelvariationandthein-planedisplacementux.

Inourinterferometer,thewavelengthofthelaseris532nm, anditsrelationshipwiththein-planedisplacementisgivenby thefollowingequation:

ux= 4sinϕx

The resulting variationof onegray level ofphase ux is approximately2.08nm.

3. Results

Five setsoftwosampleseach(naturaltoothand prosthetic crown)wereproduced.Theexperimentalprotocolwastested andvalidatedwithfourofthem.Presentedresultscorrespond solely tothe ﬁfthsample(for natural tooth and prosthetic clone).

Theinterferometricimagesdistinctlyshow thebehavior of the samplesand conﬁrm the quality ofthe mechanical apparatusandthe integrityofthededicatedsampleholder (Figs.2and3).

AnaturaltoothispresentedinFig.2.Figurecaptionsalso presentthescreenshotnumber(orimagenumber),theforce appliedtothesample,theforcedifferencefromthereference state,andforcefromthereferencestate(orphasereset).

Fig.2Adisplaysthesampleunderwhitelightbeforepaint- ing.Fig.2B–DisSIimagesrecordedatdifferentcompression levels and present typical fringes. Fringesoccur when the displacement inducesan optical phaseof 360–0^◦. The displacement inthe Xdirectioniscomputed from theoptical phase.

InFig.2B–D,differentcontinuousgraylinesdemarcatethe DEJ(redarrows)correspondingtotheimagetakeninwhite

Fig.3–Prostheticcrownsampleunderdifferentloads.(A)Whitelightimages.(B)Screenshotno.08:force36.78N,F:

2.57N(newref:41.01N).(C)Screenshotno.15:force77.56N,F:5.29N(newref:82.85N).(D)Screenshotno.33:force 64.09N,F:18.76N(newref:41.01N).

(5)

Fig.4–Naturaltoothbehaviorataloadof38.65N,F=1.74N,newref:40.39N.(A)SIimage:thesinglewhiteline

correspondstotheregionofinterest(ROI)usedforthecalculationofdisplacement.(B)Displacementcurvealongthewhite line,displacementsinnmversuspositioninpixels.Stepisabout43nm.

light(2A).TheDEJisclearlyvisiblewhentheappliedforce reaches35.5N.TheDEJisalwaysmorevisibleinfrontofthe loadingpoint.

Fig.3Apresentsaprosthetic crownsampleunderwhite lightbeforepainting,andFig.3B–Darerecordedatdifferent compression levels and present typical fringesatdifferent loads.

TheDCJappearsfrom38NandisclearlyvisibleinFig.3B–D (redarrows).Fortheprostheticcrowns,itappearsasacontin- uousgraylinecorrespondingtotheimagetakeninwhitelight (Fig.3A).

TheSIimagesdisplayedinFig. 2B–Dshowthatthenat- uralenamelcapmovesindependentlyfromthedentin.This differenceisclearlydelimitedbyalinecorrespondingtothe anatomicallocationofthedentin–enameljunction(DEJ).For theprostheticcrown,intheSIimages(Fig.3B–D)thesame kind ofshift occursatthe cement junctionofthe ceramic crownwiththedentin.TheDEJislesswellmarkedthanthe DCJduetothesmallerassociateddisplacements.Largevaria- tionsinintensity(Fig.3CandD)correspondtofringesresulting fromrigidbodymotion(inplanerotationprojectedontothe sensitivityvector).

Thehighestloadingvaluesenablingthedistinctionofthe interfacezonewere117.4Nfortheprostheticcrownand82.5N forthenaturaltooth.Beyond120N,allsamplesbehavedlike rigidbodies.Around200N,somesamplesweredestroyeddue tofracturesofthe brittlematerials.From allofthescreen- shots,differentimageswerechosenforuseincomputingthe displacementmaps.

In Fig. 4A, from left to right, we denote the transition betweenlightgrayanddarkgraycorrespondingtotheregion betweenthedentinandtheenamel.Onehorizontalwhiteline hasbeen deﬁnedasthe regionofinterest(ROI).In Fig.4B, the curve representsthe displacement change innanome- tersversusthepositionalongthewhitestraightlineshown inFig.4A.Therelativedisplacementbetweenthedentinand

the enamel is 52nm for loads between 38.6N and 40.4N (F=1.8N).

InFig.4B,thegraylevelarerisingattheendofthecurveas itisclosetotherightedgeofthetooth,itisclearlyvisibleon thenativeimagewithhighermagniﬁcation.Thisisgenerated bythesmoothinplanetiltofthepalatinecuspidfollowingthe upperdisplacementoftheupperjawrodfromthemechanical testingastheforcedecreasesfrom40.39Nto38.65N.

InFig.5A,infrontofthecuspsubjectedtotheload,the delineationmadebytheDEJisvisible,andthegraylevelsdiffer becauseofadifferentaccommodation.

Tocalculatethedisplacementvalue,sixredequalparal- lellinesweredefinedperpendicularlyacrosstheDEJ.Thered linesareseparatedby1pixelonetotheother,andsoappear asaboldredlineonthefigures.Thedisplacementistheaver- agevalueofthesix-stackedprofiles.InFig.5B,theblackcurve isthemeanvaluefittingofthebluecurvevaluesalongthe sixpaths,andthebluecurveisoneofthesixdisplacement curves.Themeanrelativedisplacementbetweenthedentin andtheenamelisabout20nmforloadsbetween39.21Nand 40.39N(F=1.18N).

Thesameanalyticalprocesswasappliedfortheprosthetic crowns(Figs.6–8).

In Fig.6A, onehorizontalblack linewas deﬁnedas the regionofinterest,asthetechnicalnoisewaslessprominent than forthe natural tooth.The area of interest is located inafringeinfrontofthestresszonefromleft toright.We denotethetransitionbetweenlightgrayanddarkgraycorre- spondingtotheregionbetweenthedentinandtheceramic cap.InFig.6B,thecurverepresentsthedisplacementchange innanometersversusthepositionalong thewhite straight lineshowninFig.6A.Therelativedisplacementbetweenthe dentinandtheceramiccapis43nmforloadsbetween82.8N and76.3N(F=6.5N).

InFig.7,toevaluatethedisplacementvalues,sixequalpar- allelstraightpathshavebeendeﬁnedintheregionofinterest

(6)

Pleasecitethisarticleinpressas:FagesM,etal.Comparativemechanicalbehaviorofdentinenamelanddentinceramicjunctionsassessedby Fig.5–Naturaltoothbehaviorataloadof39.21N,F=1.18N,newref:40.39N.(A)SIimage:sixredequalparallellines, 1pixelseparated,aredefinedacrosstheDEJ.(B)Displacementcurves:bluecurve:displacementvaluesalongoneofthesix redlines.Blackboldcurveisthemeanvaluefittingofthe6bluecurves.(Forinterpretationofthereferencestocolorinthis figurelegend,thereaderisreferredtothewebversionofthearticle.)

acrosstheDCJinthepalatinezoneinfrontoftheloadingpoint (Fig.7A).ThesixredlinesaredrawnperpendicularlytotheDCJ toaccuratelymeasurethedisplacementvariationacrossthe DCJ.Thisismuchappropriatethanthesinglehorizontalblack lineforFig.6astheredlinesenabletherealcomparisonfrom onesidetotheotherperpendiculartotheDCJ.

InFig. 7B,the displacement isdisplayedas the average valueofthesixstackedproﬁles.Theblackcurveisthemedian valueﬁttingofthebluecurvevalues.Inthiscase,therelative displacementisapproximately95nmforloadsbetween82.8N and72.8N(F=10N).ItshowsclearlytheeffectoftheDCJact- ingasanaccommodationareafortheapplieddisplacement, butwithasharperslopethanthenaturalDEJ.

TheSIimageinFig.8wasselectedtorepresentthecharac- teristicisodisplacementmapofthediscontinuityzonelocated infrontoftheloadingpointforaceramiccrown.Acolored andzoomedimageisalsopresented.Thefringeshiftinthe

region ofinterest clearlydelineates amechanical interface betweentheceramicandthedentin,whichareseparatedby thecement.

Fig.8presentsthecharacteristicfringesobtainedforahuge load difference(30.07N) andthe importantnoise occurring duetosomespeckledecorrelationeffect.ResultsfromFig.8 showitismandatorytomakestepreferencesenablingsmaller displacementloadandthusverylowdecorrelationnoise.

Fig.9showsthenaturaltoothwiththesamechargethan prosthetictooth(Fig.8)andallowstocompareDEJandDCJ.

4. Discussion

TheimportanceoftheDEJasaninterfacebindingtheenamel and dentinsurfaceshaslong beenrecognized(Tylman [2]).

OurunderstandingoftheroleandthelocationoftheDEJhas

Fig.6–Prostheticcrownbehaviorataloadof76.34N,F=6.51N,newref:82.85N.(A)SIimage:asingleblackline correspondstotheregionofinterest(ROI)usedfordisplacementcalculation.(B)Displacementcurvealongtheblackline, displacementsinnmversuspositioninpixels.Stepisabout43nm.

(7)

Fig.7–Prostheticcrownbehaviorunderaloadof72.85N,F=10N,newref:82.85N.(A)SIimage:sixredequalparallel lines,1pixelseparated,aredeﬁnedacrosstheDEJ.(B)Displacementcurves:bluecurve:displacementvaluesalongoneof thesixredlines.Blackboldcurveisthemeanvalueﬁttingofthe6bluecurves.Thedisplacementisapproximately95nm.

(Forinterpretationofthereferencestocolorinthisﬁgurelegend,thereaderisreferredtothewebversionofthearticle.)

evolvedwithongoingresearch.Theuniquemechanicalprop- ertiesofthiszonewerefirstrecognizedfrommicro-hardness profiles(Craigetal.[18,19]).WangandWeiner[7]measuredthe strainacrossthiszonewhenacompressiveloadwasapplied andsuggestedthatthisisanimportantworkingpartofthe toothduringmastication.Anasymmetrybetweenenameland dentinwasnotedpreviously(Wood etal. [20])but was not quantified.Bechtleetal.explainedthephenomenonofcrack arrestattheDEJusingtheelasticmodulusmismatchbetween thedentinandtheenamel[21].

It was found (Zaslansky [9]) using SI that a compressive load applied to the tip of the main cusp of a human premolar caused the entire enamel cap to

move essentially as a stiff body, tilting toward the buccal surface.

Recently,Barak etal. [22]highlighted theimportanceof enamelinawhole-toothdemonstrationthroughaﬁniteele- mentmodelstudyandvalidatedtheirﬁndingsbyametrology method.Chattahetal.[23]showedthattheenamelcapina minipig animalmodeliscapableofdeformingandrotating atloadsaslowas16N.Zaslanskyetal.[9]showedthatthe enamelcapofanisolatedhumanpremolardidnotdeformor rotateatloadslowerthan80N.Humanenamelisstifferthan thatoftheminipig,andthecuspsdonotdeformorcrackuntil highloadsarereached.However,inbothcases,theaimwasto preservethefunctionalityofthegrindingsurfacesoverlong

Fig.8–SIimagesample“prostheticcrown”anddisplacementsshowninacontourcoloredmap.Behaviorforaloadof 52.78N,F=30.07N,newref:82.85N.

(8)

Pleasecitethisarticleinpressas:FagesM,etal.Comparativemechanicalbehaviorofdentinenamelanddentinceramicjunctionsassessedby Fig.9–SIimagesample“naturaltooth”.Behaviorfora

loadof52.3N,F=6.5N,newref:58.8N.

periodsoftime.Theuseoftheseopposingstrategiestoachieve thesameendhasbeenattributedtophylogeneticdifferences inmasticatoryfunction.[24,25].Theminipigmolariscapable ofdeformingandrotatingatlowloads,andtheintrinsicreac- tionofthecrowntoeccentricloadingiscomplementedand evenenhancedbythestructuressupportingthetooth[23].

Inourstudy,weconsideredthehalf-cutcrownbecausethis configurationhasthemajoradvantageofobserving thein- planebehavioroftheDEJandtheDCJinanddoingsofrom eitherfarfromorclosetotheloadingzone.Zaslanskyworked withparallelepipedcutsfrom premolars[9], whileChattah etal.[23–25]studiedwholeteeth.OnSIimages,theinterfaces appeardistinctlyalongtheirentirelengths,thusshowingtheir completeloading(from35.5NfortheDEJonthenaturaltooth andfrom 36Nfortheprosthetic crown).Thefirstinterpre- tationconfirms thatthe enamel bulkand the ceramiccap willmoveslightlyunderloadingasrigidbodies.However,the mechanicalresponseisdifferentforthesameappliedload- ingforce,andthedisplacementislargerforceramicthanfor enamel.ThusinFig.8,ataloadof52.7N,theSIimageclearly showsarelativedisplacementbetweenenamelanddentin.In Fig.9,atasimilarload(52.3N),theimageidentifiestheDEJand thedifferingbehaviorsoftheenamelanddentin.However,it doesnotshowarelativedisplacementashighasinFig.8.This displacementisconfirmedinSIpicturesinFigs.6and7,with valuesof83nmand95nm,respectively.Thisbehaviordemon- stratestheaccommodationstrengthoftheDCJandconfirms itsprotectiverolefortheceramiccaps.

Moreover, we emphasize that the displacement of the crown,whetherenamelorceramic,isdifferentifthemea- surementareaisfarfromorclosetotheloadingzone.Ineach case,thelargerdisplacementislocatedoppositetotheload.

Thegraylevelsmapsalsoshowadisplacementoftheopposite cuspidbutofsmallermagnitude.

For the natural tooth, Zaslansky writes, “the asymme- try in stiffness between the buccal and lingual sides may

tainload(117.4Nfortheprostheticcrownand82.6Nforthe naturaltooth),theSIimagesshowthattheDEJandtheDCJ cannotbeobserved.Whentheinterfacesinbothsettingscan nolongeraccommodatetheloadingstress,theteethbeginto actaswholerigidobjects,andhigherloadingforcesgenerate cracksandfracturesofthesamples.

OurworktendstoconﬁrmbothZaslansky’sandWeiner’s studies:theDEJzoneisanimportantpartofthetoothstruc- ture.Moreover,wehavedemonstratedacomparablebehavior fortheDCJ.Webelievethesezonesarecrucialforthestress resistanceofthecrownstructure,whether naturalor prosthetic.Formonoblocvitreousceramicprostheticcrowns,load resistanceisaresultofthecementingprocess[27,28].

ThesoftDEJinterfaceisactuallyagradedstructure,and muchremainstobeunderstoodaboutthemannerinwhich thewholetoothbehavesunderload.Figs.6and7showthat theDCJpresentsagradedbehaviorbutwithasharperslope thantheDEJ.Aninterestinguseofdento-prostheticspacing couldbetoselectthecementjointthicknessbaseduponthe cementelasticmodulus.Thiswouldreﬁneattemptstomimic thephysiologicalbehaviorofthenaturaltooth[29],accord- ingtothebiomimeticconcept[30].Thisseemsalreadyvery importantasseveralauthorshaveinvestigatedtheeffectof differences in the resin-cement elastic moduluson stress- transmissiontoseveralcompositeorceramicreconstructions [31,32]. Other authors have studiedthe inﬂuence ofdiffer- entbondingagentsunderstressontheinternalandmarginal adaptationofcompositeorceramicreconstructions[33].

Thebiomimeticprinciplethroughtheuseofreconstruction ceramicsthathavewearcoefﬁcientsclosetothatofnatural enamelresultingfromcementswithanelasticmodulussim- ilartothatofthenaturalDEJ,wecanexpecttheconstruction ofrealbiomimeticprostheticteethinthenearfuture.

5. Conclusion

Speckleinterferometryishighlyrecommendedforperforming displacement measurements fordental biomechanics.The SIapproachallowsforthemeasurementofthemechanical propertiesofbiologicalstructuresandotherbiomaterialsthat areafewhundredmicronsthick.Thestrainaccommodation capacityofthetoothisderivedfromthedifferentialdisplace- mentbetweentheenamelandthedentin.Thedentin–enamel junctionactsasaninterfaceuntilacertainmaximalloading.

Beyondthisthreshold,theloadingaccommodationproperty

(9)

disappears,andthetooththenbehavesasarigidbody.We demonstrateasimilarbehaviorfortheprostheticcrowntooth.

Inthiscase,theenamelisreplacedbytheceramiccrownand thedentin–enameljunctionbythedentin–cementjunction.

ThesebehaviorsindicatetheinterfaceroleoftheDEJandthe DCJinthecapacityofteethtoaccommodatethestressesof theirphysiologicalfunctionsorevenoftheirparafunctions.

Forthecementcharacteristicsandthecementthicknessused inthis paper, the accommodation effect ofthe DCJ isless markedthanthatoftheDEJ.Futurestudywilldealwiththe effectofthecementcharacteristicsandthecementthickness byapplyingSIforsampleshavingdifferentcementsanddiffer- entthicknessofcementjoint,thankstotheCAD/CAMsystem.

ThelimitsofSImeasurementarethesensitivitytorigidbody motionanddecorrelationnoiseaffectingthespatialresolu- tion.Thiscanbeenhancedbyusingsmallerpixelsensorswith higherphotographicmagniﬁcation.

references

[1] ImbeniV,KruzicJJ,MarshallGW,MarshallSJ,RitchieRO.The dentin–enameljunctionandthefractureofhumanteeth.

NatureMaterials2005;4:229–32.

[2] TylmanSD.Thedentino-enameljunction.JournalofDental Research1928;8:615–22.

[3] HabelitzS,MarshallSJ,MarshallGW,BaloochM.The functionalwidthofthedentino-enameljunction determinedbyAFM-basednanoscratching.Journalof StructuralBiology2001;135:244–301.

[4] LinCP,DouglasWH.Structure-propertyrelationsandcrack resistanceatthebovinedentin–enameljunction.Journalof DentalResearch1994;73:1072–8.

[5] MeyerJM,Bodier-HoulleP,CuisinierFJG,LesotH,RuchJV.

Initialaspectsofmineralizationatthedentino-enamel junctioninembryonicmouseincisorinvivoandinvitro:a TEMcomparativestudy.InVitroCellularandDevelopmental Biology-Animal1999;35(March(3)):159–68.

[6] MarshallGW,BaloochM,GallagherRR,GanskySA,Marshall SJ.Mechanicalpropertiesofthedentin–enameljunction:

AFMstudiesofnanohardness,elasticmodulusandfracture.

JournalofBiomedicalMaterialsResearch2001;54:87–95.

[7] WangR,WeinerS.Strain–structurerelationsinhumanteeth usingMoiréfringes.JournalofBiomechanics1998;31:135–41.

[8] ZaslanskyP,ShaharR,BarakMM,FriesemA,WeinerS.

Toothandbonedeformation:structureandmaterial propertiesbyESPI.ProceedingsofSPIE2006;6341.

[9] ZaslanskyP,FriesemAA,WeinerS.Structureand mechanicalpropertiesofthesoftzoneseparatingbulk dentinandenamelincrownsofhumanteeth:insightinto toothfunction.JournalofStructuralBiology

2006;153(February(2)):188–99[Epub.2005December9].

[10] BieniekKW,MarxR.Themechanicalloadingcapacityof newall-ceramiccrownandbridgematerials.Schweizer MonatsschriftfurZahnmedizin1994;104(3):284–9.

[11] GonzagaCC,CesarPF,MirandaJrWG,YoshimuraHN.Slow crackgrowthandreliabilityofdentalceramics.Dental Materials2011;27(April(4)):394–406[Epub.2010December 24].

[12] AddisonO,SodhiA,FlemingGJ.Seatingloadparameters impactondentalceramicreinforcementconferredby cementationwithresin-cements.DentalMaterials 2010;26(September(9)):915–21.

[13] AttiaA,KernM.Inﬂuenceofcyclicloadingandlutingagents onthefractureloadoftwoall-ceramiccrownsystems.The JournalofProstheticDentistry2004;92(December(6)):551–6.

[14] GuardaGB,Gonc¸alvesLS,CorrerAB,MoraesRR,Sinhoreti MA,Correr-SobrinhoL.Lutingglassceramicrestorations usingaself-adhesiveresincementunderdifferentdentin conditions.JournalofAppliedOralScience2010;18(June (3)):244–8.

[15] FasbinderDJ.TheCerecsystem:25yearsofchairside CAD/CAMdentistry.JournaloftheAmericanDietetic Association2010;141(June(Suppl.2)):3S–4S.

[16] SlangenP,DeVeusterC,RenotteY,BerwartL,LionY.

Computer-aidedinterferometricmeasurementsofdriftand phaseshiftercalibrationforDSPI(digitalspecklepattern interferometry).OpticalEngineering1995;34(12):

3526–30.

[17] CreathK.Phase-measurementinterferometrytechniques.

In:ProgressinopticsXXVI.ElsevierSciencePublisherB.V.;

1988[Chapter5]p.349–393.

[18] CraigR,PeytonFA.Themicro-hardnessofenameland dentin.JournalofDentalResearch1958;37:661–8.

[19] CraigR,GehringP,etal.Relationofstructuretothe microhardnessofhumandentin.JournalofDentalResearch 1959;38:624–30.

[20] WoodJD,WangRZ,etal.Mappingoftoothdeformation causedbymoisturechangeusingMoiréinterferometry.

DentalMaterials2003;19:159–66.

[21] BechtleS,FettT,RizziG,HabelitzS,KlockeA,SchneiderGA.

Crackarrestwithinteethatthedentinoenameljunction causedbyelasticmodulusmismatch.Biomaterials 2010;31(May(14)):4238–47[Epub.2010February18].

[22] BarakMM,GeigerS,ChattahNLT,ShaharR,WeinerS.

Enameldictateswholetoothdeformation:aﬁniteelement modelstudyvalidatedbyametrologymethod.Journalof StructuralBiology2009;168:511–20.

[23] ChattahLT,ShaharR,WeinerS.Designstrategyofminipig molarsusingelectronicspeckleinterferometry:comparison ofdeformationunderloadbetweenthetoothmandible complexandisolatedtooth.AdvancedMaterials 2009;21:413–8.

[24] StraitDS,RichmondBG,SpencerMA,RossCF,DechowPC, WoodBA.Masticatorybiomechanicsanditsrelevanceto earlyhominidphylogeny:anexaminationofpalatal thicknessusingﬁnité-elementsanalysis.JournalofHuman Evolution2007;52(May(5)):585–99.

[25] ChattahNL,KupcikK,ShaharR,HublinJJ,WeinerS.

Structure-functionrelationofprimatelowerincisor:astudy ofthedéformationofMacacadentitionusingelectronic specklepatterninterferometry(ESPI).JournalofAnatomy 2011;218(1):87–95.

[26] NakamuraT,WakabayashiK,KinutaS,NishidaH,Miyamae M,YataniH.Mechanicalpropertiesofnewself-adhesive resin-basedcement.JournalofProsthodonticResearch 2010;54(April(2)):59–64[Epub.2009October30].

[27] AttiaA,AbdelazizKM,FreitagS,KernM.Fractureloadof compositeandfeldspathicallceramicCAD/CAMcrowns.

JournalofProstheticDentistry2006;95(February(2)):

117–23.

[28] AttiaA,KernM.Fracturestrengthofallceramiccrowns lutedusingtwobondingmethods.JournalofProsthetic Dentistry2004;91(March(3)):247–52.

[29] UrabeI,NakajimaS,SanoH,TagamiJ.Physicalpropertiesof thedentin–enameljunctionregion.AmericanJournalof Dentistry2000;13(June(3)):129–35.

[30] MagneP,BelserU.Understandingtheintacttoothandthe biomimeticprinciple.In:Magne,Belser,editors.Bonded porcelainrestorationsintheanteriordentition:a

biomimeticapproach.Chicago:QuintessencePublishingCo.;

2002.p.23–55.

(10)

Pleasecitethisarticleinpressas:FagesM,etal.Comparativemechanicalbehaviorofdentinenamelanddentinceramicjunctionsassessedby