ABSTRACT
Surfacecleaningschemesbasedontheinteractionofahigh-powerlaserpulse withasolidsurfaceandcontaminantparticleshavedrawnsubstantialattentionas thefeaturesizesshrinktonanoscaleinsemiconductoranddatastoragedevices.A numberoflasercleaningprocesses,includingdrylasercleaning(DLC),steamlasercleaning(SLC)andlasershockcleaning(LSC),havebeensuggestedand showntobeeffectiveforsub-micrometerparticleremoval[1-15].IntheLSC
method,thesurfaceisnotdirectlyexposedtolaserpulseirradiation,whichavoids thechanceforpotentialdamagetophoto-sensitiveparts,andarelativelylarge areacanbecleanedbyasinglelaserpulse[8-15].Furthermore,thereisnoneed tocontroltheliquidpurityasitisadryprocess.However,despitetheseadvantages,itisevidentthattheLSCprocessrequiresalargepulseenergyforremoval ofsub-micrometerparticles,incomparisontotheSLCprocess.AnumberofinvestigationshaveshownthattheSLCprocesscandetachsub-micrometerparticlesfromsolidsurfacesatlaserpulseenergieslowerthanthoserequiredbythe LSCprocess[1-7].However,acompleteremovalofparticlesassmallasafew tensofnanometershasneverbeenreported.Forexample,themaximumcleaning efficiencywasapproximately95%inthecaseof60-nmpolystyreneparticles[7]. Consequently,evenwiththeSLCprocess,removalofnanoscaleparticlesless than100nmindiameterisstillachallengingtask,andtheeffectivenessofthe conventionalcleaningmethodsisnotyetthoroughlyrevealedinthenanoscaleregime.Especiallythecleaningefficiencydatashowincreasinglylargescatteringas theparticledimensiondecreases,stronglydependingonthecharacteristicsofthe contaminantparticlesthatadheretosolidsurfaces.
