外文翻譯--在水中利用高密度飛秒激光對玻璃表面進(jìn)行加工 英文版.pdf

DOI:10.1007/s00339-007-3930-zAppl.Phys.A87,691695(2007)MaterialsScience&ProcessingAppliedPhysicsAy.hayasakia117d.kawamuraHigh-densitybumpformationonaglasssurfaceusingfemtosecondlaserprocessinginwaterDepartmentofOpticalScienceandTechnology,FacultyofEngineering,TheUniversityofTokushima,2-1Minamijosanjima-cho,Tokushima770-8506,JapanReceived:13November2006/Accepted:29January2007Publishedonline:29March2007Springer-Verlag2007ABSTRACTMicrometer-sizedbumpswereformedonaglasssurfaceusingafocusedfemtosecondlaserprocessinginwater.Thebumpswereformedoverawiderangesofpulseirradiationparameters,includingirradiationenergyandfocusposition.Thebumpsexhibitedawidevarietyofmorphologiesandsizesde-pendingontheparameters.Theuseofaliquid,namelyheavywater,whichreturnsafterbreakdownandcavitationbubblefor-mation,enabledustofabricatebumpswithhighspatialdensity,whichisnotpossibleusingasolidcoatingthatisablated.Ade-siredarrangementofbumpsonaglasssurfacewasfabricatedbytuningtheprocessingtimeintervaltobemorethanthedisap-pearancetimeofabubble,generatedbyfocusingafemtosecondlaserpulsenearthewater/glassinterface.PACS42.62.Cf；42.70.Ce；52.38.Mf；78.47.+p；79.20.Ds1IntroductionFemtosecondlasersarepowerfultoolsformicro-andnano-structuringoftransparentmaterialsbecausetheycanprocesswithhighspatialresolutionresultingfrommul-tiplephotonabsorption,andreducedthermaldamageduetotheultra-shortinteractiontimebetweenthelaserpulseandthematerial,aswellasvariousphysicalphenomenacausedbytheultra-highintensityofthelaserpulse111.Fem-tosecondlaserprocessingisbeingincreasinglyappliedtothedevelopmentofthree-dimensionalopticalandfluidicde-vices7,8,1014.Asthemorphologyoftheprocessedtrans-parentmaterialisrelatedtothethermaleffectsofvaporizationanddissolutionduetothermaldiffusion,interactionwiththehotvaporplume,andalow-energy-densityregioninthelaserpulse,itishighlysensitivetonotonlythephysicalproper-tiesofthematerial,butalsotothelaserirradiationparameters,suchasthewavelength,pulseduration,pulseenergy,numer-icalapertureofthefocusedbeam,andthefocusposition.Inparticular,whenafemtosecondlaserpulseisfocusednearthesurfaceofatransparentmaterial,adifferenceinthefocuspos-itiongivesrisetoalargedifferenceinthesurfacemorphology.a117Fax:+81-88-656-9435,E-mail:hayasakiopt.tokushima-u.ac.jpThetypicalsurfacemorphologyofglassprocessedbyatightly-focusedfemtosecondlaserpulse,changesfromacavitytoabumpwhenthefocuspositionchangesfromtheoutsidetotheinsideoftheglass.Thecavityissurroundedbyaring-shapedprotrusionandscattereddebris.Theirsize,andtheamountofdebrisstronglydependsonthefocuspos-itionalso.Abumpwithadiameterfromseveralhundrednanometerstoseveralmicrometersisformedbymeltingtheglasssurfacewiththemeltedglassbeingpushedupbyami-croexplosioninsidetheglass1520.Duetotherangesoffocalpositionandirradiationpulseenergy,thesurfacemelt-ingandtheinternalmicroexplosionoccursimultaneouslyandthebumpsformedareverynarrow.Bumpstypicallyexhibitssmallvariationinsizeandstructure.Inapreviousstudy,wefoundthatatransparentcoatingontheglassfordecreasingtheamountofdebrisattachedtotheglasssurfaceallowsbumpformationoveraslightlywiderrangeoffocalpositionscomparedtobareglass,whenthecoatingthicknessissufficientlylargerthanthelengthofthefocalvolume19,21.Furthermore,wefoundthatwhenthecoatingthicknessisshorterthanthelengthofthefo-calvolume,thatis,whenthecoatingsurfaceisablatedbyasinglelaserpulsefocusedattheboundarybetweenthetransparentcoatingandtheglass,bumpswereproducedoverafairlywiderangeoffocuspositionscomparedtousingathickcoating20.Fromthoseinvestigations,webelievethattheamountofcoatingmaterialablatedinthefocalvol-ume,whichdependsonthecoatingthickness,affectsthestrengthofashieldingeffectoftheplasmageneratedwhenablatingthecoating.Asaresult,thesizeandstructureoftheformedbumpcanbechanged.Thetransparentcoatingmethodhasthedisadvantagethatthespatialdensityofthebumpsislimitedtoseveralmicrometersbecauseofablationofthetransparentcoating.Inordertoachievecontrollablefabri-cationofbumpswithahighdensity,itispossibletouseliquidonthetransparentmaterialinplaceofthetransparentcoatingduringfemtosecondlaserprocessing,becausetheliquidnatu-rallyreturnsafterbreakdownandbubbleformation.Fabrica-tionofcomplexstructuresonasiliconsurfacebyfemtosecondlaserprocessinginwaterhasbeendemonstrated2224.Inthispaper,wedemonstrateformationofhigh-densitymicrometer-sizedbumpsbyfemtosecondlaserprocessinginwater.InSect.2,wedescribetheexperimentalsetupand692AppliedPhysicsAMaterialsScience&Processingprocedure.InSect.3,wedescribetheexperimentalresults.Weinvestigatedtheeffectsofirradiationparameters,includ-ingenergyandfocalposition,onthemorphologyandsizeofthebumps.Wedemonstratedthat,bytuningtheprocess-ingtimeintervaltobemorethanthedisappearancetimeofabubble2528generatedbyafemtosecondlaserpulsefo-cusednearthewater/glassinterface,wecouldfabricateade-siredstructureontheglasssurface,composedofhigh-densitybumps.InSect.4,weconcludeourstudy.2ExperimentalsetupandprocedureTheexperimentalsetupconsistedofanamplifiedfemtosecondlaserandanopticalmicroscopeandisshowninFig.1.Itwasthesameasthesetupusedinourpre-viouswork19,20.Theamplifiedfemtosecondlaserpro-ducedpulseswithapeakwavelengthof800nm,adurationof150fs,andamaximumrepetitionrateof1kHz.TheirradiationpulseenergyEatthesamplewascontrolledbyneutraldensityfilters,andisgivenbytheproductoftheenergymeasuredbeforeintroducingthelaserpulseintotheopticalmicroscope(Olympus,IX70)andthetransmittanceoftheop-ticalmicroscope,includinga40objectivelens(numericalaperture,NA=0.65).Thetransmittanceofthemicroscopewas0.69.Theprocessedareawasobservedundertransmittedilluminationbyausualcharge-coupleddevice(CCD)imagesensorwiththeframerateof30frames/s.ThefocuspositionZofthelaserpulsewasdefinedasthedistancemovedalongtheopticalaxisbythemicroscopestage.Thezeroposition(Z=0)wasdefinedasthepositionwhereastructurewasformedontheglasssurfacebyirradiationofalaserpulsewithnearablationthresholdenergy.ThestructureofthesampleisalsoshowninFig.1.Thesamplewaspreparedasfollows.Fourordinarymicroscopecoverslips(Matsunami)whichweresubjectedtoultrasoniccleaninginethanolandpurewaterwereprepared.TheywereFIGURE1Experimentalsetupandthestructureofthesample.Thespacerglasseswereremovedwhentheprocessingwasperformedatargetglass,awindowglassforsandwichingwater,andtwospacerglasseswithathicknessof130m.Poly-methylmethacrylate(PMMA)withtoluenesolventwasusedtoformwallsonthewindowglass.Aftersufficientlyevaporatingthesolventthespacerglasseswereremoved,andasmallcham-berwithasidelengthof1015mmcomposedofthePMMAwallsontheglasswasformed.WaterwasdroppedinthesmallchamberandthetargetglasswasfixedonthechamberwithasmallamountofthePMMAthatwasusedasaglue.Inthisexperiment,deuteriumoxide(heavywater,hereafterreferredtosimplyas“water”)wasusedbecauseofitslowlinearabsorptionaroundthewavelengthof800nm.Afterpro-cessing,thetargetglasswasremovedfromthechamberandsubjectedtoultrasoniccleaninginpurewaterandethanol.Thesurfacestructureoftheprocessedareawasobservedwithanatomicforcemicroscope(AFM；DigitalInstruments,Di-mension3000).3ExperimentalresultsFigure2showsstructuresprocessedinwateroverarangeofZfrom4.0to12.0mwhentheenergyEwas2.1J.Figure2aandbshowanAFMimageanditscorres-pondingprofile,whoseverticalrangeis500nm.Figure2canddshowtopandsideviewsoftheprocessedareaobservedwiththetransmissionopticalmicroscope.Figure2eshowsthediameterandheightofthebumps,whichwereobtainedfromtheAFMobservation,andthelengthofavoid,whichFIGURE2(a)AFMimagesoftheprocessedareaand(b)theirprofiles.Theirradiationenergywas2.1J.Theverticalrangeis500nm.(c)Topand(d)sideviewsobservedwithatransmissionopticalmicroscope.(e)Diameterandheightofbumpsversusfocusposition,andthelengthofvoidsformedintheglassversusfocuspositionHAYASAKIetal.High-densitybumpformationonaglasssurfaceusingfemtosecondlaserprocessinginwater693wasobtainedfromasideviewobservation.ThebumpswereformedontheglasssurfaceoverawiderangeofZ,from4.0to8.0m.AsZincreased,theheightanddiameterofthebumpsincreased.WhenZwas6.0m,thebumphadamax-imumheightof400nmandadiameterof3.6m.WhenZwas8.0m,alowbumpwithaheightof50nmwasformed.WhenZwasgreaterthan8.0m,voidswereformedinsidetheglassandnostructurewasformedontheglasssurface.ThelengthofthevoidunderthebumpalsoincreasedasZincreased.ThevoidsformedwhenZwas4to12mwerenearlyequalinlength.UndermoredetailedobservationinthesideviewshowninFig.2d,wefoundthatthevoidshaddif-ferentgraylevelswhenZwasbetween6.0and8.0m.ThedarkhueofthevoidsunderthehighbumpsatZ=3.0mandZ=6.0mwasdarkerthanthoseofthevoidsformedcom-pletelyinsidetheglass.Weexpectedthevoidinthehighbumptohavelowerdensitythantheothers,becauseaninternalmi-croexplosiondisplacedtheglassmaterialfromthefocalpointandformedthehighbump,thuscausingadecreaseindensity.Thisbumpformationphenomenonisthesameasthatob-servedinourpreviousstudyinwhichglasshavingatrans-parentpolymercoatingwasprocessed.Theprincipleofbumpformationinthatstudywasbasedonthesuppressionofthematerialemissionfromtheglasssurfacebyashieldingeffectofplasmageneratedbyablationofthepolymerandbyphys-icalblockingofthepolymer.Onedifferenceinthepresentstudyisthatthebumpformationintheglassprocessedinwa-teroccursoverawiderrangeofZ,asshowninFig.3.Theirradiationbeamparameterswerealmostthesameasourpre-viousexperiments(showninFig.3in19).TheirradiationenergywasE=0.69J.Whenprocessingglasswithapoly-mercoating,bumpformationwasobservedwhenZwas1.0to4.0m20whereaswhenprocessinginwater,bumpfor-mationwasobservedwhenZwas4.0to7.0m.Themainreasonforthedifferenceisthatthephysicalblockingofwa-terisweakerthanthatofthepolymercoating.Thisisfurthersupportedbytheresultsforstructuresprocessedwithhighpulseenergies,aboveseveralmicrojoules,discussedinthenextparagraph.Figure4showsAFMimagesoftheprocessedstructuresforvariousenergiesEwhenZ=0.BumpswereformedwhenEwas0.17to6.9J,andtheirstructuresdrasticallychangeddependingonE.Thediameterandheightofthebumpin-creasedasEincreasedto4.1J.WhenEwas4.1J,thediameterwas5.1mandtheheightwas1.57m.Withfur-therincreaseofE,bothdimensionsdecreased.WhenE2.1J,therewaslittledebrisaroundtheperipheryofthebump.Although,whenE2.1J,debriswasdistributedaroundtheperiphery,andtheamountofdebrisincreasedasEincreased.ThescatteredregionofthedebrisisindicatedbythesquaresonthesolidlinesinFig.4.Processinginwa-terproducedmorescattereddebrisaroundthebumpthanprocessingwithanappliedpolymercoating.Thisfurthersup-portstheassertionthatwaterhadweakerphysicalblockingthanthepolymercoating.Mostofthedebriswasnotremovedbyultrasoniccleaninginwater.Therefore,theglassmaterialscatteredintheliquidstateattheglass/waterinterfacead-heredtotheglasssurfaceandsolidified.Figure5showbumpsarrangedinastraightlinewithhighdensity.Thelinearly-arrangedbumpswereprocessedbyir-FIGURE3Diameterandheightofbumpsversusfocusposition.Ewas0.69JFIGURE4AFMimagesofthestructuresprocessedwith(a)E=0.69J,(b)E=2.8J,(c)E=4.1J,(d)E=4.8J,(e)E=5.5Jand(f)E=6.9J.(g)Diameterandheightofbumpanddebrisdiameterversusirradiationenergyradiatingthelaserpulsesataspatialintervalshorterthanthediameterofasinglebump.Inthiscase,thespatialintervalDwassetto2.0m,undertheconditionthatasinglebumpwithadiameterof3.6mandaheightof56nmwasformedwhenEwas3.5JandZwas6.0m.Thestructurewasprocessedbyscanningthemicroscopestagesothatasinglepulsewasirradiatedateachlocation,repeatedatarepetitionrateRof1Hz.Theshapeofthelinearly-arrangedbumpswascontrolledbychangingD,asshowninFig.6aandb.WhenDwas0.8m,thebumpsweresmoothlyconnected,toformalineofbumps.WhenDwas5.0m,thatis,whenDwassuffi-cientlylargerthanthebumpdiameter,thebumpshadisolatedpeaks.694AppliedPhysicsAMaterialsScience&ProcessingFIGURE5AFMobservationoflinearly-arrangedbumpsformedunderE=3.5J,Z=6.0m,R=1Hz,andD=2.0m.(a)and(b)arethepro-filesacrossandalongthelinearly-arrangedbumps.Theverticalrangeoftheprofilesis250nmanditshorizontallengthis60mFIGURE6Surfacestructuresformedundervariousconditions.ThesameirradiationenergyofE=2.1Jwasused.In(a)and(b),Z=6.0mandR=1Hz,andthepulseirradiationspatialintervalsof(a)D=0.8mand(b)D=5.0mweredifferent.In(c)and(d),R=1HzandD=0.5m,andthefocuspositionsof(c)Z=6.0mand(d)Z=3.0mweredifferent.In(e)and(f),Z=6.0mandD=0.5m,andtherepetitionratesof(e)R=2Hzand(f)R=5Hzweredifferent.TheAFMimagesare88m2Tofabricatebumpswithhighdensity,ZandRwerecarefullychosen,inadditiontoEandD.Withtheirradi-ationconditionsZ=6.0m,E=2.1J,D=0.5m,andR=1Hz,asmoothlineofbumpswithauniformheightwasFIGURE7Bubblesgeneratedonthewater/glassinterfaceobservedwithaCCDimagesensor,whentheelapsedtime(a)t=2/30,(b)8/30,(c)12/30,and(d)13/30s.(e)Thedisappearan