超快光學(xué)放大課件

TheAmplificationofUltra-

shortLaserPulsesFrancoisSalin

CenterforIntenseLasersandApplications(CELIA)UniversitéBordeauxI,Francewww.celia.u-bordeaux.frsalin@celia.u-bordeaux..frGillesDarpentigny(CELIA),VincentBagnoud(LLE)AntoineCourjaud,ClemensHonninger,EricMottay(AmplitudeSystemes),LucVigroux(AmplitudeTechnologies)andsomeadditionalstufffromDanMittleman,RicePulsecompressorttSolidstateamplifiertDispersivedelaylinetShortpulseoscillatorMost

ofthislecturecourtesyofTheAmplificationofUltra-

shPulseenergyvs.RepetitionrateReprate(pps)Pulseenergy(J)10910610310010-310-910-610010-3OscillatorCavity-dumpedoscillatorRegARegenRegen+multipassRegen+multi-multi-pass1WaveragepowerPulseenergyvs.RepetitionraA

tWhatarethegoalsinultrashortpulseamplification?Ipeak=EIncreasetheenergy(E),Decreasetheduration(t),Decreasetheareaofthefocus(A).MaximumintensityontargetNeededtostarttheexperimentNeededtogetusefulresultsPave=ErSignalisproportionaltothenumberofphotonsonthedetectorperintegrationtime.MaximumaveragepoweratthedetectorPulseenergyRepratePulseenergyBeamareaPulselengthAtWhatarethegoalsinultrIssuesinUltrafastAmplificationandTheirSolutionsPulselengthdiscrepancies:Multi-passamplifiers

andregenerativeamplifiers(“Regens”).Damage:Chirped-PulseAmplification(CPA)Gainsaturation:Frantz-NodvickEquationGainnarrowing:BirefringentfiltersThermaleffects:coldandwavefrontcorrectionSatellitepulses,Contrast,andAmplifiedSpontaneousEmission:Pockels’cellsSystemscostlotsofmoney:Earnmoremoney…Pockelscellpolarizergainpumpinput/outputIssuesinUltrafastAmplificatCavityDumpingBeforeweconsideramplification,recallthattheintracavitypulseenergyis~50timestheoutputpulseenergy.Sowehavemorepulseenergy.Howcanwegetatit?Whatifweinsteadusedtwohighreflectors,letthepulseenergybuildup,andthenswitchoutthepulse?ThisistheoppositeofQ-switching:itinvolvesswitchingfromminimumtomaximumloss,andit’scalled“CavityDumping.”E=ToutputEintracavityTransmissionofoutputcoupler:~2%R=100%R=98%EintracavityECavityDumpingBeforeweconsidCavitydumping:thePockelscellAPockelscellisadevicethatcanswitchapulse(inand)outofaresonator.It’susedinQ-switchesandcavitydumpers.Avoltage(afewkV)canturnacrystalintoahalf-orquarter-waveplate.VIfV=0,thepulsepolarizationdoesn’tchange.IfV=Vp,thepulsepolarizationswitchestoitsorthogonalstate.AbruptlyswitchingaPockelscellallowsustoextractapulsefromacavity.Thisallowsustoachieve~100timesthepulseenergyat1/100therepetitionrate(i.e.,100nJat1MHz).Pockelscell(voltagemaybetransverseorlongitudinal)PolarizerCavitydumping:thePockelscAmplificationofLaserPulses,inGeneralVerysimply,apowerfullaserpulseatonecolorpumpsanamplifiermedium,creatinganinversion,whichamplifiesanotherpulse.Nanosecond-pulselaseramplifierspumpedbyothernslasersarecommonplace.LaseroscillatorAmplifiermediumPumpEnergylevelsJpump(lpump/lL)lLlpumpAmplificationofLaserPulses,Single-passAmplificationMathAssumeasaturablegainmediumandJisthefluence(energy/area).Assumeallthepumpenergyisstoredintheamplifier,butsaturationeffectswilloccur.Atlowintensity,thegainislinear:Athighintensity,thegain“saturates”andhenceisconstant:Intermediatecaseinterpolatesbetweenthetwo:AmplifiermediumpumpJinJpumpJoutJsatlLlpumpJstoJsto=storedpumpfluence=Jpump(lpump/lL)Jsat=saturationfluence(materialdependent)Single-passAmplificationMathSingle-passAmplificationMathwherethesmallsignalgainperpassisgivenby:ThisdifferentialequationcanbeintegratedtoyieldtheFrantz-Nodvickequationfortheoutputofasaturatedamplifier:Single-passAmplificationMathFrantz-NodvickequationG0exp(g0L)exp(JstoJsat)Higherpumping(Jsto)meanshigherefficiencyandhighersaturationandsolowergain.Soyoucanhavehighgainorhighextractionefficiency.Butnotboth.11,21,41,61,822,22,42,600,20,40,60,81012345GainExtractionefficiency(Jout/Jsto)Jsto/JsatJout/JinFrantz-NodvickequationG0exp(Anotherproblemwithamplifyingultrashortlaserpulses…Anotherissueisthattheultrashortpulseissomuchshorterthanthe(nsorms)pumppulsethatsuppliestheenergyforamplification.Soshouldtheultrashortpulsearriveearlyorlate?Early:Late:Pumpenergyarrivestoolateandiswasted.timepumppumptimeEnergydecaysandiswasted.Inbothcases,pumppulseenergyiswasted,andamplificationispoor.AnotherproblemwithamplifyinSoweneedmanypasses.Allultrashort-pulseamplifiersaremulti-pass.Thisapproachachievesmuchgreaterefficiency.timepumpTheultrashortpulsereturnsmanytimestoeventuallyextractmostoftheenergy.Soweneedmanypasses.AllultTwomainamplificationmethodsMulti-passamplifierpumpinputoutputgainPockelscellpolarizergainpumpinput/outputRegenerativeamplifierTwomainamplificationmethodsAnothermulti-passamplifierAPockelscell(PC)andapairofpolarizersareusedtoinjectasinglepulseintotheamplifier.Anothermulti-passamplifierARegenerativeamplifiergeometriesThisisusedfor10-20-Hzrepetitionrates.IthasalargerspotsizeintheTi:sapphirerod.PockelscellFaradayrotatorthin-filmpolarizerPockelscellTheTi:Sapphirerodis~20-mmlonganddopedfor90%absorption.ThisdesignisoftenusedforkHz-repetition-rateamplifiers.RegenerativeamplifiergeometrPulseintensitiesinsideanamplifiercanbecomesohighthatdamage(oratleastsmall-scaleself-focusing)occurs.Solution: Expandthebeamand

uselargeamplifiermedia.Okay,wedidthat.Butthat’sstillnotenough.Solution: Expandthepulseintime,too.Okay,sowhatnext?PulseintensitiesinsideanamChirped-PulseAmplificationChirped-pulse

amplificationin-

volvesstretchingthe

pulsebeforeamplifyingit,

andthencompressingitlater.Wecanstretchthepulsebyafactorof10,000,amplifyit,andthenrecompressit!G.Mourouandcoworkers1983CPAisTHEbigdevelopment.PulsecompressorttSolidstateamplifiertDispersivedelaylinetShortpulseoscillatorChirped-PulseAmplificationChiStretchingandcompressingultrashortpulsesdf2ffdgratinggratingOkay,thislooksjustlikea“zero-dispersionstretcher”usedinpulseshaping.Butwhend≠f,it’sadispersivestretcherandcanstretchfspulsesbyafactorof10,000!Withtheoppositesignofd-f,wecancompressthepulse.PulsestretcherStretchingandcompressingultApulsestretcherThisdevicestretchesan18-fspulseto600ps—afactorof30,000!Araytraceofthevariouswavelengthsinthestretcher:

Pulsestretchercharacteristics:Inputpulsewidth:18fsOutputpulseduration:600psBandwidthpassed:>105nmPulseenergyout:~0.5nJApulsestretcherThisdevicesAlexandriteTi:sapphireExcimers0,00010,0010,010,11101001101001000104105106Nd:GlassDyesDirectAmplificationFluence(J/cm2)PulseDuration(fs)CPAvs.DirectAmplificationCPAachievesthefluenceoflongpulsesbutatashorterpulselength!AlexandriteTi:sapphireExcimersRegenerativeChirped-PulseAmplificationat~100kHzreprateswithacwpumpCoherent

RegAamplifierAfsoscillatorrequiresonly~5Wofgreenlaserpower.AnArgonlaserprovidesupto50W.Usetheresttopumpanamplifier.Today,weuseanintracavity-doubledNd:YLFpumplaser(~10W).Microjoulesat250kHzrepetitionrates!RegenerativeChirped-PulseAmpRegenerativechirped-pulseamplificationwithakHzpulsedpump

Wavelength:800nm

(Repetitionratesof1to50kHz)HighEnergy:<130fs,>2mJat1kHzPicosecond:~80ps,>0.7mJat1kHzShortPulse:<50fs,>0.7mJat1kHzSpectraPhysicsregen:the“Spitfire”Regenerativechirped-pulseampPumplaserforultrafastamplifiersCoherent“Corona”highpower,Q-switchedgreenlaserinacompactandmorereliablediode-pumpedpackage15mJ(ns)ata10kHzreprate(150Wavepower!)Pumplaserforultrafastampli1kHz10kHz100kHz20mJ1.8mJ0.2mJ20W18W20W3mm1mm250μmExtractedenergyBeamdiameterPumppower100WAveragePowerReprateAveragepowerforhigh-powerTi:SapphireregensTheseaveragepowersarehigh.Andthispumppowerisalso.Ifyouwantsub-100fspulses,however,theenergieswillbeless.1kHz10kHz100kHz20mJ1.8mJ0CPAisthebasisofthousandsofsystems.It’savailablecommerciallyinnumerousforms.Itworks!Buttherearesomeissues,especiallyifyoutrytopushforreallyhighenergies:Amplifiedspontaneousemission(ASE)Gainsaturation:gainvs.extractionefficiencyGainnarrowingThermalaberrationsContrastratioDamagethresholdvsextractionefficiencyCPAisthebasisofthousandsAmplifiedSpontaneousEmission(ASE)Fluorescencefromthegainmediumisamplifiedbefore(andafter)theultrashortpulsearrives.Thisyieldsa10-30nsbackgroundwithlowpeakpowerbutlargeenergy.Dependsonthenoisepresentintheamplifieratt=0ASEsharesthegainandtheexcitedpopulationwiththepulse.Amplificationreducesthecontrastbyafactorofupto10.AmplifiedSpontaneousEmissionGainNarrowing(andASE)Oneachpassthroughanamplifier,thepulsespectrumgetsmultipliedbythegainspectrum,whichnarrowstheoutputspectrum—andlengthensthepulse!Asaresult,thepulselengthens,anditcanbedifficulttodistinguishtheultrashortpulsefromthelongerAmplifiedSpontaneousEmission(ASE)GainNarrowing(andASE)OneacGainnarrowingexampleTi:sapphiregaincrosssection10-fssech2pulsein00.8100.511.522.536507007508008509009501000NormalizedspectralintensityCrosssection(*10^-19cm^2)Wavelength(nm)65-nmFWHM32-nmFWHMFactorof2lossinbandwidthfor107gainMostTerawattsystemshave>1010smallsignalgainlonger

pulseoutGainnarrowingexampleTi:sapphBeatinggainnarrowingBirefringentplatePolarizerPolarizerEEEwith

filterwithoutfilterSpectrumWavelength(nm)Gain&modulation650850900Wavelength(nanometers)70075080011.522.595020%GainmodulationBeforeAfterIntroducesomelossatthegainpeaktooffsetthehighgainthere.GainandlossSpectrum:beforeandafterBeatinggainnarrowingBirefrinGain-NarrowingConclusionGainnarrowingcanbebeaten.Wecanuseuptohalfofthegainbandwidthfora4-levelsystem.Sub-20fsinTi:sapphireSub-200fsinNd:glassSub-100fsinYb:XXGain-NarrowingConclusionGain00.81750800850Intensity(arb.units)Wavelength(nm)Verybroadspectracanbecreatedthisway.A100-nmbandwidthat800nmcansupporta10-fspulse.00.81750800850IntensiHeatdepositioncauseslensingandsmall-scaleself-focusing.Thesethermalaberrationsincreasethebeamsizeandreducetheavailableintensity.Ipeak=EATWewantasmallfocusedspotsize,butthermalaberrationsincreasethebeamsize,nottomentionscrewingitup,too.Nowtheaveragepowermatters.Therepetitionrateiscrucial,andwe’dlikeittobehigh,buthighaveragepowermeansmorethermalaberrations…ThermalEffectsinAmplifiersHeatdepositioncauseslensingLowtemperatureminimizeslensing.CalculationsforkHzsystemsCryogeniccoolingresultsinalmostnofocalpowerInsapphire,conductivityincreasesanddn/dTdecreasesasTdecreases.Murnane,Kapteyn,andcoworkersLowtemperatureminimizeslensStaticWave-frontCorrection2.5timesimprovementinpeakintensityhasbeenachievedCUOSStaticWave-frontCorrection2.DynamicCorrectionofSpatialDistortion50mmdiameter37actuatorsCUOSDynamicCorrectionofSpatialContrastratioWhydoesittakeover2yearsbetweenthefirstannouncementofanewlasersourceandthefirstsuccessfulexperimentusingit?Becausethepulsehasleadingandfollowingsatellitepulsesthatwreakhavocinanyexperiment.Ifapulseof1018W/cm2peakpowerhasa“l(fā)ittle”satellitepulseonemillionthasstrong,that’sstill1TW/cm2!Thiscandosomeseriousdamage!Ionizationoccursat1011W/cm2soat1021W/cm2weneeda1010contrastratio!ContrastratioWhydoesittakeMajorsourcesofpoorcontrastNanosecondscale: pre-pulsesfromoscillator pre-pulsesfromamplifier ASEfromamplifierPicosecondscale: reflectionsintheamplifier spectralphaseoramplitudedistortionsMajorsourcesofpoorcontrast0-1-2-3-4-5-6-7-8-9-10FrontBacktimeSpectralphaseaberrationsPre-pulsesASE0ps10nsnsFWHMAmplifiedpulsesoftenhavepoorcontrast.Log(Energy)Pre-pulsesdothemostdamage,messingupamediumbeforehand.0FrontBacktimeSpectralphaseTypical3rd-orderautocorrelationAmplifiedpulseshavepre-andpost-pulses.Typical3rd-orderautocorrelatAPockelscell“PulsePicker”APockelscellcanpickapulsefromatrainandsuppresssatellites.Todoso,wemustswitchthevoltagefrom0tokVandbackto0,typicallyinafewns.VTimeVoltagefewnsSwitchinghighvoltagetwiceinafewnsisquitedifficult,requiringavalanchetransistors,microwavetriodes,orotherhigh-speedelectronics.APockelscell“PulsePicker”AamplifieroscillatorstretchercompressorPockelscells10-2-10-310nsPockelscellssuppresspre-andpost-pulses.Unfortunately,Pockelscellsaren’tperfect.Theyleak~1%.amplifieroscillatorstretchercoContrastimprovementrecipesAPockelscellimprovesthecontrastbyafew100to1000.Weneedatleast3Pockelscellsworkinginthebestconditions:onaxis(donottiltPockels’cells)broadbandhigh-contrastpolarizers(notdielectric)fastrisetime(<<2ns10-90%)collimatedbeams

TemperaturedriftisalsoaprobleminPockelscells.Also:Goodpumpsynchronizationgivesafactor3-10ContrastimprovementrecipesAMultiple-stagemulti-passamplifiers1kHzMulti-passsystemattheUniversityofColorado(MurnaneandKapteyn)0.2TW4mJ,20fspulselengthMultiple-stagemulti-passamplClosedloopcryogeniccooling100oK5x1J,20HzNd:YAGlasers10fsoscillatorstretcher100HzRegenerativeamplifier200mJ,30fs,100HzHighenergy,highcontrast100-HzsystematCELIAClosedloopcryogeniccoolingPumpbeamUltrashortpulse--nearfieldUltrashortpulse--farfieldAmplified-pulsebeamshapesPumpbeamUltrashortpulse--neA1-JouleApparatusA1-JouleApparatusMulti-JouleSystemsNd:Glass

Energy>20Joules

Pulsewidth<500fs

Power>40TW

Repetitionrates~everyhourTerawattLaserSystem

Ti:sapphire

Energy>1Joule

Pulsewidth<35fs

Power>10TW

Repetitionratesto1kHzYoucanbuytheselasers!Multi-JouleSystemsNd:Glass

EnLawrenceLivermoreNationalLabsHigh-PowerAmplifiersLaserSpecificationsUSP1996USP(InDevelopment)Janus1996TotalEnergy1Joule10Joule6-10JoulePulseLength100fs30-100fs600fsPeakIntensity5.e19W/cm25-10.e20W/cm25.e18W/cm2Wavelength800nm,400nm800nm,400nm1053nm,526nmPulseContrast<10-7in400nm<1.e-7in400nm<10-9in526nmRepRate1shotin3minutes1shotin20minutes1shotin3minutesLawrenceLivermoreNationalLaEvenHigherIntensities!NationalIgnitionFacility(underconstruction)192shapedpulses10.4kJperbeaminUV(done)21kJperbeaminIR(done)>1.8MJtotalenergy(planned)Pulses0.2to25nsinlengthEvenHigherIntensities!NationWhattodowithsuchhighintensitiesWhattodowithsuchhighinte知識(shí)回顧KnowledgeReview祝您成功！知識(shí)回顧KnowledgeReview祝您成功！TheAmplificationofUltra-

shortLaserPulsesFrancoisSalin

CenterforIntenseLasersandApplications(CELIA)UniversitéBordeauxI,Francewww.celia.u-bordeaux.frsalin@celia.u-bordeaux..frGillesDarpentigny(CELIA),VincentBagnoud(LLE)AntoineCourjaud,ClemensHonninger,EricMottay(AmplitudeSystemes),LucVigroux(AmplitudeTechnologies)andsomeadditionalstufffromDanMittleman,RicePulsecompressorttSolidstateamplifiertDispersivedelaylinetShortpulseoscillatorMost

ofthislecturecourtesyofTheAmplificationofUltra-

shPulseenergyvs.RepetitionrateReprate(pps)Pulseenergy(J)10910610310010-310-910-610010-3OscillatorCavity-dumpedoscillatorRegARegenRegen+multipassRegen+multi-multi-pass1WaveragepowerPulseenergyvs.RepetitionraA

tWhatarethegoalsinultrashortpulseamplification?Ipeak=EIncreasetheenergy(E),Decreasetheduration(t),Decreasetheareaofthefocus(A).MaximumintensityontargetNeededtostarttheexperimentNeededtogetusefulresultsPave=ErSignalisproportionaltothenumberofphotonsonthedetectorperintegrationtime.MaximumaveragepoweratthedetectorPulseenergyRepratePulseenergyBeamareaPulselengthAtWhatarethegoalsinultrIssuesinUltrafastAmplificationandTheirSolutionsPulselengthdiscrepancies:Multi-passamplifiers

andregenerativeamplifiers(“Regens”).Damage:Chirped-PulseAmplification(CPA)Gainsaturation:Frantz-NodvickEquationGainnarrowing:BirefringentfiltersThermaleffects:coldandwavefrontcorrectionSatellitepulses,Contrast,andAmplifiedSpontaneousEmission:Pockels’cellsSystemscostlotsofmoney:Earnmoremoney…Pockelscellpolarizergainpumpinput/outputIssuesinUltrafastAmplificatCavityDumpingBeforeweconsideramplification,recallthattheintracavitypulseenergyis~50timestheoutputpulseenergy.Sowehavemorepulseenergy.Howcanwegetatit?Whatifweinsteadusedtwohighreflectors,letthepulseenergybuildup,andthenswitchoutthepulse?ThisistheoppositeofQ-switching:itinvolvesswitchingfromminimumtomaximumloss,andit’scalled“CavityDumping.”E=ToutputEintracavityTransmissionofoutputcoupler:~2%R=100%R=98%EintracavityECavityDumpingBeforeweconsidCavitydumping:thePockelscellAPockelscellisadevicethatcanswitchapulse(inand)outofaresonator.It’susedinQ-switchesandcavitydumpers.Avoltage(afewkV)canturnacrystalintoahalf-orquarter-waveplate.VIfV=0,thepulsepolarizationdoesn’tchange.IfV=Vp,thepulsepolarizationswitchestoitsorthogonalstate.AbruptlyswitchingaPockelscellallowsustoextractapulsefromacavity.Thisallowsustoachieve~100timesthepulseenergyat1/100therepetitionrate(i.e.,100nJat1MHz).Pockelscell(voltagemaybetransverseorlongitudinal)PolarizerCavitydumping:thePockelscAmplificationofLaserPulses,inGeneralVerysimply,apowerfullaserpulseatonecolorpumpsanamplifiermedium,creatinganinversion,whichamplifiesanotherpulse.Nanosecond-pulselaseramplifierspumpedbyothernslasersarecommonplace.LaseroscillatorAmplifiermediumPumpEnergylevelsJpump(lpump/lL)lLlpumpAmplificationofLaserPulses,Single-passAmplificationMathAssumeasaturablegainmediumandJisthefluence(energy/area).Assumeallthepumpenergyisstoredintheamplifier,butsaturationeffectswilloccur.Atlowintensity,thegainislinear:Athighintensity,thegain“saturates”andhenceisconstant:Intermediatecaseinterpolatesbetweenthetwo:AmplifiermediumpumpJinJpumpJoutJsatlLlpumpJstoJsto=storedpumpfluence=Jpump(lpump/lL)Jsat=saturationfluence(materialdependent)Single-passAmplificationMathSingle-passAmplificationMathwherethesmallsignalgainperpassisgivenby:ThisdifferentialequationcanbeintegratedtoyieldtheFrantz-Nodvickequationfortheoutputofasaturatedamplifier:Single-passAmplificationMathFrantz-NodvickequationG0exp(g0L)exp(JstoJsat)Higherpumping(Jsto)meanshigherefficiencyandhighersaturationandsolowergain.Soyoucanhavehighgainorhighextractionefficiency.Butnotboth.11,21,41,61,822,22,42,600,20,40,60,81012345GainExtractionefficiency(Jout/Jsto)Jsto/JsatJout/JinFrantz-NodvickequationG0exp(Anotherproblemwithamplifyingultrashortlaserpulses…Anotherissueisthattheultrashortpulseissomuchshorterthanthe(nsorms)pumppulsethatsuppliestheenergyforamplification.Soshouldtheultrashortpulsearriveearlyorlate?Early:Late:Pumpenergyarrivestoolateandiswasted.timepumppumptimeEnergydecaysandiswasted.Inbothcases,pumppulseenergyiswasted,andamplificationispoor.AnotherproblemwithamplifyinSoweneedmanypasses.Allultrashort-pulseamplifiersaremulti-pass.Thisapproachachievesmuchgreaterefficiency.timepumpTheultrashortpulsereturnsmanytimestoeventuallyextractmostoftheenergy.Soweneedmanypasses.AllultTwomainamplificationmethodsMulti-passamplifierpumpinputoutputgainPockelscellpolarizergainpumpinput/outputRegenerativeamplifierTwomainamplificationmethodsAnothermulti-passamplifierAPockelscell(PC)andapairofpolarizersareusedtoinjectasinglepulseintotheamplifier.Anothermulti-passamplifierARegenerativeamplifiergeometriesThisisusedfor10-20-Hzrepetitionrates.IthasalargerspotsizeintheTi:sapphirerod.PockelscellFaradayrotatorthin-filmpolarizerPockelscellTheTi:Sapphirerodis~20-mmlonganddopedfor90%absorption.ThisdesignisoftenusedforkHz-repetition-rateamplifiers.RegenerativeamplifiergeometrPulseintensitiesinsideanamplifiercanbecomesohighthatdamage(oratleastsmall-scaleself-focusing)occurs.Solution: Expandthebeamand

uselargeamplifiermedia.Okay,wedidthat.Butthat’sstillnotenough.Solution: Expandthepulseintime,too.Okay,sowhatnext?PulseintensitiesinsideanamChirped-PulseAmplificationChirped-pulse

amplificationin-

volvesstretchingthe

pulsebeforeamplifyingit,

andthencompressingitlater.Wecanstretchthepulsebyafactorof10,000,amplifyit,andthenrecompressit!G.Mourouandcoworkers1983CPAisTHEbigdevelopment.PulsecompressorttSolidstateamplifiertDispersivedelaylinetShortpulseoscillatorChirped-PulseAmplificationChiStretchingandcompressingultrashortpulsesdf2ffdgratinggratingOkay,thislooksjustlikea“zero-dispersionstretcher”usedinpulseshaping.Butwhend≠f,it’sadispersivestretcherandcanstretchfspulsesbyafactorof10,000!Withtheoppositesignofd-f,wecancompressthepulse.PulsestretcherStretchingandcompressingultApulsestretcherThisdevicestretchesan18-fspulseto600ps—afactorof30,000!Araytraceofthevariouswavelengthsinthestretcher:

Pulsestretchercharacteristics:Inputpulsewidth:18fsOutputpulseduration:600psBandwidthpassed:>105nmPulseenergyout:~0.5nJApulsestretcherThisdevicesAlexandriteTi:sapphireExcimers0,00010,0010,010,11101001101001000104105106Nd:GlassDyesDirectAmplificationFluence(J/cm2)PulseDuration(fs)CPAvs.DirectAmplificationCPAachievesthefluenceoflongpulsesbutatashorterpulselength!AlexandriteTi:sapphireExcimersRegenerativeChirped-PulseAmplificationat~100kHzreprateswithacwpumpCoherent

RegAamplifierAfsoscillatorrequiresonly~5Wofgreenlaserpower.AnArgonlaserprovidesupto50W.Usetheresttopumpanamplifier.Today,weuseanintracavity-doubledNd:YLFpumplaser(~10W).Microjoulesat250kHzrepetitionrates!RegenerativeChirped-PulseAmpRegenerativechirped-pulseamplificationwithakHzpulsedpump

Wavelength:800nm

(Repetitionratesof1to50kHz)HighEnergy:<130fs,>2mJat1kHzPicosecond:~80ps,>0.7mJat1kHzShortPulse:<50fs,>0.7mJat1kHzSpectraPhysicsregen:the“Spitfire”Regenerativechirped-pulseampPumplaserforultrafastamplifiersCoherent“Corona”highpower,Q-switchedgreenlaserinacompactandmorereliablediode-pumpedpackage15mJ(ns)ata10kHzreprate(150Wavepower!)Pumplaserforultrafastampli1kHz10kHz100kHz20mJ1.8mJ0.2mJ20W18W20W3mm1mm250μmExtractedenergyBeamdiameterPumppower100WAveragePowerReprateAveragepowerforhigh-powerTi:SapphireregensTheseaveragepowersarehigh.Andthispumppowerisalso.Ifyouwantsub-100fspulses,however,theenergieswillbeless.1kHz10kHz100kHz20mJ1.8mJ0CPAisthebasisofthousandsofsystems.It’savailablecommerciallyinnumerousforms.Itworks!Buttherearesomeissues,especiallyifyoutrytopushforreallyhighenergies:Amplifiedspontaneousemission(ASE)Gainsaturation:gainvs.extractionefficiencyGainnarrowingThermalaberrationsContrastratioDamagethresholdvsextractionefficiencyCPAisthebasisofthousandsAmplifiedSpontaneousEmission(ASE)Fluorescencefromthegainmediumisamplifiedbefore(andafter)theultrashortpulsearrives.Thisyieldsa10-30nsbackgroundwithlowpeakpowerbutlargeenergy.Dependsonthenoisepresentintheamplifieratt=0ASEsharesthegainandtheexcitedpopulationwiththepulse.Amplificationreducesthecontrastbyafactorofupto10.AmplifiedSpontaneousEmissionGainNarrowing(andASE)Oneachpassthroughanamplifier,thepulsespectrumgetsmultipliedbythegainspectrum,whichnarrowstheoutputspectrum—andlengthensthepulse!Asaresult,thepulselengthens,anditcanbedifficulttodistinguishtheultrashortpulsefromthelongerAmplifiedSpontaneousEmission(ASE)GainNarrowing(andASE)OneacGainnarrowingexampleTi:sapphiregaincrosssection10-fssech2pulsein00.8100.511.522.536507007508008509009501000NormalizedspectralintensityCrosssection(*10^-19cm^2)Wavelength(nm)65-nmFWHM32-nmFWHMFactorof2lossinbandwidthfor107gainMostTerawattsystemshave>1010smallsignalgainlonger

pulseoutGainnarrowingexampleTi:sapphBeatinggainnarrowingBirefringentplatePolarizerPolarizerEEEwith

filterwithoutfilterSpectrumWavelength(nm)Gain&modulation650850900Wavelength(nanometers)70075080011.522.595020%GainmodulationBeforeAfterIntroducesomelossatthegainpeaktooffsetthehighgainthere.GainandlossSpectrum:beforeandafterBeatinggainnarrowingBirefrinGain-NarrowingConclusionGainnarrowingcanbebeaten.Wecanuseuptohalfofthegainbandwidthfora4-levelsystem.Sub-20fsinTi:sapphireSub-200f