Secondary Electron Yield Measurements of TiN Coating and TiZrV Getter Film

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

SLAC-TN-03-052\revised27thAugust20041arXiv:physics/0310071v2 [physics.acc-ph] 30 Aug 2004SecondaryElectronYieldMeasurementsofTiNCoatingandTiZrVGetterFilmF.LePimpec,F.King,R.E.Kirby,M.PiviSLAC,2575SandHillRoad,MenloPark,CA940259thOctober2003AbstractInthebeampipeofthepositronMainDampingRing(MDR)oftheNextLinearCollider(NLC),ionizationofresidualgasesandsecondaryelectronemissiongiverisetoanelectroncloudwhichcancausethelossofthecirculatingbeam.Onepathtoavoidtheelectroncloudistoensurethatthevacuumwallhaslowsecondaryemis-sionyieldand,therefore,weneedtoknowthesecondaryemissionyield(SEY)forcandidatewallcoatings.WereportonSEYmeasurementsatSLAContitaniumni-tride(TiN)andtitanium-zirconium-vanadium(TiZrV)thinsputter-deposited lms,aswellasdescribeourexperimentalsetup.1IntroductionBeam-inducedmultipacting,whichisdrivenbytheelectric eldofsuccessivepositivelychargedbunches,arisesfromaresonantmotionofelectronsthatwereinitiallygeneratedorbygasionizationorbysecondaryemissionfromthevacuumwall.Theseelectronsthenbouncebackandforthbetweenoppositewallsofthevacuumchamber.Theelectronclouddensitydependsoncharacteristicsofthepositivelychargedcirculatingbeam(bunchlength,chargeandspacing)andthesecondaryelectronyieldandspectrumofthewallsurfacefromwhichthestartingelectronsarise.Theelectroncloude ect(ECE),duetomultipacting,hasbeenobservedorisexpectedatmanystoragerings[1].Thespacechargeofthecloud,ifsu cient,canleadtoalossofthebeamor,atleast,toadrasticreductioninbunchluminosity.

InordertominimizetheelectroncloudproblemwhichmightariseintheNLC,wearelookingtoasolutioninvolvingsurfacecoatingofthesecondaryelectronemittingvacuumwall.TheSEYoftechnicalsurfaceshasbeenmeasuredinthepastatSLAC[2]

[3],atCERNFig.1[4][5]andinotherlabs[6].InthispaperwepresentmeasurementsoftheSEYofmaterialspreviouslymeasuredandknowntohavelowSEYliketitanium-nitridethin lm(TiN)andtitanium-zirconium-vanadiumgetter lm(TiZrV).

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

Figure1:SEYofbakedtechnicalsurfaces.350 Cfor24hr[10]2ExperimentDescription

ThesystemusedtomeasureSEYiscomposedoftwocoupledstainlesssteel(S/S)UHVchamberswherethepressureisinthelow10 10Torrscaleinthemeasurementchamberandhigh10 9Torrscaleinthe”loadlock”chamber,Fig.2.Samplesindividuallyscrewedtoacarrierplate,areloaded rstontoanaluminiumtransferplateintheloadlockchamber,evacuatedtoalow10 8Torrscale,andthentransferredtothemeasurementchamber.

Themeasurementchamberhastwoelectrongunsandasoft(1.49keV)x-raysource.Oneelectrongun(energy,1-20keV)isusedforAugerelectronspectroscopy(AES)lightelementsurfacecontaminationanalysis.Thex-raysourceisusedtoexcitephotoelectronsforsurfacechemicalvalenceanalysis,calledESCA(ElectronSpectroscopyforChemicalAnalysis).TiNstoichiometryismeasuredbyESCAtechniquewhichisalsocalledXPS(X-rayPhotoelectronSpectroscopy).

TheprincipleofXPSistocollectphotoelectronsejectedbyx-raysofknownenergynearthesurface(1-5nminformationdepth).TheemittedelectronshaveanenergyEkwhichisgivenbyequation1

Ek=hν Eb Φ(1)

wherehνistheenergyoftheincidentphoton,EbthebindingenergyoftheelectronrelativetotheFermilevelofthematerialandΦthespectrometerworkfunction.Thespectrumofthemeasuredkineticenergygivesthespectrumofthebindingenergyofthephotoelectrons.

Thex-raysourceisalsousedforexcitingsecondaryX-rayFluorescence(XRF)forthicknessmeasurementofthedepositedTiNoverlayers.Thesecondelectrongun(0-3keV)isusedmeasuretheSEY,andcanalsobeusedtoelectronconditionthesurface.

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

Aniongunisavailableforcleaningthesamplesbysputteringandforionconditioning

surfaces.

Figure2:Experimentalsystemusedforsurfaceanalysis

Afterallsamples(uptotenorso)aretransferredintothemeasurementchamber,onesampleatatimeisloaded,onitsindividualcarrierplate,ontoamanipulatorarm(VacuumGenerators”Omniax”).TheOmniax carrierplateholderdesignisshowninFig.4.Twothermocouplesareinstalledonthisholderplateaswellasheating lamentandsampleconnectionwires.Samplescanbeheatedviaatungstenwire lament,Fig.3,byradiationorbyelectronbombardment.Electronbombardmentisachievedbybiasingthe lamentnegatively.TheOmniaxsampleholderisinsulatedfromgroundviaanaluminaceramicandbyseveralthermalshieldplatesinsulatedviafoursapphireballs,Fig.3andFig.4.

Thesamplecarrierplatesaremadeofmolybdenum(forattainingthehighesttem-peratures)orstainlesssteel,Fig.5.Thisplatewillsupportonesamplewhichisheldbycornerscrewsontoitssurface.ThisdesignallowstheplatetothenslideintotherailsoftheOmniaxplateholder,cfFig.3.AnexampleofanaluminiumprototypecarrierplateisshowninFig.6andFig.7.NotethatonFig.6theholesforthescrewsarenotdrilled.Inthiscon guration,thethermocouplesarenotattacheddirectlytothecarrierplatebutindirectlythroughtheholderplate.Sampletemperaturesarecomparedtothethermocoupletemperaturesbyusingablackbody-calibratedinfraredpyrometer(0.8µm-1.3µmbandpass)whichiscorrectedforabsorptionbythe7056glassviewportofthemeasurementchamberandthesampleemissivity( )of0.2-0.4.

AgoodwaytomonitortheactivationprocessoftheTiZrVnon-evaporablegetter

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

Figure3:Omniaxsampleholderplatewithheating lament

visible

Figure4:DrawingoftheOmniax SampleHolderplatewithouttheheating lament.

cVacuumGeneratorLtd

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

Figure5:SketchoftheSamplecarrierplatemadeinstainlesssteel.Dimensionsarein

inches

Figure6:Sampleholderproto-

typeinaluminium.Theblack

dotsarethelocationofthescrew

samplemounting

holesFigure7:Bottomsideofthesam-pleholder

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

(NEG)istorecordthedecreaseofthesurfaceoxygenconcentrationwithXPS.DuringtheNEGactivation,thesurfacegoesfromanoxidizedstatetoapartiallymetallicstate.ThebacksideoftheNEGwasheated,directlyviaaholeintheplateholder,byelectronbombardment.TheminimumtemperatureneededtoactivatethisNEGis180 C

[8][11].Thethermocouplesdisplayedatemperatureof216 Cfor2hours.InordertodeterminethetemperatureoftheNEG,aS/Ssampleholder( 0.31)washeated,withahotplate,tovarioustemperaturewithathermocouplespotweldedtoit,withandwithout7056glassviewportcorrection.MatchingthereadingofthepyrometerofthisexperimentalcalibrationwiththereadingrecordedduringaninsituheatingofthesameS/SsampleholderallowedustodeterminetheactualtemperatureoftheNEG,assumingthesameemissivity,TNEG=232 C.Anin-situtemperaturemeasurementwiththeblackbodycalibratedpyrometergivesforthesameS/Ssurface201 CwhenthethermocouplesontheOmniaxcarrierplateholderread216 C.

TheexperimentalelectroniccircuitisidenticaltotheonepresentedinFig.8[3].Thecomputercontrolledelectronbeamcomingfromthegunisdecoupledfromthetargetmeasurementcircuitry.However,thegroundiscommontoboth.Inshort,thetargetisattachedtoabiasingvoltagesourceandanelectrometerconnectedinseriestothedatagatheringcomputerAnalogDigitalConverter(ADC).

Figure8:Electroniccircuitryusedtomeasurethesecondaryemissionyield

3SEYmeasurementmethodology

Inordertocharacterizeoursystem,wehavemeasuredtheSEYin-situofinertgas-ioncleanedmaterials:carbon,copperandgold.Resultsarenotpresentedherebutagreewithwidelypublishedresultsforionsputtercleanedsampleofthesematerials.Aftercon rmingtheSEYfromthesereferencematerials,weproceededtomeasureaTiN-coated

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

aluminiumsamplesprovidedbyBNLandaTiZrVsputterdeposited lmonstainlesssteelsubstrate,obtainedfromCERN,

Fig.9.Figure9:Samplesmeasuredthusfar.

SEY(δ)de nitionisgiveninequation2.Inpracticeequation3isusedbecauseitcontainsparametersmeasureddirectlyintheexperiment.

δ=NumberofelectronsleavingthesurfaceIP(3)

WhereIPistheprimarycurrentorthecurrentleavingtheelectrongunandimpingingonthesurfaceofthesampleandITisthetotalcurrentmeasuredonthesample(IT=IP+IS).ISisthesecondaryelectroncurrentleavingthetarget.

Thespectrumofsecondaryelectroncurrentleavingthetargetiscomposedoftruesecondaries(0eVto40eV,byconvention),re-di usedprimaryelectronsexitingaftersu eringlossesinthesample(40eVtoEp)andfromincidentprimaryelectrons(Ep)elasticallyre ectedfromthesurface,Fig.10.Themajorityoftheelectronsleavingthesurfacearetruesecondaries.

Inordertomeasuretheprimarycurrentleavingtheelectrongun,thesampleisbiasedat+150V.Thebiasvoltagepreventsallre-di usedandsecondaryelectronsoflessthan150eVfromleavingthesample.Elasticallyre ectedelectronsarenotcollectedandcouldstrikenearbysurfaces,creatingsecondaryelectronsthatarethencollectedbythesamplebias.Thise ectissmallbecausethere ectivityat100eVisafewpercent.Weestimatetheerrorinprimarybeamcurrentmeasurementtobesmall(1-2%)becausethelossofelasticsisbalancedbythegainofnearbysecondaries.TheSEYaround100eVforbaked

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

stainlesssteeliscloseto1.1,Fig.1.Withregardtotheguncurrentasafunctionofenergy,itstartsatzeroforzeroenergyandsmoothlyincreasedtoitssaturationvalueat70eV.Wemeasurethemagnitudeandfunctionaldependenceofthebeamcurrentuptosomewhathighervalue(100eV)anduseaconstructedlookuptableofthebeamcurrentforSEYcalculations.Notbiasingthesamplewithahighenoughvoltagewillleadtoanunderestimationofthebeamcurrent.Thisiseasilyunderstoodfromthesecondaryspectrum,Fig.10.Theselected2nAguncurrentismeasuredforagunenergyof0-100eVbyenergystepsof10eV(0-3000eVrange)or2eV(0-300eVrange).

ActualmeasurementoftheSEYisdonebybiasingthesampleto-20V.Thisretarding eldrepelsmostsecondariesfromadjacentpartsofthesystemthatareexcitedbytheelasticallyre ectedprimarybeam.SEYmeasurementsaredonetwice,oncebetween0eVto3000eVwith10eVsteps,thenbetween0eVto300eV,with2eVsteps.The nalenergyoftheprimaryelectronsisrespectively2980eVand280eVbecauseofthebias.TheprimarybeamcurrentfunctionismeasuredandrecordedeachtimebeforeanSEYmeasurement,andwiththesamestepinenergyfortheelectronbeam.Afreshcurrentlookuptableiscreatedwitheachmeasurement.

Thepurposeofthesecondmeasurement,0eVto300eV,istotrytounderstandthestructureoftheSEYcurveatverylowenergy.Severalpointsareimportant,though.

1.Becauseofthenegativesamplebias,primaryelectronsnear0eVatthesampleareassuredtobeleavingthegun(20eVdeparture)andarrivingat(20eV-20V)eV.

2.BecauseofthealgorithmusedtocalculatetheSEYfromtheprimaryandsamplecurrentsat0eVincidentenergy,adivideby0blowupoccurs.Toavoidthisproblem,the rstpointat0eVisforcedtovalueone.The rst”true”datapointisat2eVfor300eVrangeand10eVfor3000eVrange.

3.Theuncertaintyinthesampleelectrometercurrentreadingissetbytheinputoperationalampli erbiasleakage,±20pA.

ingsamplecurrenttodetermineSEYexcludestheelasticallyre ectedelectronsfromthecalculation(theywouldneedtobecollectedbyanexternaltothesamplegridstructure).ThatservestoincreasetheSEY(lesssamplecurrent)by1-3%.Thisisfortuitouslybalancedbythefactthatthe-20Vbiasdoesnotrepel100%ofnearbysurfacesecondaries.

TheconsequenceofallofthesepointsisthattheSEYmeasurementsarejustthat,”secondary”.Whichmeansthatitdoesnotincludetheelastics.TheSEYmeasurementsarehowever,accurate.ItisimportanttonotlookattheSEYatlowprimaryenergyandtrytoconcludesomethingaboutelasticre ectivity.Databelow20eVcomesfromabandstructureandareacombinationofdi ractionfromthecrystallinestructureandenergyabsorptionbythematerial[12].Surfacee ectssuchasroughnesscanalsochangetheSEY.

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

Figure10:Spectrumofasec-

ondaryelectronbeamfroma

300eVincidentprimarybeam

impingingonaTiNonAlsub-

stratesample

[13]Figure11:TiNthicknessmea-suredbyx-ray uorescenceoftheTiK-lineoffourBNLsamples

4ResultsandComments

TheTiNcoating,madeatBrookhavenNationalLaboratory(BNL),wasdepositedontoaluminiumalloysubstrates,followingthesamerecipedescribedin[9].FortheSpallationNeutronSource(SNS)project,thecoatingwasdoneonS/S.AccordingtoBNL,theexpected lmthicknessisaround1000 A.WemeasuredtheactualsamplethicknessesusingXRF,Fig.11.

TheprincipleofXRFistocollectsecondaryx-raysgeneratedandexitingthesamplewhenbombardedbyprimaryincidentx-rays.Thesecondary uorescenceyieldishighestsomewhatabovetheK-absorptionedgeofTi,soprimaryx-raysof7keVareusedfortheexcitationoftheTi-Kαline(4.51keV).ATiN lmofsimilarknownthickness(byRutherfordbackscatterspectrometry,performedatanoutsidelab)isusedtocalibratethetechnique.

TheheightofthemeasuredTi-Kαline,usingaSiLi-driftedx-raydetector,islinearlyproportionaltothenumberofTiatomsinthe lm.ResultsareshownforafewsamplesinFig.11.

SEYmeasurementsresultsofsixdi erent”asreceived”TiNsamplesaredisplayedinFig.12andFig.13.Theelectronbeamimpingingontothesurfaceisoftheorderof2nAoveranareaoflessthanamm2.Typicallythebeamsizeisbetween0.2mmto0.4mmindiameter.ThelowcurrentisnecessaryinordertoavoidsurfaceconditioningduringSEYmeasurement.Thesizeofthebeamcanbecheckedbyusinga uorescentscreen,orisinferredfromsecondaryelectronmicroscopicalimaging(availableonthemeasurementsystemandusedtopreciselychoosethepointofSEYmeasurement).

TheSEYofthesamplesvariesfrom1.5to2.5,withthethickest lmsamplesdisplayingthelowerSEY.However,wehavenodataconcerningtheroughness,andtheroughnesscanbeafactorwhichcanchangetheSEY.Thecommonlyacceptedhypothesisisthat,foragivenchemicalsurface,theroughersurfacehasalowerSEYthanasmootherone

[4].TheirregularityintheSEY,atnearmaximum,forsampleTiN#4andTiN#6canbeduetoanonuniformspotemittingsecondarieswithtwodi erentyields.Theresult

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

Figure12:SEYofdi erentTiN

sampleforelectronenergybe-

tween0-2980

eVFigure13:SEYofdi erentTiNsampleforelectronenergybe-tween0-280eV.

couldbethesuperpositionofthetwoSEYcurves.Asthegunappearsveryreliableinrepetitivemeasurementswedonotbelievetheirregularityisanartifactduetothesystem.OthermeasurementsoftheSEYforunbakedS/S(δ 2at250eV)oranasreceivedNEG,

Fig.14,donotshowstheseshoulders.XPSsurveyofTiN#6showspresenceofmagnesium.Itsexpectedin uenceontheSEYisnotknown.

Finally,analternativecoatingtoTiNissputter-depositedTiZrVgetter(～2µmthick-ness).ThisNEG,whenactivated,showsadrasticreductionofitsSEY,Fig.14andFig.15.Theseresultscon rmwhathasbeeninvestigatedelsewhere[8].ItisalsointerestingtofollowthebehaviouroftheSEYcurveswhenthesampleisjustexposedtoonlyaresidualgasbackgroundof～3.10 10Torrforanextendedperiodoftime.TheSEYoftheTiZrVgoesupwithtimewhenexposedtoevensuchgoodvacuum.Interestinglyenough,itisclaimedin[8]thatthein uenceintheSEYoftheNEGafterexposureto30000L(1L=10 6Torr.s)ofCOorCO2israthersmall.δmaxwillincreasefrom1.1(CERNfullyactivatedNEG)to1.35(max)whileinUHV[8]. (SEY)is0.25usingCERNresultsandis0.3forours,1.3≤δmax≤1.6,Fig.14.

Figure14:SEYofTiZrVafter

di erentprocess,electronenergy

between0-2980eV

Figure15:SEYofTiZrVafterdi erentprocess,electronenergybetween0-280eV

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

Ifweconsider,at rst,thatourSEYmeasurementiserrorless,onecanarguethatouractivationwasnotcomplete.Hence,someinitialairformedoxideisstillpresentandthesurfacechemistryisnotidenticaltoafullyactivatedNEG,henceexplainingourδafteractivationof1.3.InthiscasetheSEYshouldstillnotincreaseabove1.35[8]asthegetterdoesnothavemanypumpingsitesleft.However,thishypothesismighthavetobediscardedastheminimaltemperaturewearecertaintohaveachievedisatleast201 C.Asaresultofapoweroutageofanhour,duringtheelevendaysofmonitoring,aniongauge,aresidualgasanalyserandtheionpumpwereturnedo .Attheswitchonofthegauges,hot lamentsreleasegasesinthesystembeforebeingpumpedawaybytheionpump.Therecordedpressureinthesystemwasattheswitchonofthegauges～6.10 10Torr.Thiskindofincidentalsocanhappeninanaccelerator,anditisinterestingtoseethatthise ectleadstotherecontaminationofthegetter.Betweenthedayselevenanddaystwenty,thesystemwasusedtoXPSothersamples.Duetothetransferofsamplesfromtheloadlockchambertothemeasurementchamber,thepressureroseupmomentarilyto～2.10 9Torr.

IfCOorCO2exposuredonotseemtoa ecttheincreaseofδ,airexposuredoes[8].Itispossiblethatopeningourbakedmeasurementchamber(P～3.10 10Torr)totheloadlockchamber(P～9.10 9Torr),whichisfrequentlyopentoair,canbeconsiderasanairexposure.Hence,thereisnocontradictionbetweenCERN[8]andourresults.

XPSanalysiswascarriedouttoobservetheevolutionofthecarbonchemistryduringthis20daysandcomparedtotheXPSspectrumtakenaftertheendoftheactivation.TheXPSspectrumshowsaslightriseoftheoxygenpeakandadoublevalencycarbon1speak.Theelementalcarbonpeakhasanenergyof～285eV,theoxidizedcarbonpeakhasanenergyof～288eV,andtheatypicalmetal-carbidepeakisaround～283eV.Theoxidizedcarbonpeakaftertheendoftheactivationat288eVwasbarelypresent,Fig.16,blueplot.OnlyTheCpeakandTi-Cpeakispresent,theNEGbeingat180 C.Acleanly-scrapedcarbonsurface,measuredatroomtemperature,willonlypresentapeakat285eV.Aftersixteenhoursofpumping,greenplot,theoxidizedstateofthecarbonshowsup,andthecarbidepeakisgone.Thisstate,oxidized,becomesdominantafterelevendaysofpumpingandkeepincreasingafter20days,graycurve.InthelastcasestheNEGwasatroomtemperature.TheplotspresentedinFig.16area toftheactualdata.ThisincreaseoftheC-OpeakimpliesthattheSEYshouldalsoincrease,asisusuallythecaseforoxidizedmetalsurfaces.

Theroughness(R)oftheTiZrV lmisunknown.Thisroughnessa ectsnotonlytheSEY[4]butalsothepumpingcapacityandspeedoftheNEG.TiZrVdepositedonS/Sisrelativelysmooth(R 1)andismuchrougheronanaluminiumsubstrate[11].Let’sassumethatoursystemwasfortwentydaysat～3.10 10Torr,theCObeingpresentat15%ofthetotalspectrum,andthatthestickingcoe cientσforCOis0.4.Thestickingcoe cientdecreaseswhentheTiZrVNEGhaspumpedalmosthalfofamonolayer(1ML～1015molecules-cm 2forasmoothsurface)byafactor10afterreaching1ML,andbyafactor100after10ML[11].Amonolayerofasurface(MLs)canbede nedas:MLs=ML×R[10].Thefrequencyofcollisionν(molecules.s 1.cm 2)equation4andthetotalamountofCOpercm 2pumpedbytheNEGinonedayisgivenbyequation5.Pν=3.51231022(4)NCO=ν×86400×σ(5)MTwherePisthepressureinTorr,MtheatomicmassandTisthetemperatureinK.AfteronedayofpumpingtheNEGwouldhavepumped～0.6ML.Assumingthat

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

Figure16:XPSCpeakofaTiZrV lmaftertheendofactivation

fortendaysthestickingcoe cientis0.04,theNEGwouldhavepumpedanadditional～0.6ML.AccordingtotheroughcalculationforthetotalamountofCOpumpedbytheNEG,theXPSof11daysand20daysshouldbesimilar,astheNEGisbasicallysaturated.ThisincreaseoftheC-Opeak,Fig.16,suggeststhatthesurfacestillhadsomeremainingpumpingspeedafterelevendaysofvacuumexposure.HencehavingaroughnessR>1.NEGprovidesanicesolutionforadistributedpumpinginsideavacuumchamber,whenactivated.Oursmallsample,afewcm2,beingat300KpumpedalloftheresidualgasinthechamberexceptCH4.InafewmeterslongNEGcoatedchamberthecontaminationcomesfromoutgassingsurfacesoutsidethischamber,sincetheNEG lmisadi usionbarrierforoutgassingmoleculesofthesubstrate.DependingoftheaveragevacuuminthemachineandthelengthoftheNEGchambers,therecontaminationofthesurfacemighttakelongerthanforoursample.Hence,theSEYmightnotincreaseasrapidlyasmeasuredhere,Fig.14andFig.15.

Also,ithastobetakenintoaccountthatthepumpinglifetimeofthin lmNEGdependsonitsthicknessandthenumberofactivationcycles[11].Thein uenceofthethicknessofthe lmshouldalsobetakenintoaccount,whencalculatingtheimpedancefortheimagecurrentonthevacuumchamberwall,duetoapassingparticlebeam.Therequirementontheconductivityandthegoodmechanicalperformanceafteranin-situbakeofthesubstrateisofimportance.

Forthemaindampingrings,currentdesignwillusealuminium-alloychamber.Aluminiumloosesmechanicalstrengthwhenheatedabove150 C.AnadequatesubstrateforTiNorTiZrVwhichful llsthemechanicalrequirementsisthealloy,Al6061,orAl6060whichiseasiertoextrude.Theirconductivityis～1.4timeslowerthanforpurealuminium.SEYmeasurementsonaTiN lmdepositedonAl6061hasalreadybeencarriedoutatSLAC

[13].

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

5Conclusion

WehavepresentedareportonthestatusoftheSEYexperimentcarriedoutatSLAC.Descriptionofourexperimentalsystemhasbeenpresented.

Firstresultsonas-receivedTiNsampleandonanas-receivedTiZrVgetterhavebeenshown.Inthecaseofthegetter,thein uenceoftheactivationandrecontaminationbyitspumpingactionwereinvestigated.ThemaximumSEYδincreasedfrom～1.3to～1.5afterelevendaysandto～1.6aftertwenty-twodaysofexposuretoavacuumof～3.10 10Torr.OurSEYresultsseemtodisagreewithCERN[8].Firstofall,ourstartingδmaxis1.3compareto1.1[8].Second,wehaveanincreaseinδmaxabovetheCERN-predicted1.35[8].Thismattershouldbeinvestigatedfurtherastheimplicationtoelectronclouddevelopmentisofimportanceforapositivelychargedbeamrunninginanaccelerator.

Valuesofδforenergiesbelow20eVshouldbeusedcarefullyifpluggedintosimulation.ItisplannedtoinvestigatefurtherthispartoftheSEYcurve,ashasbeendoneatCERN

[14].

Additionally,wewillstudythein uenceofvarioustreatments,suchasthein-situbakeoutofTiN,electronconditioningappliedfortheNLCcase,andalsothein uenceofionconditioning.Byconditioningwemeanbombardingthesurfacewithagivenspectruminenergyofelectronsandions.Di erentspeciesofionscanalsobeinvestigated.

Ithastobestressedthatconditioning(dosee ect)isaverye cientwayofloweringtheSEYofanytechnicalsurfacestoalmostthesamevalue(δ=1.2);andthusindependentlyoftheinitialδ[4].However,suchmeasurementsaretimeconsumingandmightnotberelevanttotheoperationofanaccelerator,dependingonthe uxofelectronsassociatedwithpost-commissioningproductionoperation.

6Acknowledgments

WewouldliketothankP.HeandH.C.HseuhatBNLforprovidingtheTiNsamplesandtheESTgroupfromC.BenvenutiatCERNfortheTiZrVsample.Wealsothank

A.WolskiatLBNLforshepherdingtheproductionofsampleplates,andinthenearfuture,forthin lmsamplescomingfromLBNL.MostvaluablewastheworkofG.ColletandE.Garwin,SLAC,forconvertingandbakingtheXPSsystemforuseonSEYmeasurements.

References

[1]M.Pivietal.RecentElectron-CloudSimulationResultsfortheMainDampingRingsoftheNLCandtheTESLALinearColliders.InPAC,Portland,Or,USA,2003.SLAC-PUB-9814.

[2]P.PrietoandR.E.Kirby.X-rayphotoelectronspectroscopystudyofthedifer-encebetweenreactivelyevaporatedanddirectsputter-depositedTiN lmsandtheiroxidationprperties.JournalofVacuumScienceandTechnology,A13(6),1995.

In the beam pipe of the positron Main Damping Ring (MDR) of the Next Linear Collider (NLC), ionization of residual gases and secondary electron emission give rise to an electron cloud which can cause the loss of the circulating beam. One path to avoid the

[3]R.E.KirbyF.K.King.SecondaryEmissionYieldfromPEP-IIacceleratormaterial.

NuclearInstrumentsandMethodsinPhysicsResearchA,A469,2001.

[4]N.Hilleretetal.http://clic-structures-working-group.web.cern.ch/clic-structures-

working-group/min/2000/18-9-2000/noel.pdf.

[5]N.Hilleretetal.TheSecondaryElectronYieldofTechnicalMaterialsanditsVari-

ationwithSurfaceTreatments.InEPAC,Vienna,Austria,2000.

[6]S.Kato,M.Nishiwaki.StudyonElectronEmissionfromSomeMetalsandCarbon

MaterialsandtheSurfaceCharacterization.In49thAVS,2002.

[7]urent,U.IrisoAriz.MultipactorTestsofaNEGcoatedChamber.Technical

report,CERN-LHC-VAC03-04,2003.

[8]C.Scheurlein.TheActivationofNon-evaporableGettersMonitoredbyAES,XPS,

SSIMSandSecondaryElectronYieldMeasurements.Technicalreport,CERN-THESIS-2002-026,2002.

[9]P.Heetal.DevelopmentofTiNcoatingforSNSRingVacuumChambers.InPAC

2001,2001.

[10]F.LePimpec.EtudedelaD´esorptionMol´eculaireInduiteparTransitionsElectron-

iquesdanslesSurfacesTechniques.PhDthesis,Universit´ePierreetMarieCurieParis6,2000.http://documents.cern.ch/archive/electronic/cern/preprints/thesis/thesis-2000-017.pdf.

[11]C.Benvenutietal.VacuumPropertiesofTiZrVnon-evaporablegetter lms.Vac-

uum,60:57,2001.

[12]E.Tamuraetal.Energy-DependenceofInnerPotentialinFefromLow-Energy

ElectronAbsorption(TargetCurrent).Technicalreport,SLAC-PUB-3594,1985.

[13]R.E.Kirby,F.K.King.SecondaryElectronEmissionfromAcceleratorMaterials:

Transparencies.Technicalreport,SLAC-PUB-8380,2000.

[14]R.Cimino,I.Collins.ASurfaceScienceApproachtoStudyElectron

CloudPhenomena.InDampingRingWorkshop,DaresburyUK,2003.http://www.astec.ac.uk/conf/dampingring/proceedings.html.

本文来源：https://www.bwwdw.com/article/8ya1.html