DEVELOPMENTOFTHGEMBASEDDETECTORSFORAT-TPCAPPLICATIONSByStefanHermannRostATHESISSubmittedtoMichiganStateUniversityinpartialentoftherequirementsforthedegreeofPhysics-MasterofScience2016ABSTRACTDEVELOPMENTOFTHGEMBASEDDETECTORSFORAT-TPCAPPLICATIONSByStefanHermannRostStudyingunstablenucleiawayfromthelineofstabilityusinglowenergyreactionsisofhighinterestinmodernnuclearphysics.Thelowenergyionsobservedinthesereactionslosemostoftheirenergywithinthetarget,thereforetheconstructionofanActiveTargetTimeProjectionChamber(AT-TPC),detectingthetrackwithinthetarget,isanattractivesolution.Inordertomaximizethereactionyieldwhileminimizingbackgroundoperationinpuregaswithoutquenchgasisnecessary.Inpuregas,however,thetransitiontostreamermode(discharge)occursalreadyatlowgainwhenusingcommonstructures(e.g.GEM(gaselectronmultiplier)),thereforeanewtypeofdetectorfunctioningathighgaininpureelemtalgasisrequired.TwoTHickGaseousElectronMultipliers(THGEM)incascadewerefoundtooperateatagainofupto104inpureheliumforpressuresbetween200torrand600torr,whilealsoprovidingagoodenergyresolution.TheTHGEMisahole-typegaseouselectronmultiplierproducedbymultilayerprintedcircuitboard(PCB)technology.Itconsistsofadenselyperforatedassemblyof0.6mmthickFR-4substrate,sandwichedbetweenthinmetallicelectrodestrata.ThreeTHGEMswerestackedtogethertoformanovelgaseousmultiplier(M-THGEM),providingahightoftheelectronavalanchewithinthehole,resultinginagainofupto104:5inpurelow-pressurehelium.Itdemonstratedaverygoodenergyresolutionintheexperimentandanestimatedionbacwofaslowas13%insimulations.Furthermore,itisextremelysuitabletocoverlargeareasduetoitsrobustness,makingthethreelayerM-THGEMsuitableforalargedetectorsuchastheAT-TPC.ACKNOWLEDGMENTSIwouldliketothankmyadvisor,Prof.DanielBazin,forhisunwaveringsupportduringmyworkattheNationalSuperconductingLaboratory.WhenIneededhelp,hisdoorwasalwaysopen.ThisprojectwouldnothavebeenpossiblewithoutMarcoCortesi.Heintroducedmetotheresearchofgaseousdetectordevelopment,actingasmymentorandteacherduringmytimeinthelaboratory.Hispatientexplanationsguidedmethroughmanyobstaclesalongtheway.Prof.WolfgangMittig'sknowledgeableinputwasavaluableresourcethroughoutthere-search.Mysincerethanksalsogotomygroup,forthefruitfuldiscussionsonandresults.Thesupportofthemechanicalandelectronicengineerswascriticaltothedevelopmentoftheexperiment.iiiTABLEOFCONTENTSLISTOFTABLES....................................viLISTOFFIGURES...................................viiKEYTOSYMBOLS..................................xiiKEYTOABBREVIATIONS.............................xiiiChapter1Introduction...............................1Chapter2Operatingprinciplesofgaseousdetectors.............32.1Avalancheprocess.................................32.1.1Gain....................................32.1.1.1Longtermgainstability....................42.1.1.2Penning..........................52.1.1.3Impurities............................52.1.2Ionfeedback................................62.1.3Photonfeedback.............................62.1.4Limitofgain...............................72.2Hole-typedetectorstructures..........................82.2.1TwocascadeTHGEMs..........................92.2.2Three-layerM-THGEM.........................102.3Signalformation.................................11Chapter3Simulations................................133.1Fieldsimulation..................................133.1.1TwocascadeTHGEMs..........................133.1.2Three-layerM-THGEM.........................153.1.2.1Symmetricvoltages.......................153.1.2.2Asymmetricvoltages......................163.2Particledrift....................................163.2.1RangeintheAT-TPC..........................163.2.1.1SRIM(energylosscalculation)................163.2.1.2MCNPX(alternateenergylosscalculation).........173.2.2Electroncollection.......................193.2.3Gain....................................203.2.3.1tgeometriesandgases.................203.2.3.2TwocascadeTHGEMs.....................213.2.3.3Three-layerM-THGEM....................243.2.4Ionbacw................................27iv3.2.4.1Transferandionfeedback................273.2.4.2oftherelativealignmentbetweenthetwoTHGEMsonelectronandiontransfer..................293.2.4.3CollectionofIonsinthethree-layerM-THGEM.......303.2.4.4Ionfeedbackvs.Gainfort.......34Chapter4MeasurementsforthetwocascadeTHGEMs...........364.1ExperimentalSetup...............................364.1.1Gashandlingsystem...........................364.1.2Voltagepowersupplies..........................374.1.3Readoutelectronics............................374.2PerformanceinHelium..............................384.2.1Gain....................................384.2.2Energyresolution.............................394.3PerformanceinHydrogen............................404.3.1Gain....................................404.3.2Electrondriftvelocity..........................414.4Furtherinterpretationofmeasurements.....................42Chapter5TowardsanewgenerationofM-THGEMs:Three-layerM-THGEM..................................44Chapter6Conclusionandoutlook........................48APPENDICES......................................50AppendixAAdditionalgraphs............................51AppendixBTechnicaladvice.............................53AppendixCManualforperformingsimulationsofTHGEMswithintheNSCL..54BIBLIOGRAPHY....................................76vLISTOFTABLESTable3.1:RangeofalphaparticlesinheliumsimulatedusingSRIM......17Table3.2:RangeofalphaparticlesinhydrogensimulatedusingMCNPX....19Table4.1:DriftvelocityofelectronsinhydrogenaccordingtoFig.4.5......41Table4.2:pcalculatedfromthelinearforthecurvesinFig.4.6.......43viLISTOFFIGURESFigure1.1:VisualizationoftheAT-TPCworkingprinciple.Figuretakenfrom[1]1Figure2.1:Illustrationofonesingleavalanche...................4Figure2.2:Visualizationofthesecondaryemissionfromtheelectrodeduetophotons.Figuretakenfrom[9]......................6Figure2.3:Illustrationoftheformationofstreamersandtheresultingareaofionizedgas(lefttoright).InspiredbyRef.[13]............8Figure2.4:SchematicdrawingofatwocascadeTHGEMsetup,likeitisusedintheAT-TPC................................9Figure2.5:SetupoftwocascadeTHGEMsincludingthemicromegas.Theimageiscreatedin[5]andshowsthedriftlinesofelectronsinvacuum.10Figure2.6:SetupofthethreelayerM-THGEMincludingthevisualizationofanavalanchecreatedbyasingleelectron.Theimageismadeusing[5]................................11Figure3.1:SimulationoftheelectricaldipoleontopofaTHGEM(left).PlotoftheelectricalmagnitudeswithinthewholestructureoftwocascadeTHGEMs.Thedriftissetto0.5kVcm,theinductionissetto1.5kVcmandthetransferis1.5kVcm.ThevoltageacrossoneTHGEMis1500V......................14Figure3.2:SimulationoftheelectricstrengthwithinathreelayerM-THGEM(left).ThevoltageacrossalllayersoftheM-THGEMisthesame.Plotoftheelectricstrengthalongaverticalcutthroughtheplotontheleft,showingtheenhancementinthecenteroftheM-THGEM(right)............................15Figure3.3:SimulationofthestrengthwithinathreelayerM-THGEM.Ontheleft,thevoltageappliedtoalllayersisequivalent,ontherightthetoplayerisreducedto3(70%)...................16Figure3.4:RangewithintheAT-TPCusingpureHeliumat400torr.Theenergylossperangstromisplottedasfunctionofthepenetrationdepth.The3.6mthickpara-aramidwindowistakenintoaccount........17viiFigure3.5:RangewithintheAT-TPCusingpureHydrogenat400torr.TheresultingenergydistributionisshowninFig.3.6...........18Figure3.6:DistributionoftheeneregydepositioninhydrogenwithintheAT-TPC,takingintoaccountthewindowandangledistributionofthesource...................................18Figure3.7:Collectionforathree-layerM-THGEMinpureheliumat200torr.Thepercentagesgivendescribethepotentialinthelayercomparedtotheotherlayersfortasymmetricvoltage..........................20Figure3.8:Gaincurvesforalltgeometriesdiscussedinthiswork.Thedataisgeneratedusing[5]atapressureof600torrforallcurves...................................21Figure3.9:Gaincurvesfortpressuresinpurehelium,usingatransferof0.25kVcmandaninductionanddriftof0.5kVcm....22Figure3.10:Gaincurvesfortpressuresinpurehydrogen,usingatransferof0.25kVcmandaninductionanddriftof0.5kVcm....23Figure3.11:Gaincurvesfordtpressuresinpurehelium,usinganinductionanddriftof0.5kVcm.......................25Figure3.12:Gaincurvesfortpressuresinpurehydrogen,usinganin-ductionof1.5kVcmanddriftof0.1kVcm.Thegraphincludesgainsforthetwotoperationmodes,symmetricvoltagesandasymmetricvoltagesTHGEM70%)................26Figure3.13:Plotoftheiontracks(red)andelectrontracks(orange)inoneavalancheforasetupwithtwocascadeTHGEMs.TheplotsaregeneratedinH2at400torrwithavoltageof1025VacrosseachTHGEMusing[5].Thepotentialare:drift0.1kVcm,induction1.5kVcm,1.5kVcm(left)and0.125kVcm(right)......27Figure3.14:CorrespondingtoFig.3.13thehistogramsontheleftshowtheend-pointsforionsandelectronsusingahightransferwhereasontherightthesevaluesareshownforaweaktransfer......28viiiFigure3.15:Plotoftheiontracks(red)andelectrontracks(orange)inoneavalanchefortwocascadeTHGEMs,whicharenotaligned.Theplotsaregen-eratedinH2at400torrwithavoltageof1050VacrosseachTHGEMusing[5].Thedrifteldissetto0.1kVcm,theissetto0.125kVcmandtheinductionis1.5kVcm.Theleftpictureshowsa3Dviewandtherightdiagrama2Dillustration.......30Figure3.16:Statisticsoftheendpointsforions(right)andelectrons(left)inthetwocascadeTHGEMs,whicharenotalignedlikeshowninFig.3.15.TheplotsaregeneratedinH2at400torrwithavoltageof1050VacrosseachTHGEMusing[5].Thedriftissetto0.1kVcm,theissetto0.125kVcmandtheinductionis1.5kVcm..31Figure3.17:Plotoftheiontracks(red)andelectrontracks(orange)inoneavalancheforathree-layerM-THGEM.TheplotsaregeneratedinH2at400torrwithavoltageof650VacrosseachTHGEMusing[5].Thedriftissetto0.1kVcmandtheinductionis1.5kVcm.Theleftpictureshowsa3Dviewandtherightdiagrama2Dillustration...32Figure3.18:CorrespondingtoFig.3.17thehistogramsontheleftshowtheend-pointsforionsandelectronsusingahighsymmetricvoltageof650VforeachTHGEM,whereasontherighttheTHGEMisworkingasacollector(asymmetricvoltageItissetto70%ofthefullvoltage..............................33Figure3.19:Ionbacw(numberofionsatthecathodeovernumberelectronsattheanode)plottedagainstthegainforathree-layerM-THGEMandtwocascadesinglelayerTHGEMsinH2at400torr.ThedecreaseinIBFforhighergainisclearlyvisibleforeachsetup.Theerrorisindicatedasthe˙/2spreadofthesimulation..............34Figure4.1:Calibrationcurvesforthechargeforthetwotam-Theslopeofthetwocurvesare6electronsbinand60electronsbinrespectively................................37Figure4.2:Comparisonofthegaininpureheliumfortpressures.Thehorizontallinesindicatethemaximumachievablegain.InadditionthreemeasurementsbyMarcoCortesi[9]usingUV-Lightareincluded.38Figure4.3:Visualizationfor2000eventsinheliumat600torr(left).Theenergyspectrumcalculatedfromtheseevents(right).............39ixFigure4.4:Comparisonofthegaininpurehydrogenfortpressures.Theendpointofeachcurveindicatesthemaximumachievablegainatthatpressure..................................40Figure4.5:Visualizationof1000eventsinhydrogenforerentpressures....41Figure4.6:Comparisonofthegaininpurehydrogenfortpressures.Thecurvesinthisgraphareusedtotheparameterp..........42Figure5.1:Aschematicdrawing(left(topview)andtopright(sideview))andaphotograph(bottomright)ofthemountedM-THGEM........44Figure5.2:Spectrumofa5.5MeValphaparticlecrossingthepAT-TPCactivevolume,recordedwiththethreelayerM-THGEMoperatedinHe-basedmixture(10%CO2)........................45Figure5.3:GainmeasurementforthethreelayerM-THGEM(schematicdraw-ing(a))inHe/(10%)CO2(b)andinpureHe(c)............46Figure5.4:ComparisonoftheelectricmapforathreelayerM-THGEMholesymmetricallybiased(a)andasymmetricbiased(b).3representsthereducedfractionoftheappliedpotentialtothemultiplierstage.Measurementsofegainin200torrHe/10%CO2fortthreelayerM-THGEMc.................47Figure5.5:Comparisonoftheelectroncollection(a)andionbacw(b)inHe/10%CO2fortasymmetricM-THGEMs.47FigureA.1:Comparisonbetweentheexperimentaldata(Fig.4.2)inheliumforthetwocascadeTHGEMsandthesimulatedresults(Fig.3.9).ThetdiscrepancycanbeexplainedbythemissingPenningin[5]...............................51FigureA.2:Comparisonbetweentheexperimentaldata(Fig.4.4)inhydrogenforthetwocascadeTHGEMsandthesimulatedresults(Fig.3.10).Thetdiscrepancycanbeexplainedbyresidueoxygen,capturingfreeelectrons...............................52FigureC.1:\recordscript"functioninMaxwell..................56FigureC.2:StartMobaxtermandenterhostname.................59FigureC.3:Correctsettingsinnotepad++,tomodifytheforsubmittingtothebatchsystem.Thecorrectlineending\LF"isshown.......70xFigureC.4:Wrong(windows)linebreakintheforsubmittingtothebatchsystem...................................71xiKEYTOSYMBOLSEtrans-ElectrictransferEdrift-ElectricdriftEind-ElectricinductionV-Voltage-VoltageacrossoneTHGEM-ElectricalI-currentq-chargeEW-weightingG-Gain-Townsendcot-electronmeanfreepathxiiKEYTOABBREVIATIONSAT-TPC-ActiveTargetTimeProjectionChamberGEM-GaseousElectronMultipliersTHGEM-THickGaseousElectronMultipliersM-THGEM-MultilayerTHGEMFWHM-FullWidthHalfMaximumIBF-IonBacwxiiiChapter1IntroductionStudyingthestructuralevolutionandthereactiondynamicsofunstablenucleiawayfromthelineofstabilityisofhighinterestinmodernnuclearphysics[1].Modernacceleratorfacilities(e.g.NSCL[2])provideradioactivebeamsoverawidedomainofthenuclearchart.Whentheseheavy-ionbeamscollidewithalight-iontarget,mostofthereactionenergyiscarriedawaybytheheavyions.Indetectingthelightionswithlowrecoilenergyof0.1-10MeV[1],alargeportionoftheirkineticenergyislostwithinthetarget,thereforeconventionaldetectors(HiRA[3],HELIOS[4])canonlydetectpartofthekineticenergy,oftenresultinginapoorenergyresolution.Figure1.1:VisualizationoftheAT-TPCworkingprinciple.Figuretakenfrom[1]Theconstructionofadetectorforwhichtrackingofparticlesispossiblewithinthetargetcancircumventthisissue.ActiveTargetTimeProjectionChambers(AT-TPC)aregaseousdetectorsinwhichalargevolumeofalow-massgasactssimultaneouslyasionization1mediumandtarget.Theelectronscreatedbyaparticletravellingthroughthevolumearedetectedwithspatialresolutionattheendofthechamber(Fig.1.1).Combiningthisdatawithtiminginformation,thetrackofaparticleintheactivevolumecanbereconstructedin3D[1].AnAT-TPCrequiresanelectronmultipliercapableofdeliveringalargedynamicrangeandhighgaininlow-pressureoperationinpureelemtalgas(H2,D2,3He,4He,etc.).Theoperationwithoutaquencherisimportantbecauseitmaximizesthereactionyieldwithminimalbackgroundsubstraction.Withouttheuseofaquenchgasonlyreactionswiththedesiredtargetatomsaccure,forexamplewhenusingaprotontargetonlyreactionswithH2arevisible,insteadofhavingadditionalcollisionswithaquenchgaslikeCO2.Itistoreachstablehigh-gainoperationofproportionalgaseousdetectorsundertheseconditions(lowpressureandnoquencher),mainlyduetotheconsiderablephoton-mediatedsecondarythatleadtoanearlytransitionfromproportionalavalanchemodetostreamermode(discharge).THGEMs,especiallyinacascadesetuporasmultilayerTHGEM(M-THGEM),cansolvetheseissues.THGEMsareaholetypestructureof0.5mm-1mmthickness,consistingoftwocopperlayersseperatedbyanisolator.Avoltageisappliedacrossthegapbetweenthetwoelectrodes.Throughavalancheprocesseselectronsaremultipliedinthisgap.Itisessentialtodeterminetheenergyresolutionandthegain,dependingonthevoltageV.Thisinvestigationisthesubjectofthepresentworkthroughexperimentaldataandsimulationsperformedin[5].Itisfollowedbytheinvestigationofaninno-vativenewTHGEMconcept:amultilayerTHGEM(M-THGEM)thatconsistsofmultipleTHGEMscombinedintoonemodule,tlyimprovingtheperformance.2Chapter2Operatingprinciplesofgaseousdetectors2.1Avalancheprocess2.1.1GainGain,(multiplication)ofaincomingelectronsignal,inagaseousmultiplierisachievedviathecreationofnewelectron-ionpairs.Anewelectron-ionpairiscreatedwhentheenergytransferredfromacollisionisgreaterthantheionizationenergyofthegasatomormolecule.Oncetheelectronhasgainedenoughenergyfromtheelectriceld(byacceleration),anewelectron-ionpaircanbecreated.Thisprocesscreatesanavalancheofelectronsifthedriftlengthinthemultiplicationareaislongenough(Fig.2.1).Duringtheaccelerationprocessofaelectron,elasticandinelasticcollisionswithgasatomsormoleculesoccur.Theexcitedgasatoms,canemitaphotoncausingsecondarydescribedinsection2.1.3and2.1.4.Thefrequencyofcollisionsdependsontheelectronsmeanfreepath,whichincreaseswhenthegaspressureisdecreased.MultiplicationinagascanbedescribedusingtheTownsendcot()G=ed;3Figure2.1:Illustrationofonesingleavalanche.withdbeingthelengthofthemultiplicationzone.Athighreducedelectricstrength,thestTownsendcotdependsmainlyonthemeanfreepathoftheelectrons[6].Inthiscase,theenergygainedbetweentwocollisionsishigherthantheionizationenergyandthemultiplicationcanbeassumedasinverselyproportionaltotheelectronsmeanfreepath()=1[6],sinceeachcollisionleadstomultiplication.Forlowreducedelectricstrength(smaller20Volt/torr/cm)theTownsendcocientisproportionaltotheelectric[7].2.1.1.1LongtermgainstabilityThelongtermgainstabilityofTHGEMorGEMbasedelectronmultipliersismainlyrelatedtoradiation-inducedchargingupoftheisolator(e.g.,FR4)andvariationofthegascompo-sitionduetooutgassingofcomponents,watervaporandresidualgases[8].Thevariationofthegainduetotheseeisrelatedtothegaspressureandthestrengthwithintheholes[9].Chargingupoftheisolatorsubstratecausesadecreaseintheelectricstrengthandthereforedecreasesthegain.TheholegeometryandtheholediameterhaveagreatonthisIttypicallyoccursonamuchshortertimescale(underonehour)comparedtothechangeofthegasmixture.ToachieveastableoperationoftheTHGEM,itshould4berunningforseveralhoursinorderforthegasmixturetostabilize(givenaconstantgasw)[9].Impuritiesinthegasusuallydecreasethegain(section2.1.1.3),althoughveryfewmixturesmaycauseanincreaseofgain(section2.1.1.2).2.1.1.2PenningAmixtureofapuregaswithasmallamountofquenchgas,forwhichtheexcitationlevelofthepuregasishigherthantheionizationenergyofthequenchgas,iscalledPenningmixture[10].Forthismixtureitisespeciallyeasytocreatenewelectron-ionpairs,becausetheexcitedstateofthepuregascanionizethequenchgas.Thiscanincreasethemultipli-cationbyseveralordersofmagnitude.ExamplesforaPenningmixtureareneonwithargonorheliumwithnitrogen.2.1.1.3ImpuritiesResidualgasescanhavetsometimesintended.ThePenning(2.1.1.2)canbeusedtoincreasethetotalgainofagasmixture.Othergases(e.g.oxygen)haveahighelectronicyandcancaptureelectrons[11].Thiscanbeaproblem,sincebycapturingelectronsspatialresolutionandenergyresolutionislost-theelectronisrecombined.Thisalsomeansthegainistlylowered.Duringtheavalancheprocessphotonsareproduced(section2.1.3).Inordertocapturethesephotons,complexmoleculeswithmanytvibrationalandrotationalmodescanbeutilized.Theeigenmodesfromthesemolecules(i.e.CO2,CH4)canbeexcitedwithabroadrangeofphotonsandabsorbthem[12].Acombinationoftwogaseswiththeseproperties,forexampleheliumwithcarbondioxyde(10%)iscommontouseforconventionalGEMstructuresinordertocapturethephotonscreatedduringtheavalancheprocess.The5gasregulatingtheamountoffreephotonsiscalledaquencher.2.1.2IonfeedbackWheneveranelectronisgeneratedinanavalanchethecorrespondingiondriftsintheoppositedirection.Theseionscan,forexample,causeadistortionoftheelectricandthereforedecreasethespatialresolution.Inaphotomultiplier,theseionsevencauseadditionalwavesofelectronswhenevertheyhitaelectrode.Itisthereforeimportanttocontroltheionbacw.ThiscanbedonebycapturingtheionsontheupperlayeroftheTHGEM(section3.2.4).2.1.3PhotonfeedbackDuringtheavalancheprocessmanyphotonsareproduced,mainlyduetoexcitationofthegasatomsinsteadofionization.Thesephotonscancausesecondaryionizationprocessesinthegasandontheelectrodes.InaPenningmixture(section2.1.1.2)andalsoinpuregassuchaphotoncanionizeagasatom.Figure2.2:Visualizationofthesecondaryemissionfromtheelectrodeduetophotons.Figuretakenfrom[9].Thisleadstothecreationofmanynewavalanches.Whenaphotonhitsanelectrode,anelectroncanbeemittedduetothephotoelectric(Fig.2.2).Inbothcasesthespatial6resolutionandtheenergyresolutionbecausetheisnotproportionaltotheoriginalsignalanymore[13].Awellknownmethodtocapturethesephotonsandpreventsecondaryionizationistoaddaquencher(i.e.CO2,CH4)tothegasmixture(section2.1.1.3).Thesephotonscanalsobeabsorbedbytheisolatorsubstrate,iftheavalancheiswithinthehole.Nomatterwhichwaythephotonsareregulated,theresultisahighermaximumachiev-ablegain,becauseproportionaloperationispossibleathighervoltages.Ifthephotonsaren'tcaptured,theavalanchegrowsexponentially,fedbyadditionalsmallavalanches,eventuallycausingacompleteionizationofthegaswithinthehole.Theionizedgasthenbecomesaconductorbetweenthetwoelectrodes,causingasparkandvoltagebreakdown[13].2.1.4LimitofgainRaether'slimit,anempiricalvalueforthemaximumavalanchesize,isd<20,meaningamaximumgainof108[13].Butbeforethisempiricallimitisreached,therecanbeashortcircuitduetoaconductingbandofionizedgasbetweenthetwoelectrodes,whennoquenchgasisused.Thisiscalledstreamermode.Formationofadditionalavalanchesbeforeandaftertheoriginalone,duetosecondaryionizationbyphotons,leadtothestreamermode.Theavalanchebecomesonelongconductingbandwhentheisdeformedduetospacechargecausedbytheavalancheitself.Theavalanchescausedbyphotonmediatedsecondarythendrifttowardstheexistingavalanche,causingittofurthergrowanddeformtheevenmore.Thereforetheformationofstreamersisaselfenhancingprocess.ItisshownFig.2.3.Dependingonthegasmixtureandgeometrythiscanlimitthemaximumgaintlybyseveralordersofmagnitude[13].7Figure2.3:Illustrationoftheformationofstreamersandtheresultingareaofionizedgas(lefttoright).InspiredbyRef.[13]2.2Hole-typedetectorstructuresAschematicaldrawingoftwocascadeTHGEMsisshowninFig.2.4.THGEMsconsistoftwocopperelectrodesandacorematerial(i.e.FR-4,Kapton,Kevlar).Holesaredrilledmechanicallythroughtheselayers,whichareproducedbymulti-layerprintedcircuitboard(PCB)technology.Theproductionprocessisdbychemicallyetchingasmallrim(100m)aroundtheholesintobothoutercoppersurfaces.Rimshavetheadvantageofreducingtheprobabilityofdischargeduetomechanicaldefectsandthusincreasethegainlimit[14].THGEMsarealotmorerobustmechanicallyandeasiertohandlecomparedtomicromegasortraditionalGEMs,duetotheirthicknessofabout0.6mm.PrimaryelectronsaretransferredtotheTHGEMsetupbyadrift(Edrift).AftermultiplicationintheTHGEMwithavoltagediofV,theyaretransferredtothesecondTHGEMundertheofatransfer(Etrans).TheyarethenextractedfromthesecondTHGEMbytheinduction(Eind).Theenceoftheseonthe8Figure2.4:SchematicdrawingofatwocascadeTHGEMsetup,likeitisusedintheAT-TPC.optimumelectrongainandminimumionbacwisimportanttounderstandforasuccesfuloperation.2.2.1TwocascadeTHGEMsTwosingle-layerTHGEMscanbecombinedintoamoreemodule.Bycombiningthemitispossibletoachievethesamegainwithasmallervoltageacrosseachmodulecomparedtothatofasinglemodule.ThetwoTHGEMsareplaced2mmapartwithatransferof0:25kVcm.AportionoftheionbacwfromthesecondTHGEMisabsorbedbythesecondTHGEMandthereforereduced.IntheAT-TPCthissetupiscombinedwithamicromegas.ThefullmoduleisshowninFig.2.5.TheeldstrengthratiobetweenthedipolewithintheTHGEMandthedrifttransferandinductiondeterminetheelectroncollectionciency,transferandextractioniencyrespectively(section9Figure2.5:SetupoftwocascadeTHGEMsincludingthemicromegas.Theimageiscreatedin[5]andshowsthedriftlinesofelectronsinvacuum.3.2.2).IntheAT-TPCthemicromegasislocated128mabovethereadoutpads.Whenoperat-ingtheTHGEMasthemultiplicationdevice,thereisnomultiplicationnecessarywithinthemicromegas.Thus,thetransperancyofthemicromegascanbeusedtosetthegainforeverysinglepadbyapplyingabiasonindividualreadoutpadswiththemicromegasgrounded.EachTHGEMconsistsofasubstrateof0.6mmthickness.Thecopperstratais0.01mmthick.Thediameterandpitchoftheholesare0.5mmand1mmrespectively.2.2.2Three-layerM-THGEMCombiningthreeTHGEMstoonesinglemodule(threelayerM-THGEM)allowsforahighertoftheavalanchewithinthehole,whichgivesabettercontroloverphoton-mediatedsecondaryThiseliminatestheneedofatransferbecauseallelectronsandionsaretransferredfromonelayertothenext.SuchasetupisshowninFig.2.6.The10Figure2.6:SetupofthethreelayerM-THGEMincludingthevisualizationofanavalanchecreatedbyasingleelectron.Theimageismadeusing[5].moduleismechanicallysturdy,about2mmthickandveryrobust.Thesetupcannotonlybeoperatedwithsymmetricvoltagesappliedtoalllayersbutalsowithasymmetricvoltages,meaningthelayeractsasancollectorandthemultiplicationhappensonlyinthesecondandthirdTHGEM(Fig.2.6).ThispresentstheadvantagethattheavalanchecanbewithintheM-THGEM(section2.1.3,5).Thedimensionsareequivalenttoasingle-layerTHGEM,thereforethetotalthicknessofthethreelayerM-THGEMis1.8mm.2.3SignalformationWhenchargedparticlestravelthroughtheTHGEM,theycreateasignalintheelectrodestheypassby.Theinducedsignaldependsonthechargeoftheparticleanditsvelocityrelativetotheweighting11TheRamo-Shockleytheoremdescribesthecurrentinducedintheelectrode(IR)[15]:IR=qEWR~v(2.1)IRcurrentintheelectrodeqelectricchargeEWRweightingoftheelectrodevvelocityofthechargewithrespecttotheweightingTheweightingisastheinducedbyoneunitofpotential(positive)onthereadoutelectrode,whenallotherelectrodesareatzeropotential[15].Theinducedcurrentintegratedovertimeequalsexactlythechargeoftheparticle,inthiscaseanelectron.Asaresult,themultiplicationwithintheTHGEMcanbemeasured,integratingthecurrentinducedinoneTHGEMelectrode(thelast),giventheprimarynumberofelectronsisknown.12Chapter3Simulations3.1FieldsimulationTheelectrichastobecalculatedbeforetheelectronandiondriftcanbesimulated.ThereforesimulationareperformedinANSIIMaxwell11[16].Thiselectricdataisthenusedin[5]todeterminetheexpectedmultiplicationoftheTHGEMs.3.1.1TwocascadeTHGEMsBetweenthetwoelectrodesofoneTHGEMtheissimilartoadipoleMorepreciselytwohalfdipoleseparatedbyasectionwithparallellines(Fig.3.1).FarawayfromtheTHGEMthedrifthasparallellines.InFig.3.1thetopedgeofoneTHGEMisshown.AniondriftingawayfromtheavalanchecreatedwithintheTHGEMwouldfollowthelinesshown,dependingonitsvelocity.Itisclearthatinsomecasesitwillcollidewiththecopperstratawherethedipolelinescurveintothesurface,inothercasesitwillleavetheTHGEM.ThereforethecollectionofionsisdeterminedbytheshapeoflinesontopoftheTHGEM,drivenbytheratiobetweenthedriftandthedipoleThisratioalsodeterminestheelectroncollection.Forthetransferbetweenthe13Figure3.1:SimulationoftheelectricaldipoleontopofaTHGEM(left).PlotoftheelectricalmagnitudeswithinthewholestructureoftwocascadeTHGEMs.Thedriftissetto0.5kVcm,theinductionissetto1.5kVcmandthetransferis1.5kVcm.ThevoltageacrossoneTHGEMis1500V.twoTHGEMs,itisimportanttoagoodcompromisebetweenahighelectronextractionusingastrongtransfercombinedwithaweakdipoleandahighcollectioninthesecondTHGEM,whichisobtainedwithalowtransferandahighdipoleItwouldbeclearhowtodeterminetheoptimaltransferstrengthinthisscenario,butthesecondaryionizationduetothephotonmediatedalsoplaysatrole.Ifthetransferistoostrong,theavalancheextendsoutoftheholeandphotonsreachthecoppersubstrate,causingtheTHGEMtotransitionintonon-proportionalmodeandeventuallystreamermode(discharge).143.1.2Three-layerM-THGEM3.1.2.1SymmetricvoltagesForthethreelayerM-THGEMtheinvolvingthetransferdonotexist,be-causeallelectronsandionsaretransferredtothenextlayer.InFig.3.2,thesimulatedFigure3.2:SimulationoftheelectricstrengthwithinathreelayerM-THGEM(left).ThevoltageacrossalllayersoftheM-THGEMisthesame.Plotoftheelectricstrengthalongaverticalcutthroughtheplotontheleft,showingtheenhancementinthecenteroftheM-THGEM(right).strengthwithinathreelayerM-THGEMisshown.ItisremarkablethatthereisaldenhancementwithinthemiddleelementofthethreelayerM-THGEM.Sincethemultipli-cationscalesquadraticallywithintheelectricthisecttheavalancheinthemiddleoftheM-THGEM(section5).Thistsuppressesthephotonmediatedsecondarybecauseallphotonsareabsorbedbytheisolator.153.1.2.2AsymmetricvoltagesInordertoachieveanevenhighert,thethreelayerM-THGEMcanbeoperatedinanasymmetricvoltageration,wherethevoltageacrosstheTHGEMisloweredto3oftheoriginalvoltage.ThisgivestheadvantageofacompletelysuppressedphotonfeedbackonthetopoftheTHGEMandahigherioncollection.Itwillbediscussedindetailinsection5.Figure3.3:SimulationofthestrengthwithinathreelayerM-THGEM.Ontheleft,thevoltageappliedtoalllayersisequivalent,ontherightthetoplayerisreducedto3(70%).3.2Particledrift3.2.1RangeintheAT-TPC3.2.1.1SRIM(energylosscalculation)Inordertomeasurethegainandenergyresolution,itisimportanttoknowhowmuchenergyisdepositedinthegasvolumebyanionizingparticle,suchasanalphaparticlewith5.5MeV.Dependingonthegaspressurethealphaparticlemightnotbestoppedwithinthegasvolumeandthereforenotdeposititsfullenergywithintheactivevolume.Figure3.4showstheenergylossinelectronvoltperangstromdependingonthedistancetraveledintheAT-TPC.Therateoftheenergydepositionincreasesatlowerenergy(velocity)ofthealphaparticle.ThisBraggcurvewillbeobservedinreverseform,whenmeasuringtheenergy16Figure3.4:RangewithintheAT-TPCusingpureHeliumat400torr.Theenergylossperangstromisplottedasfunctionofthepenetrationdepth.The3.6mthickpara-aramidwindowistakenintoaccount.depositionasafunctionoftime.TherangesimulatedinSRIMfortpressuresinpressurerange200torr73.5cm400torr37cm600torr24.5cmTable3.1:RangeofalphaparticlesinheliumsimulatedusingSRIMheliumispresentedinTab.3.1.Dependingontheangleofemission,analphaparticlewillnotbestoppedwithintheactivevolumeforagaspressureof200torrand400torr,becausetheradiusoftheAT-TPCisonly25cm.3.2.1.2MCNPX(alternateenergylosscalculation)Forhydrogen,therangeofalphaparticlesissimulatedusingMCNPX[17].Thissoftwaretakesthegeometryofthedetectorandtheangulardistributionofthesourceintoaccount.17AgraphicdemonstrationofonesimulationisshowninFig.3.5.Figure3.5:RangewithintheAT-TPCusingpureHydrogenat400torr.TheresultingenergydistributionisshowninFig.3.6.Figure3.6:DistributionoftheeneregydepositioninhydrogenwithintheAT-TPC,takingintoaccountthewindowandangledistributionofthesource.Thepara-armidwindowisalsoincludedinthesimulation.TherangeinH2isalittle18bitshorterthaninhelium.ThevaluesfortpressuresareshowninTab.3.2.Sincepressurerange200torr64cm400torr32cm600torr21.5cmTable3.2:RangeofalphaparticlesinhydrogensimulatedusingMCNPX.MCNPXalsoincludestheangulardistributioninitscalculation,itispossibletocalculatetheexpectedenergydistributionofalphaparticlesdetectedintheAT-TPCusinganAmericium-241source(Fig.3.6).3.2.2ElectroncollectionAhighelectroncollectionisimportanttoobtainfulltracksandagoodenergyresolution.ItismainlydeterminedbytheratiobetweenthedriftandtheatthelayeroftheTHGEM.Therefore,thecollectionismostlyindependentofthegeometricFigure3.7showsthecollectionfortasymmetricvoltageofthethreelayerM-THGEM.Inordertoobtaineachdatapointthedriftlinesof200bunches(each143electrons)havebeencalculated,andthefractionofthoseenteringtheTHGEMplotted.Theerrorbarsaregivenbythestandarddeviationofthe200bunchesofelectrons.Itisclearlyvisiblethatwithahigherdipolethecollectionincreases.Foroperationinpurehydrogen,thecollectioniscloseto100%foralldriftnormallyusedduetothehighdipolenecessaryformultiplication.Butforheliumitcanbettomakesurethattheratiobetweenthedipoleandthedriftallowsforahighcollection,sincethiswillensureagoodspatialresolution,andisessentialfortheenergyresolution.Ifpartoftheprimaryelectronsarelosttheenergyresolutionwillbedegraded.19Figure3.7:Collectionforathree-layerM-THGEMinpureheliumat200torr.Thepercentagesgivendescribethepotentialinthelayercomparedtotheotherlayersfortasymmetricvoltage3.2.3Gain3.2.3.1tgeometriesandgasesFigure3.8givesanoverviewofallthetgeometriesandgasesusedinthiswork.ItshowsclearlythatathreelayerM-THGEMachievesthesamegainasatwocascadeTHGEMwithatlylowervoltagepermodule.Also,theingainbetweenthesymmetricandasymmetricvoltageinthethreelayerM-THGEMisap-parent.Thisoverviewalsoshowsthatforoperationinpurehydrogen,therequiredvoltage20istlyhigher.Themaximumachievablegainisnotgivenbytheendofthecurves,Figure3.8:Gaincurvesforallntgeometriesdiscussedinthiswork.Thedataisgeneratedusingg[5]atapressureof600torrforallcurves.butratherbythemaximumavalanchesize,whichcanbecalculatedinareasonableamountofcomputingtime.Allgainsareplottedagainsttheassociatedreducedbias,whichistheappliedvoltagetooneTHGEMelementdividedbythepressureintorr.3.2.3.2TwocascadeTHGEMsForeachdatapointatotalof1000completeavalanchesweresimulated,trackingeveryionandelectronproducedintheavalanche.Theroutine\avalanche"wasusedin[5].The1000avalanchesweresplitinto10bunchesof100avalanches.Theerrorshownisgiven21bythestandarddeviationofthese10bunches.TheelectricmapwascalculatedinFigure3.9:Gaincurvesfortpressuresinpurehelium,usingatransferof0.25kVcmandaninductionanddriftof0.5kVcm.Maxwell,thenthegasdriftdatamodeledinMagboltz,andtheactualdriftperformedin[5].Itisimportanttousetherightunitcells,withthecorrectsymmetriesandasuitableintegrationstep.ItisalsoimportanttouseanarrowenoughmeshfortheelementsimulationinMaxwell.Acompletemanualonhowtoperformthesimulationsisgivenintheappendix(sectionC).doesnotincludetheofphotonmediatedfeedback,aquencherorthePenningThereforethesesimulationsdonotfullymodelreality.Whencomparing22Figure3.10:Gaincurvesfortpressuresinpurehydrogen,usingatransferof0.25kVcmandaninductionanddriftof0.5kVcm.thesimulatedresultsforpureheliumtothemeasurements(section4),thesimulationusinghelium(Fig.3.9)showsasignitlylowergain,mostlikelyduetoaresidueofnitrogenintheactivevolume.NitrogenandheliumareaPenningmixture(section2.1.1.2),thusalreadysmallamountsofnitrogencansigtlyincreasethegain.Whencomparingtheresultsforhydrogentothemeasurements(section4),theoppositeisobserved.Thesimulatedgains(Fig.3.10)arealothigherthenthemeasuredones,mostlikelyduetoresidualoxygenintheactivevolumefromoutgassingofcomponentsorleaks,becauseoxygenfunctionsasaelectroncapturinggas(section2.1.1.3).Adirectcomparison23forbothgasesisgivenintheappendix(sectionA).Thesedeviationsaside,thesimulationsshowtheexpectedexponentialbehaviorverywell,andfromthesmallspreadofthegainforasinglevalue(errorbars),itcanbeassumedthattheenergyresolutionispotentiallyverygood.ThevaluesfortheweresettovaluescommonfortheAT-TPCtogiveanimpressionofthegaintobeexpectedinoperation(transfer=0.25kVcm,inductionanddrift=0.5kVcm).3.2.3.3Three-layerM-THGEMThesimulationsforthethreelayerM-THGEMaredoneinexactlythesamewayasforthetwocascadeTHGEMs,exceptfromusingatgeometry.Inthiscase,thevaluesforthedrifteldandinductionweresettoensureoptimaloperation.Whenexaminingthepresentedresults,itisimportanttokeepinmindthattheydonotshowanyinformationaboutthemaximumachievablegain.ThethreelayerM-THGEMhasalowergainthanthetwocascadeTHGEMsforthesametotalvoltagebecausetheTHGEMlayeroperatesmainlyasanelectroncollector.24Figure3.11:Gaincurvesfordtpressuresinpurehelium,usinganinductionanddriftof0.5kVcm.25Figure3.12:Gaincurvesfortpressuresinpurehydrogen,usinganinductionof1.5kVcmanddriftof0.1kVcm.Thegraphincludesgainsforthetworentoperationmodes,symmetricvoltagesandasymmetricvoltagesTHGEM70%).263.2.4Ionbacw3.2.4.1TransferandionfeedbackForthefollowingdiscussion,theionbacw(IBF)isasthenumberofionscollectedatthecathodedividedbythenumberofelectronsattheanode.IBF=#ionsatcathode#electronsatanodeThedetectorpadsaretheanode.Thesimulationsfortheionbacowareperformedsimilartotheonesforthegain,buthereitisimportanttogive[5]thecorrectexperimentalvaluesforthetransportpropertiesofgaseousionssincethesearenotprovidedbyMagboltz.TheiondriftdatawasobtainedfromRef.[18]andRef.[19].Figure3.13showstheFigure3.13:Plotoftheiontracks(red)andelectrontracks(orange)inoneavalancheforasetupwithtwocascadeTHGEMs.TheplotsaregeneratedinH2at400torrwithavoltageof1025VacrosseachTHGEMusing[5].Thepotentialare:drift0.1kVcm,induction1.5kVcm,1.5kVcm(left)and0.125kVcm(right).27Figure3.14:CorrespondingtoFig.3.13thehistogramsontheleftshowtheendpointsforionsandelectronsusingahightransferld,whereasontherightthesevaluesareshownforaweaktransfer2Dplotofonesingleavalanche(electronsinorange,ionsinred).Theisolatingsubstrate(yellow)isbetweenthecopperstrata(brown).Noteverylayershowstheholebecauseofthe2Dprojection,thusshowingthewholeouterrimofthehole.Theclearlyshowstheuenceofthetransfer,likementionedinsection3.1.1.Averyhightransfer28(1.5kVcm)transfersallionsandelectrons.InthatcasetheionsarecollectedonlyontopoftheTHGEM.InbetweenthetwoTHGEMsthetransferisrelativelystrongcomparedtothedipoleld.Foralowtransfer(0.125kVcm)ionsarealsocollectedontopofthesecondTHGEM,butsincetheextractionforelectronsfromtheTHGEMisthenloweraswell,someelectronsarelost.Sooveralltheionbacwisnotreduced,justtheoverallgainislowered.ThestatisticsonwhereelectronandionpathsendisshowninFig.3.14.Thestatisticsisdoneonlyfor10avalanches,andthetotalnumberofelectronsshownisaabsolutevalue(toevaluatethequalityofthestatistic).Toputthissimulationintoperspective,oneshouldverifythattheappliedstrongtransferdoesnotcausetheavalachetoextendfromthehole.Thiswouldleadtophotonmediatedsecondaryionizationandtransitionintothenon-proportionalmode.3.2.4.2oftherelativealignmentbetweenthetwoTHGEMsonelec-tronandiontransferInthesetupusingtwocascadeTHGEMs,theholesarenotnecessarilyaligned.InthepresentsetupoftheAT-TPCtheyarenotaligned.Whenthisisthecasethetransparencyforchargedparticlesislowered,stronglydependingonthedistancebetweenthetwoTHGEMsandthegaspressure.InsimulationsperformedbyBin-Longetal.[20]ithasbeenshownthatthealignmentoftheTHGEMsmakesasigtdintheionbacw.However,thisisnotthecaseforthepresentsetup.Allions/electronsaretransferredindependentofthealignmentduetothelargeholediameterandlongdistancebetweenthetwoTHGEMsincomparisontotheelectronsmeanfreepath().ThiscanbevisuallyseeninFig.3.15andFig.3.16wherenotasingleelectronislostontopofthesecondTHGEM,29Figure3.15:Plotoftheiontracks(red)andelectrontracks(orange)inoneavalanchefortwocascadeTHGEMs,whicharenotaligned.TheplotsaregeneratedinH2at400torrwithavoltageof1050VacrosseachTHGEMusing[5].Thedriftissetto0.1kVcm,theissetto0.125kVcmandtheinductionis1.5kVcm.Theleftpictureshowsa3Dviewandtherightdiagrama2Dillustration.whereitcouldpossibly\crash"intothesecondTHGEMduetothemisalignment.Inthe2DgraphofFig.3.15,itisimportanttokeepinmindthattheionsandelectronswhichseemtogothroughthematerialareactuallygoingthroughaholebehindthefront.3.2.4.3CollectionofIonsinthethree-layerM-THGEMSincethereisnotransferinthethreelayerM-THGEM,rathertheTHGEMsaredirectlyattachedtoeachother,allionsandelectronsaretransferredindependentoftheappliedvoltages,aslongastheyareappliedsymmetrically.Onesingleevent(avalanche)isshowninFig.3.17toillustratethisbehavior.Lookingatthestatisticsfor1000electrons,comparingthethreelayerM-THGEMincollectormode(asymmetricvoltages)andsymmetricvoltages,itisremarkablethatasmall30Figure3.16:Statisticsoftheendpointsforions(right)andelectrons(left)inthetwocascadeTHGEMs,whicharenotalignedlikeshowninFig.3.15.TheplotsaregeneratedinH2at400torrwithavoltageof1050VacrosseachTHGEMusing[5].Thedriftissetto0.1kVcm,theissetto0.125kVcmandtheinductionis1.5kVcm.fractionofionsisactuallyabsorbedwithinthethreelayerM-THGEMincollectormode,whereasallelectronsarestilltransferred.31Figure3.17:Plotoftheiontracks(red)andelectrontracks(orange)inoneavalancheforathree-layerM-THGEM.TheplotsaregeneratedinH2at400torrwithavoltageof650VacrosseachTHGEMusing[5].Thedriftissetto0.1kVcmandtheinductionis1.5kVcm.Theleftpictureshowsa3Dviewandtherightdiagrama2Dillustration.32Figure3.18:CorrespondingtoFig.3.17thehistogramsontheleftshowtheendpointsforionsandelectronsusingahighsymmetricvoltageof650VforeachTHGEM,whereasontherighttheTHGEMisworkingasacollector(asymmetricvoltageion).Itissetto70%ofthefullvoltage.333.2.4.4Ionfeedbackvs.GainfortFigure3.19displaystheionbacwversusthegain.ItshowsthatthethreelayerM-THGEMisbetterthanthetwocascadeTHGEMsintermsoflowionbacw.ThethreelayerM-THGEMincollectormodeprovidesthelowestionbacw.Theadvantagesofthisnewgenerationofdetectorsarediscussedinsection5.ForthetwocascadeTHGEMsitFigure3.19:Ionbacw(numberofionsatthecathodeovernumberelectronsattheanode)plottedagainstthegainforathree-layerM-THGEMandtwocascadesinglelayerTHGEMsinH2at400torr.ThedecreaseinIBFforhighergainisclearlyvisibleforeachsetup.Theerrorisindicatedasthe˙/2spreadofthesimulation.isinterestingthatthehightransfer(collectionontopoftheTHGEMonly)hasactuallyalowerionbacwthanthewiththelowtransfer(collection34onbothTHGEMs).Thisisduetothefactthatincaseofthelowtransferelectronsarealsolost(whichwouldotherwisebemulitpliedinthesecondTHGEM),andthereforethegainislower.35Chapter4MeasurementsforthetwocascadeTHGEMs4.1ExperimentalSetupTheexperimentalsetupconsistsoftheAT-TPCdetector(Fig.1.1),whichhasaradiusof25cmandalengthof100cm.Thereisathinpara-aramid(3.6m)windowwithinthecathodetowhichacollimatedAmericium-241alphasourceisattached.TwocascadeTHGEMsaremountedattheanode,whilethesecondlayerofthesecondTHGEMiscon-nectedtothereadoutelectronics.Thegaspressurecanberegulatedwithintheactivevolume.Thedriftcanbesetusingahighvoltagepowersupply.4.1.1GashandlingsystemThegashandlingsystemiscustommadeandconsistsofaDatametricsDresser1404valvecontrollerandanelectronicmultimeter(1400Datametrics)toregulatethepressurewithinthedetectorvesselandasecondequivalentsetuptocontrolthenitrogenpressureintheoutervolume.Theconstantgaswisregulatedbyawmeterconnectedtothevacuumlineforbothvolumes.364.1.2VoltagepowersuppliesThehighvoltagesourcefortheexperimentisaHeinzingerPNChp100000-1negpowersupply,connectedtothecathode.4.1.3ReadoutelectronicsTomeasurethegain,thesignalinducedinthelastTHGEMisused(section2.3).ThissignalisintegratedusingachargeORTECModel109A.Thischargemeasuresthenumberofcollectedelectrons.Figure4.1showsthecalibrationcurveforthechargemeasuredusingaRC-circuitwitha3.3pFcapacitorandafunctiongenerator.TheintegratedsignalisdigitizedusingaFASTERmodule[21].TheFASTERanalog-to-digitalconverter(ADC)providesrentoptions,includingpuls-shaping(33sconstant),fastdatarecordingandingtorecordalargenumberofsingleevents.Itispossibletoaddalinearintegrationoruseitasamulti-channelanalyser.Figure4.1:CalibrationcurvesforthechargeforthetwotTheslopeofthetwocurvesare6electronsbinand60electronsbinrespectively.374.2PerformanceinHelium4.2.1GainUsingtheFASTERmoduleasmulti-channelanalyzerthepeakofthesignalismeasuredandconvertedtoanumberofelectronsusingthecalibration.AnalyzingthesimulationgivenFigure4.2:Comparisonofthegaininpureheliumfortpressures.Thehorizontallinesindicatethemaximumachievablegain.InadditionthreemeasurementsbyMarcoCortesi[9]usingUV-Lightareincluded.bySRIM(section3.2.1.1),theaverageenergylossinthewindowis48104eV1010mwithathicknessof3:6m(thethicknessissmallerthanthehistogramintervalofSRIM(1cm))thereforethetotalenergylosswithinthewindowis0.48MeV.Thetotalenergylossinthegasistherefore5.02MeV(alphaparticlefullystopped).Usingtheion-electronpairproductionenergyof46.3eVgiveninRef.[22],thenumberofprimaryelectronsisabout108,000.The38egainwascalculatedusingthisnumber.Themainerrorsareduetothecalibrationofthereadoutelectronicsandtheangulardistributionofthesource.Also,theremightbesomeresidualgas,whichcanmodifythegain.4.2.2EnergyresolutionTwothousandeventswererecordedusingtheFASTERmodule.TheyareshowninFig.4.3.Onealphaparticleeventsandtwoalphaparticleevents,meaningtwoarrivingattheFigure4.3:Visualizationfor2000eventsinheliumat600torr(left).Theenergyspectrumcalculatedfromtheseevents(right).sametime(overlapping),areclearlyvisible.TheFWHMofthedistributionis8.5%,whenconstructingthehistogramfromintegratedsignalforeachsingleevent.Thatdoesn'tmeantheenergyresolutionis8.5%becausetheenergydistributionitselfhasalsoawidth,buttheenergyresolutionisatleast8.5%.Theintegrationtimeismarkedinred.Themainerrorinthiscaseisthetriggeraswellastheabovementioned.394.3PerformanceinHydrogen4.3.1GainSimilarlytotheprocedurewithhelium,thesamecalculationcanbedoneforhydrogenusingtheenergydepositioncalculatedbyMCPNXandshowninFig.3.6.Figure4.4:Comparisonofthegaininpurehydrogenfortpressures.Theendpointofeachcurveindicatesthemaximumachievablegainatthatpressure.404.3.2ElectrondriftvelocityUsingthetimemeasurementsinFig.4.5.Thedriftvelocitycanbecalculated.Itistodeterminetheexactpulselengthforthisdata,sotheerrorisatleast10%.pressuredriftvelocity100torr10105cmsec200torr9:14105cmsec300torr9105cmsec400torr8105cmsecTable4.1:DriftvelocityofelectronsinhydrogenaccordingtoFig.4.5.Figure4.5:Visualizationof1000eventsinhydrogenfortpressures.414.4FurtherinterpretationofmeasurementsTovalidatethesemeasurements,onecananalyzetheslopeofthegaincurvesforentpressures.Neglectingsecondary(photonmediatedts,spacechargeandat-tachment),thegainG(electronmultiplication)canbedescribedusingtheTownsendcot[9]:G=edThethicknessdofthemultiplicationzoneisproportionaltothenumberofTHGEMs(#THGEM),sinceallTHGEMsusedhavethesamethickness.ForlowreducedelectricFigure4.6:Comparisonofthegaininpurehydrogenfortpressures.Thecurvesinthisgraphareusedtotheparameterp.(smallerthan20Volt/torr/cm),theTownsendcocientisproportionaltothe42appliedvoltageV[7].Therefore:G=ed/eV#THGEMFittingallcurvesinFig.4.4(Fig.4.6)withG/epV/eV#THGEMtheparameterponlydependsonthegeometryofthep/#THGEMThereforeplottingtheresultsfoundinsection4againstthevoltageinsteadofthereducedbiasshouldleadtostraightlineswiththesameinclination,sinceforG=cepVthelogarithmicplotislog(G)=log(c)+log(e)pV(Fig.4.6).Thevaluesdonotpreciselymatch,makingitclearthatthereissomeerrorinthemeasurementsortheassumptions,forthebehaviorofthegainaretosimple.TheresultsarepresentedinTab.4.2.pressurep100torr0.02381V200torr0.02051V300torr0.01531V400torr0.01441V500torr0.01351VTable4.2:pcalculatedfromthelinearforthecurvesinFig.4.6.43Chapter5TowardsanewgenerationofM-THGEMs:Three-layerM-THGEMThethree-layerM-THGEMwasplacedinthePrototypeActive-TargetTimeProjectionChamber(pAT-TPC)[1]inordertoinvestigateitsperformance.ThepAT-TPChasadi-ameterof28cm,alengthof50cm,andthesameentrancewindowasthefullsizeAT-TPC(3.6mpara-armid).TheareacoveredbytheM-THGEMhasadiameterof25cm.Figure5.1showsaschematicdrawingsandaphotographoftheM-THGEMelectrodemountedontopofthepAT-TPCreadoutelectronics.ThesamereadoutelectronicandcollimatedFigure5.1:Aschematicdrawing(left(topview)andtopright(sideview))andaphotograph(bottomright)ofthemountedM-THGEM.241-Amwasused,asinthefullsizeAT-TPC.Figure5.2showsanexampleofanenergyspectrummeasuredinHe/(10%)CO2(blackgraph),comparedtotheexpectedspectrumcomputedbyaMonteCarlosimulationus-44ingMCNPX(bluegraph).Thesharppeakcorrespondstotheenergyreleasedbyparticlesemittedintheforwarddirection,whichcrossthewholeactivevolumealongtheaxis.Thetotalenergydepositedinthegasinthiscaseis2.4MeV,withameasuredenergydisper-sionof5%(comparedtothe4.5%computedwithMCNPXsimulations).Thelow-energytailofthespectrumcorrespondstoalphaparticlesemittedatlargerangleswithrespecttothecylindricalaxis.Thesealphaparticlesexitthedetectoractivevolumeandstopontheexternalwallsofthecage.Thesmallbetweenthecomputedandthesimulatedspectrumismostlikelyduetouncertaintieswhenmodelingtheexactgeometryofthecollimatorforthe241-AmsourceinMCNPX.Figure5.2:Spectrumofa5.5MeValphaparticlecrossingthepAT-TPCactivevolume,recordedwiththethreelayerM-THGEMoperatedinHe-basedmixture(10%CO2).45Figure5.3:GainmeasurementforthethreelayerM-THGEM(schematicdrawing(a))inHe/(10%)CO2(b)andinpureHe(c).Amaximumgainof104:5wasobservedforbothgases.Thedecreaseofmaximumachiev-ablegainathighpressureisduetothesmalleralphaparticlesrange.Thisresultsindenserionization(multiplicationinashortertimeinterval),thereforethemaximumavalanchesizeisalreadyreachedatlowergain.Conversely,atverylowpressure(below100torr)thedetectorshowslargeinstabilitieswhenrelativehighvoltageisappliedtotheM-THGEMelectrodes.OperatingtheM-THGEMwithareducedvoltageforthestagehasseveraladvan-tages.TheavalancheissqueezedtowardsthelowerregionoftheM-THGEM,preventingphoton-mediatedfromthetopelectrodesurface.Inaddition,themultiplierstagethenactsascollectorofpositivecharges,sothattheionbacwtothedriftregionisreduced.UsingMaxwellthemapforasymmetricallybiased(Fig.5.4(a))andanasymmetricallybiased(reducedto3=70%)(Fig.5.4(b))M-THGEMwerecalculated.ThegaincurvesoftasymmetricM-THGEMsetups,measuredin200torrHe/10%CO2(Fig.5.4(b)),shifttowardshigherreducedbiasbutreachthesamemax-imumachievablegain.Whilefullelectroncollectionency(Fig.5.5(a))isreachedatalowhole-to-driftratioof20(correspondingtoagainofafewhundred),theasymmetric46setupshowsaslightlybettercollectionofionsintheM-THGEMelectrode,reducingtheionbackw.Theseresultsweresimulatedusing[5],asdescribedinsection3.Amulti-cascadeelementisexpectedtoreducetheIBFfurther,below10%.Figure5.4:ComparisonoftheelectriceldmapforathreelayerM-THGEMholesym-metricallybiased(a)andasymmetricbiased(b).3representsthereducedfractionoftheappliedpotentialtothemultiplierstage.Measurementsofegainin200torrHe/10%CO2fordtthreelayerM-THGEMFigure5.5:Comparisonoftheelectroncollectiony(a)andionbacw(b)inHe/10%CO2fordtasymmetricM-THGEM47Chapter6ConclusionandoutlookTheTHGEMisahole-typegaseouselectronmultiplierproducedbymultilayerprintedcircuitboard(PCB)technology.Itconsistsofadenselyperforatedassemblyof0.6mmthickFR-4substrate,sandwichedbetweenthinmetallicelectrodestrata.TwoTHickGaseousElectronMultipliers(THGEM)incascadewerefoundtooperateatagainofupto104inpureheliumforpressuresbetween200torrand600torr,whilealsoprovidingagoodenergyresolution.InpurehydrogenitwasnotpossibletotestthelimitoftheTHGEMsetupbecausethefeedthroughintheAT-TPCwasnotdesignedforvoltagesabove1.5kVandthereforelimitedthemaximumvoltageapplied.Neverthelessagainof102:5wasmeasuredatapressureof100torr.ThreeTHGEMsweresandwichedtogethertoformanovelgaseousmultiplier(M-THGEM),providingahightoftheelectronavalanchewithinthehole,resultinginagainofupto104:5inpurelow-pressurehelium.DuetothegeoemtricstructureoftheM-THGEM,theldisenhancedinthemiddleTHGEM,theavalanchetothatarea.ThethreelayerM-THGEMalsoshowedaverygoodenergyresolutionof5%,comparedtothe4.5%computedwiththeMCNPXsimulation.In[5],simulationsoftheionbacwwasfoundtobeaslowas13%incertainconditions.Furthermore,thisnoveldetectorisextremelysuitabletocoverlargeareasduetoitsrobustnessandthicknessof1.8mm.ThismakesthethreelayerM-THGEMaperfectdevicefortheoperationofalargeAT-TPC.InthenearfuturethePenningshouldbeincludedinthesimulations,sothatthese48matchtheexperimentalresultsforpurehelium.TheAT-TPCshouldbeequippedwithathreelayerM-THGEMtoenablehighgaininlowpressurehydrogenandincreasetheenergyresolution,aswellasdecreasetheriskoffailureduetothetwoTHGEMstouching(mechanicalbending).ForfurtherdevelopmentoftheM-THGEMitmightbepossibletouseceramicinsteadofFR-4,toreducethechargingupAfutureprojectcouldbetoapplycaesiumiodide(CsI)tothetoplayer,whichwouldemitelectronsduetoX-raysorphotons.TheseelectronscouldthenbemultipliedintheM-THGEM.Inthatoperationmodeonewouldhavetomakesurethattheionbacwdoesn'thitthetoplayerbychoosingtheappropriatestrengthorgeometryandpossiblyuseacascadesetup.49APPENDICES50AppendixAAdditionalgraphsFigureA.1:Comparisonbetweentheexperimentaldata(Fig.4.2)inheliumforthetwocascadeTHGEMsandthesimulatedresults(Fig.3.9).ThetdiscrepancycanbeexplainedbythemissingPenningin[5].51FigureA.2:Comparisonbetweentheexperimentaldata(Fig.4.4)inhydrogenforthetwocascadeTHGEMsandthesimulatedresults(Fig.3.10).Thetdiscrepancycanbeexplainedbyresidueoxygen,capturingfreeelectrons.52AppendixBTechnicaladviceThefollowingisalistoftechnicaltipsinordertocompletetheexperimentsuccessfully:WhenscrewingstainlesssteelscrewsinstainlesssteelNEVERuseanautomaticscrew-driver!Bestiswhenyoueven(iftheplaceallowsit)putsomelubricantwiththescrew,topreventiteatingintothematerialGroundallmeasuringdevicestwiceoreventhreetimesinordertoreducenoise.Groundingcablescanbeusedorevenaluminiumfoilworkswell.Whenpumping(vacuum)bepatient!Itisbesttopumpbelow500mtorr(thisisalotlessinactualgaspressure,becausethesystemshowsthepressureswronginpurenoblegases).Themaximumpressureofthegashandlingsystemis700torr.Whenyoutrytoputmore,anoverpressurevalvewillopenandyousimplyloseallyourgastotheenvironment.Makesuretoalsovacuumalltubesbeforwithgas,havedetectorchamberclosedwhiledoingsowhenyouaredoingitrightbeforetheoperationduetoleaksinthetube.53AppendixCManualforperformingsimulationsofTHGEMswithintheNSCLThereisademofolderstoredinthegroupdriveofProf.DanielBazin,whichmakestheuseofthismanualeasier,butitisnotnecessary.C.1OverviewBeforethedetaileddescriptionisgivenhereaquickoverviewofwhatishappening.Theaimistoautomatesimulationsformanytgaspressuresandvoltagesasfaraspossibleandthenruntheminparallelonabatchsystem.Theideaistocreatthenecessaryfolders,thenrunANSIIMaxwell11andperformthesimulationinldusingMagboltz.Requiredsoftware:Windows7(itistestedonwindows7butshouldalsorunonotheroperationsystems)9(thisisinstalledonthetankandonthebatchsystem(seaside))ANSIIMaxwell11(notanewerversion)theinstallationisprovidedinthefolder,andthelicenseserverfornewerversionsworksMobaXterm_v8.6oranyotherlinuxterminal(ssh)54notepad++Origin2016oranyotherplottingsoftwareThroughouttheprocessitwillbenecessarytoadapttheandpathsineachscript,whendoingsopleasebesuretoincludethecorrect\/"or\\"withit,dependingonwetherthissppartwillberunninginLinuxorWindows.Thesuggestionistousenotepad++,sinceitisfastandt(multipletabs).C.2CreatingthefoldersandrunGotothefolder:I:\analysis\attpc\demofolder_simulation\triple_layer_voltage_30_handrunthecreate_all_voltages.bat.Thiscreatesafolderforeachvoltageandcopiestherequiredintoit.Itthenrunsavisualbasicscriptinordertoperformethemaxwellsimulation.Withtheloopthevoltagesaresp(\from",\step",\to").createallvoltages.bat1setlocalenabledelayedexpansion2for/l%%nin(1000,25,1025)do(3mkdir%%nVsymauto4rmdir/S/Qthgem1.mxwlresults5@remnewdirectoryiscreatedandoldcurrentdirectoryinmaxwellisdelted6@rempause7@remtheVBscriptformaxwelliseditedsothatitrunsthecurrentvoltage,thisworksbyfindingtheentry999andreplacingit8settxtfile=fullscript2edit.vbs9setnewfile=fullscript%%nV.vbs10ifexist!newfile!del/f/q!newfile!11for/f"tokens="%%ain(!txtfile!)do(12setnewline=%%a13setnewline=!newline:^999=^%%n!14echo!newline!>>!newfile!15)16@rempause17@remrequiredfilesaremovedintothecurrentfolder5518movefullscript%%nV.vbs"I:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hn%%nVsymautonfullscript%%nV.vbs"19copycopymaxwell.bat"I:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hn%%nVsymauton"20copyrun"I:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hn%%nVsymauton"21copygarftrans"I:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hn%%nVsymauton"22@rempause23@remecho!volt!24@remtheVBscriptisrun25cdI:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hn%%nVsymauto26start/wfullscript%%nV.vbs27cdI:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hn%%nVsymauto28callcopymaxwell.bat29@remallfilesneedeforgarfieldarecopiedfrommaxwell30cdI:nanalysisnattpcndemofoldersimulationntriplelayervoltage30h31)32pauseThescripttakesthevisualbasicscript(fullscript_2_edit.vbs)andreplacesthevoltageinitwiththeonespintheloop.C.3RunningMaxwellandcreatingmapsFigureC.1:\recordscript"functioninMaxwellThefullscript_2_edit.vbscanbecreatedinMaxwellbyusingthecommand56\recordscript"sothatitgivesytobeusedfortgeometriesandpurposes.Thescriptshouldincludeeditingoneormoreglobalvariables,whichthenappearinthevisualbasic(VB)scriptandcanbeeditedbythecreate_all_voltages.bat.Thereforethevariabletoeditshouldbesetto999,sothatthecreate_all_voltages.batscriptcanit.Alsothenamesofthemaxwellfolderpath,andmaxwellfolderhavetobeuptodateinthescriptandneedtobechangedwhenitiscopiedtoanewfolder(newsetofsimulations).fullscript2edit.vbs1'2'ScriptRecordedbyMaxwellVersion11.1.13'10:02PMFeb08,20164'5DimoAnsoftApp6DimoDesktop7DimoProject8DimoDesign9DimoEditor10DimoModule11SetoAnsoftApp=CreateObject("AnsoftMaxwell.MaxwellScriptInterface")12SetoDesktop=oAnsoftApp.GetAppDesktop()13oDesktop.RestoreWindow14oDesktop.OpenProject"I:/analysis/attpc/demofoldersimulation/triplelayervoltage30h/thgem1.mxwl"15SetoProject=oDesktop.SetActiveProject("thgem1")16SetoDesign=oProject.SetActiveDesign("MaxwellDesign1")17oProject.ChangePropertyArray("NAME:AllTabs",Array("NAME:ProjectVariableTab",Array("NAME:PropServers",18"ProjectVariables"),Array("NAME:ChangedProps",Array("NAME:$voltagesym","Value:=","999"))))19oProject.Save20oDesign.AnalyzeAllNominal21SetoModule=oDesign.GetModule("FieldsReporter")22oModule.CalcStack"clear"23oModule.EnterQty"Voltage"24oModule.CalculatorWrite25"I:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hn999Vsymautonphi.reg",Array("Solution:=",26"Setup1:LastAdaptive"),Array("$voltagesym:=","999","a:=","1mm","d:=",27"0.5mm","h:=","0.1mm","t:=","0.6mm")28oModule.CalcStack"clear"29oModule.EnterQty"E"30oModule.CalculatorWrite31"I:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hn99957Vsymautone.reg",Array("Solution:=",32"Setup1:LastAdaptive"),Array("$voltagesym:=","999","a:=","1mm","d:=",33"0.5mm","h:=","0.1mm","t:=","0.6mm")34oModule.CalcStack"clear"35oModule.EnterQty"D"36oModule.CalculatorWrite37"I:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hn999Vsymautond.reg",Array("Solution:=",38"Setup1:LastAdaptive"),Array("$voltagesym:=","999","a:=","1mm","d:=",39"0.5mm","h:=","0.1mm","t:=","0.6mm")40oProject.Save41oDesktop.CloseProject"thgem1"Thisscriptalsosavesthethreeneededfrommaxwellfor[5](thepotential,theelectricandtheelectricIncaseitshouldbedonemanually,dothefollowing:rightclickonoverlays->calculator->input->quantities->E-Field/D-Field/Voltage->output->write.Thesearealsothestepstobefollowedwhencreatinganewscript.OncetheVBscriptiscreate_all_voltages.batrunsthenextscriptcopy_maxwell.battocopythethreefromthemaxwellfolderelds.shd,current.hyd,current.pnt)intothecurrentvoltagefolder.copymaxwell.bat1echoCopyfilesstarts:2echofjxcopy/Y/S"I:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hnthgem1.mxwlresultsnMaxwellDesign1.resultsncurrent.hyd"current.hyd3echofjxcopy"I:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hnthgem1.mxwlresultsnMaxwellDesign1.resultsncurrent.pnt"current.pnt/Y/S4echofjxcopy"I:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hnthgem1.mxwlresultsnMaxwellDesign1.resultsnfields.shd"fields.shd/Y/S5@rempauseWheneverthelenameofthemaxwellischanged,thefolderpathchangesandthisthenneedstobechangedincopy_maxwell.bat.Nextcreate_all_voltages.batcopiestherequiredfor[5]intothecurrent58voltagefolder(run,garftrans).create_all_voltages.batwithdeletingthecurrentmaxwellworkingdirectory.Thisisveryimportantotherwisetheupwithmoretrashoneveryrun.Itwillnowstartthenextiterationofthe\for"loop.C.4RunontankNowthatallthenecessaryareinthefolder,itisgoodtodoatestrunifeverythingworkswithvisualoutputontankbeforerunninghundredsofjobsontheseasidesystem.OpenMobaXterm_Personal_8.6.exeandstartanewsession,nowselectSSHandenterFigureC.2:StartMobaxtermandenterhostname.thehostnametank.nscl.msu.edu",clickOK.NowloginwithyourNSCLusernameandpassword.Gotothedirectorywhereyouareworking,ifyouworkinthesharedanalysisdrivewherethedemofolderis,yougettherebyentering:1cd/mnt/analysis/attpc/demofoldersimulation/triplelayervoltage30h/59Nowstartbyentering:1moduleloadgarfield92garfield9314<runThereisaespeciallyforrunningwiththegraphicinterface,thegeometryinfor-mationcanbeincludedatthedesignatedposition,sothatthegraphicsoutputshowstheTHGEM.runview.txt1<garftrans2GLOBALnum=13GLOBALpresmin=1004GLOBALpresmax=7005GLOBALpresstep=10067//justassignedforloopprecheck8Globalbin`temp.bin`9GlobalEHeHe=010GlobalKHeHe=011Globalstatus=012Globalcode=01314Vectorpressures15400161718ForpresInpressuresDo19ForkFrom1Step1TonumDo20!OPTNOGRID21!opttimestamp22!optlineary23!replabelstextcolourblacktextfonthigz24!repnumberstextcolourblacktextfonthigz25!reptitletextcolourblacktextfonthigz26!repcommenttextcolourblacktextfonthigz27!repmessagetextcolourforegroundtextfonthigz28!layoutxnumber=0.015,xdecade=0.025,...29ynumber=0.007,ydecade=0.015,...30xlabel=0.010,ylabel=0.010,...31title=0.01032GlobalplotdriftTrue33Globald`d.reg`34Globale`e.reg`35Globalphi`phi.reg`6036Globalbin`3Dfpresg.bin`37Globalexistfalse38&CELL39cellid"Gain"40Callinquirefile(bin,exist)41IfexistThen42readfieldmapfbing43Else44fieldmapfilesfphigfegfdgMAXWELL11...45xmirrorperiodicymirrorperiodic...46plotmap47say"geometrycreated"48savefieldmapfbing49say"filesaved"50Endif51solids52ForxFrom0.08664Step0.08662To0.08664Do53ForyFrom0.054Step0.052To0.054Do54holecentrefx,yg0.03005direction001...55halflengths0.050.050.0005...56radius0.035...57n=5...58conductor359holecentrefx,yg+0.00direction001...60halflengths0.050.050.03...61radius0.025...62n=5...63dielectric64holecentrefx,yg+0.03005direction001...65halflengths0.050.050.0005...66radius0.025...67n=5...68conductor369holecentrefx,yg+0.0601direction001...70halflengths0.050.050.03...71radius0.025...72n=5...73dielectric74holecentrefx,yg+0.09015direction001...75halflengths0.050.050.0005...76radius0.025...77n=5...78conductor379holecentrefx,yg+0.1202direction001...80halflengths0.050.050.03...81radius0.025...82n=5...83dielectric84holecentrefx,yg+0.15025direction001...85halflengths0.050.050.0005...86radius0.035...87n=5...88conductor389Enddo6190Enddo91ForxFrom0.08665Step0.08662To0.08665Do92ForyFrom0.055Step0.052To0.055Do93holecentrefx,yg0.03005direction001...94halflengths0.050.050.0005...95radius0.035...96n=5...97conductor398holecentrefx,yg+0.00direction001...99halflengths0.050.050.03...100radius0.025...101n=5...102dielectric103holecentrefx,yg+0.03005direction001...104halflengths0.050.050.0005...105radius0.025...106n=5...107conductor3108holecentrefx,yg+0.0601direction001...109halflengths0.050.050.03...110radius0.025...111n=5...112dielectric113holecentrefx,yg+0.09015direction001...114halflengths0.050.050.0005...115radius0.025...116n=5...117conductor3118holecentrefx,yg+0.1202direction001...119halflengths0.050.050.03...120radius0.025...121n=5...122dielectric123holecentrefx,yg+0.15025direction001...124halflengths0.050.050.0005...125radius0.035...126n=5...127conductor3128Enddo129Enddo130131132say">>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>1"133134//135136Globalgasfile`Hgasfpresg`137Globalgasmember`exbfpresg`138Globalp=pres139Globalpbar=pres/750.06//(pinbar)(oftablegiven)140Globalt=300141142&GAS14362144Callinquiremember(gasfile,gasmember,`gas`,exist)145IfexistThen146getfgasfile,gasmemberg147Else148pressurefpgTorr149temperatureftg150magboltzH2100...151e/prange0.001135.152writefgasfile,gasmemberg153Endif154155//ADDIonmobility156say"magboltzfinished>addionmobility"157VectorEHeHeKHeHe1580.0016.01594.0016.01605.0016.01616.0016.01628.0016.016310.016.016412.016.016515.015.916620.015.816725.015.716830.015.516940.015.217050.014.917160.014.517280.013.917310013.417412013.217515013.117620013.117725013.217830013.317940013.7180181GlobalEHeHe=EHeHe/(0.010354300)182GlobalKHeHe=KHeHe1e6/pbar183Callfitexponential(EHeHe,KHeHe,1e8,p0,p1,p2,p3,ep0,ep1,ep2,ep3)184addionmobilityexp(fp0g+fp1gep+fp2gep^2+fp3gep^3)185//extrapolationslowionmobilityconstanthighionmobilitylinear186addionmobilityKHeHevsEHeHe187188say">>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>2"189190Globalpitch=0.01191Globalpitchx=sqrt(3/4)pitch192Globaln=50193Globaltot=n194Globalstatus,code195Globaltd196Globalxd197Globalyd63198Globalzd199Globalnd200Globalewelll=0201Globalmultip202Globaltimep203Globalions=0204Globalelect=0205Globalconta=0206207&DRIFT208area0.20.20.20.20.20.5CUT...209viewy=0rotate180.210view6y+2x+3z=03Drotate90.211integrationparametersmcdistint0.001212Callbookhistogram(elec,50,0,2500)213Callbookhistogram(ion,50,0,2500)214Callbookhistogram(created,50,0.02,+0.02)215Callbookhistogram(lost,50,0.02,+0.02)216Callbookhistogram(ende,50,0.201,+0.501)217Callbookhistogram(endion,50,0.201,+0.501)218ForiFrom1Step1TonDo219Globalz=0.4220Globalx=0.0866+(RNDUNIFORM20.0866)221Globaly=0.05+(RNDUNIFORM20.05)222223Callplotdriftarea224Callavalanche(x,y,z,`plotelectron,plotion`,ne,ni,...225`ycreated`,created,`ylost`,lost,`ze`,ende,`zion`,endion)226Callhistogramtomatrix(endion,a,min,max)227Callhistogramtomatrix(ende,b,min,max)228Say"Electrons:fneg,ions:fnig,ionfeedback:fa[50]g,electronfeedback:fb[1]g,fraction:fa[50]/b[1]g,gain:fb[1]/ng(avalanchefig)"229Callfillhistogram(elec,ne)230Callfillhistogram(ion,ni)231Globalions=ions+ni232Globalelect=elect+ne233If(ne<1)Then234Globalconta=conta+1235Endif236Enddo237Callfitexponential(elec,a,b,ea,eb,`plot`)238Say"Slope:f1/bg"239!optionslogy240Callplothistogram(elec,`Electrons`,`Numberofelectronsafteravalanche`)241Callplotend242Callhplot(ion,`Ions`,`Numberofionsproducedinavalanche`)243Callplotend244Callhplot(created,`y[cm]`,`Productionpointofelectrons`)245Callplotend246Callhplot(lost,`y[cm]`,`Absorptionpointofelectrons`)247Callplotend248Callhplot(ende,`z[cm]`,`Endpointofelectrons`)249Callplotend250Callhplot(endion,`z[cm]`,`Endpointofions`)64251Callplotend252253>ionbackflowfpresg.txt254Say"Electrons:felectg,ions:fionsg,ionfeedback:fa[50]g,electronfeedback:fb[1]g,fraction:fa[50]/b[1]g,gain:fb[1]/ng,numbernotdetected=fcontag,totalnumber=fng,collectionefficiency=f(nconta)/ng,run=fkg"255>256Enddo257Enddo258&MAINHopefullyallcommentingwithintheisclearandselfexplaining.Forusingatwocascadesetuporatwocascadesetupwhichisnotalignedthefollowingtwogeometriescanbeused.Theonlythingtokeepinmindwhencreatinggeometriesisthatthereisalimitationontheamountof\forloops"inganotalignedgeometry.txt1solids2ForxFrom0.08664Step0.08662To0.08664Do3ForyFrom0.054Step0.052To0.054Do4holecentrefx,yg0.0005direction001...5halflengths0.050.050.0005...6radius0.035...7n=5...8conductor39holecentrefx,yg+0.031direction001...10halflengths0.050.050.03...11radius0.025...12n=5...13dielectric14holecentrefx,yg+0.0615direction001...15halflengths0.050.050.0005...16radius0.035...17n=5...18conductor319Enddo20ForyFrom0.055Step0.052To0.055Do21holecentrefx,yg+0.2625direction001...22halflengths0.050.050.0005...23radius0.035...24n=5...25conductor326holecentrefx,yg+0.293direction001...27halflengths0.050.050.03...28radius0.025...29n=5...30dielectric6531holecentrefx,yg+0.3235direction001...32halflengths0.050.050.0005...33radius0.035...34n=5...35conductor336Enddo37Enddo38ForxFrom0.08665Step0.08662To0.08665Do39ForyFrom0.055Step0.052To0.055Do40holecentrefx,yg0.0005direction001...41halflengths0.050.050.0005...42radius0.035...43n=5...44conductor345holecentrefx,yg+0.031direction001...46halflengths0.050.050.03...47radius0.025...48n=5...49dielectric50holecentrefx,yg+0.0615direction001...51halflengths0.050.050.0005...52radius0.035...53n=5...54conductor355Enddo56ForyFrom0.054Step0.052To0.054Do57holecentrefx,yg+0.2625direction001...58halflengths0.050.050.0005...59radius0.035...60n=5...61conductor362holecentrefx,yg+0.293direction001...63halflengths0.050.050.03...64radius0.025...65n=5...66dielectric67holecentrefx,yg+0.3235direction001...68halflengths0.050.050.0005...69radius0.035...70n=5...71conductor372Enddo73Enddoalignedgeometry.txt1solids2ForxFrom0.08664Step0.08662To0.08664Do3ForyFrom0.054Step0.052To0.054Do4holecentrefx,yg0.0005direction001...5halflengths0.050.050.0005...6radius0.035...7n=5...668conductor39holecentrefx,yg+0.031direction001...10halflengths0.050.050.03...11radius0.025...12n=5...13dielectric14holecentrefx,yg+0.0615direction001...15halflengths0.050.050.0005...16radius0.035...17n=5...18conductor319holecentrefx,yg+0.2625direction001...20halflengths0.050.050.0005...21radius0.035...22n=5...23conductor324holecentrefx,yg+0.293direction001...25halflengths0.050.050.03...26radius0.025...27n=5...28dielectric29holecentrefx,yg+0.3235direction001...30halflengths0.050.050.0005...31radius0.035...32n=5...33conductor334Enddo35Enddo36ForxFrom0.08665Step0.08662To0.08665Do37ForyFrom0.055Step0.052To0.055Do38holecentrefx,yg0.0005direction001...39halflengths0.050.050.0005...40radius0.035...41n=5...42conductor343holecentrefx,yg+0.031direction001...44halflengths0.050.050.03...45radius0.025...46n=5...47dielectric48holecentrefx,yg+0.0615direction001...49halflengths0.050.050.0005...50radius0.035...51n=5...52conductor353holecentrefx,yg+0.2625direction001...54halflengths0.050.050.0005...55radius0.035...56n=5...57conductor358holecentrefx,yg+0.293direction001...59halflengths0.050.050.03...60radius0.025...61n=5...6762dielectric63holecentrefx,yg+0.3235direction001...64halflengths0.050.050.0005...65radius0.035...66n=5...67conductor368Enddo69EnddoC.5Runontheseasidebatchsystem-createjobTorunmanytvoltagesatthesametimetheseasidebatchsystemcanbeused.Touseiteachseperatjobmusthaveitsownjobthesearecreatedwithcreate_job_files.batandcanthenbesubmittedasonesinglecreatejob1setlocalenabledelayedexpansion2for/l%%nin(1000,25,1025)do(3settxtfile=jobstandard.sh4setnewfile=job.sh5ifexist!newfile!del/f/q!newfile!6for/f"tokens="%%ain(!txtfile!)do(7setnewline=%%a8setnewline=!newline:^999=^%%n!9echo!newline!>>!newfile!10)11@remreplacingallthe999inthejobfilewiththecurrentvoltageandthencopyingit12movejob.sh"I:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hn%%nVsymautonjob.sh"13copyrun"I:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hn%%nVsymauton"14@remcopyingtherunfileincaseitchanged15cdI:nanalysisnattpcndemofoldersimulationn16@remcreatingthefilewhichwillbesubmitted17echocd/mnt/analysis/attpc/demofoldersimulation/triplelayervoltage30h/%%nVsymauto/>>submit.sh18echoqsub/mnt/analysis/attpc/demofoldersimulation/triplelayervoltage30h/%%nVsymauto/job.sh>>submit.sh19cdI:nanalysisnattpcndemofoldersimulationntriplelayervoltage30hn20)68create_job_files.bateditsthejob_standard.shforeachvoltageandcopiesitthenintothespfolder.Forthisalsotheandfolderpathsaswellasthevoltagesneedtobechangedandmatchwithallotherscripts.jobstandard.sh1#pbstemplatescript234###Setthejobname56##PBSNTarget789###HavePBSmailyouresults1011#PBSMrost@nscl.msu.edu121314#Emailgeneratedatb)eginning,a)bort,ande)ndofjobs1516#PBSmbae1718#PBSlnodes=11920#PBSlwalltime=72:00:00212223###Combinestdout/stderr2425###Note,thisoutputdirectorymushexist,orjobwillfailwith:2627###AbortedbyPBSServer2829###Jobcannotbeexecuted3031###Also,don'tuse"~"inpath,alsoseemstofail.3233#PBSjoe3435#PBSo$PBSJOBID.$PBSJOBNAME.out363738moduleloadgarfield9394041#gotowhereqsubwasrunfrom42cd$PBSOWORKDIR434445#ormaybe694647#cd/mnt/analysis/attpc/Rost/..4849printenv505152535455cd/mnt/analysis/attpc/demofoldersimulation/triplelayervoltage30h/999Vsymauto/#changesfolderpathinlinux56#thejinthenextlineisvitalbecauseitpassesontherunfilesothatitreadlinebyline57catrunjgarfield9Foreachoftheseanentryismadeintheforsubmissiontotheseasidesystem.FigureC.3:Correctsettingsinnotepad++,tomodifytheforsubmittingtothebatchsystem.Thecorrectlineending\LF"isshown.submit.shcannotbesubmittedtotheseasidesystem.Thelinebreakhastobechangedfromwindowstolinuxformat.Thisstepiscrucialandcannotbeskipped.notepadinitssettingsothatitlookslikeFig.C.3.Basicallychangethelineendingsettingtolinuxandenablethatlineendingsaredisplayed.Nowcreateanew(submit_linux.sh)70FigureC.4:Wrong(windows)linebreakintheforsubmittingtothebatchsystem.andcopyallthecontentfromsubmit.shintothenewItshouldnowonlysay\LF"atthelineending,likeinFig.C.3,not\CRLF"likeinFig.C.4.Logontotheseasidesystem:openMobaXterm_Personal_8.6.exeandstartanewsession,nowselectSSHandenterthehostname\seaside.nscl.msu.edu",clickOK.NowloginwithyourNSCLusernameandpassword.Gotothedirectorywheresubmit_linux.shislocated.Ifyouworkinthesharedanalysisdrivewherethedemofolderis,yougettherebyentering:1cd/mnt/analysis/attpc/demofoldersimulation/Nowsimplytype1submitlinux.shandpressenter.Awholebunchofjobswillbesubmitted,youcanviewyourcurrentstatuswith1qstataajobcanbedeletedwith1qdel[jobID]71Hint:Ifyouhavetodeletemanyjobs,doanexcelsheetandcopypasteitintotheterminal(withrightclick).C.6ExtractingthedatawithMatlabresultiontransport.m1clearall;2closeall;3addpath('H:nMyDocumentsnMatlabscriptsnmatlabfrag')%notneededunlessusingmlf2pdfscript(notneeded)4save='symmetricvoltagefield500Vpercm';%notparticularalyimportantcanbechangedlaterwhensaving5pfad='I:nanalysisnattpcndemofoldersimulationntriplelayervoltage30h';%filepath>hastobechanged6gasstep=200;%definitionofstepsforgasandvoltages7gasstart=200;8gasfinish=600;9voltagestart=300;10voltagefinish=1300;11voltagestep=25;12gasp=gasstart:gasstep:gasfinish;%definitionovectorsneededwhenreadingfiles13voltage=voltagestart:voltagestep:voltagefinish;14multir=zeros(size(voltage,2),size(gasp,2));15effir=zeros(size(voltage,2),size(gasp,2));16forivoltage=1:size(voltage,2);%redingfiles(votlages)17strvoltage=num2str(voltage(ivoltage));18fullpfad=[pfad'n'strvoltage'Vsymauto'];%creatingfielpathwithinloop>editwhenfilenamechanges19forigasp=1:size(gasp,2);20ifgasp(igasp)==75021gasp(igasp)=760;22end23strgas=num2str(gasp(igasp));24name=[fullpfad'nionbackflow'strgas'.txt'];25ifexist(name,'file')%readingoffileswhichexist26k=0;27m=0;28fid=fopen(name);29tline=fgetl(fid);30%tline=fgetl(fid);%thislinecanbedeleted31whileischar(tline)32%fori=1:200;33k=k+1;34chararray(k)=textscan(tline,'%s','Delimiter','nt,:');35forp=1:size(chararrayfkg,1)36tempcharstrm=chararrayfkgfpg;7237test='fraction';38ifstrcmp(tempcharstrm,test)39m=m+1;40multir(ivoltage,igasp,m)=str2num(chararrayfkgfp+3g);41effir(ivoltage,igasp,m)=str2num(chararrayfkgfp+1g);42end43end44tline=fgetl(fid);45end46fclose(fid);47end48ifmod(igasp,2)==049Progress=igasp/size(gasp,2)%thisprogressloopisnotcompletebutgivesanideathatthefileisrunning50end51end52end53display('readingofdatacompleted')54%%55forg1=1:1:size(multir,2);%calculatingallthemeanvaluesandstandarddiviationfortheruns56forv1=1:1:size(multir,1);57tempmultifull=multir(v1,g1,:);58tempeffifull=effir(v1,g1,:);59tempeffi=tempeffifull(tempeffifull~=0);60tempmulti=tempmultifull(tempmultifull~=0);61multistd(v1,g1)=std(tempmulti,1,3);62effistd(v1,g1)=std(tempeffi,1,3);63multimean(v1,g1)=mean(tempmulti,3);64effimean(v1,g1)=mean(tempeffi,3);65end66end67effimean(isnan(effimean))=0;%makingsurethereareno"emptyfields"68effistd(isnan(effistd))=0;69multimean(isnan(multimean))=0;70multistd(isnan(multistd))=0;71%multistd=std(multir,1,3);72%effistd=std(effir,1,3);73%multimean=mean(multir,3);74%effimean=mean(effir,3);75%%76maxmulti=5E6;%cutofffordata>editasneeded77startcutoff=1;78plotstep=200;79plotstart=gasstart;%outputselection80plotfinish=gasfinish;81graphnumber=0;82forgasevl=plotstart:plotstep:plotfinish;83%gasevl=150;84gasposition=(gasevlgasstart)/gasstep+1;85cutoff=0;86start=1;7387voltagenumber=0;88fort1=1:size(voltage,2)89ifmultimean(t1,gasposition)<maxmulti90ifmultimean(t1,gasposition)>startcutoff91voltagenumber=voltagenumber+1;92plotmulti(voltagenumber,1)=voltage(t1);93plotmulti(voltagenumber,2)=multimean(t1,gasposition);94plotmulti(voltagenumber,3)=multistd(t1,gasposition);9596ploteffi(voltagenumber,1)=voltage(t1);97ploteffi(voltagenumber,2)=effimean(t1,gasposition);98ploteffi(voltagenumber,3)=effistd(t1,gasposition);99end100end101end102ifexist('ploteffi');%plottingandsaving103graphnumber=graphnumber+1;104color=rand(1,3);105ifgasevl==750106gasevl=760;107end108figure(1)109errorbar(ploteffi(:,1),ploteffi(:,2),ploteffi(:,3),'','color',color)110gridon;111holdon;112legd1fgraphnumberg=['gaspressure:'num2str(gasevl)'torr'];113xlabel'voltage[V]';114ylabel'ionfeedbackIa/Ic';115fig1=gcf;116%mlf2pdf(1,['multiplication'num2str(gasevl)]);117118figure(2)119errorbar(plotmulti(:,1),plotmulti(:,2),plotmulti(:,3)3,'','color',color)120set(gca,'yscale','log')121gridon;122holdon;123legd2fgraphnumberg=['gaspressure:'num2str(gasevl)'torr'];124xlabel'voltage[V]';125ylabel'multiplication';126fig2=gcf;127csvwrite(['finalmulti'num2str(gasevl)'torr.csv'],plotmulti);%savingascsvdata128csvwrite(['finalion'num2str(gasevl)'torr.csv'],ploteffi);129%mlf2pdf(1,['efficiency'num2str(gasevl)]);130end131clearploteffiplotmulti%neededsothatnosizeerroroccureforthenextpressure132end133figure(1)134legend(legd1,'Location','southeast')135holdoff13674137figure(2)138legend(legd2,'Location','eastoutside')139holdoff140clearlegd1legd2graphnumber%needetocreateanewtitleandlegendnextrunAfterextractingthedataintospreadsheets,itcanbeplottedwithanygoodplottingsoftware,forexampleOrigin.C.7Summary1.Prepareallbyupdatingthefolderpathandnames2.runcreate_all_voltages.bat3.runcreate_job_files.bat4.editthelinebreakinsubmit.sh5.logontoseaside.nscl.msu.eduandsubmitsubmit_linux.sh75BIBLIOGRAPHY76BIBLIOGRAPHY[1]D.Suzuki,M.Ford,D.Bazin,W.Mittig,W.G.Lynch,T.Ahn,S.Aune,E.Galyaev,A.Fritsch,J.Gilbert,F.Montes,A.Shore,J.Yurkon,J.J.Kolata,J.Browne,A.Howard,A.L.Roberts,andX.D.Tang.PrototypeAT-TPC:Towardanewgenera-tionactivetargettimeprojectionchamberforradioactivebeamexperiments.NuclearInstrumentsandMethodsinPhysicsResearchSectionA:Accelerators,Spectrometers,DetectorsandAssociatedEquipment,691:39{54,November2012.[2]Nationalsuperconductingcyclotronlaboratory,michiganstateuniversity,eastlansing,mi48824,usa.[3]M.S.Wallace,M.A.Famiano,M.-J.vanGoethem,A.M.Rogers,W.G.Lynch,J.Clif-ford,F.Delaunay,J.Lee,S.Labostov,M.Mocko,L.Morris,A.Moroni,B.E.Nett,D.J.Oostdyk,R.Krishnasamy,M.B.Tsang,R.T.deSouza,S.Hudan,L.G.Sobotka,R.J.Charity,J.Elson,andG.L.Engel.Thehighresolutionarray(HiRA)forrareiso-topebeamexperiments.NuclearInstrumentsandMethodsinPhysicsResearchSectionA:Accelerators,Spectrometers,DetectorsandAssociatedEquipment,583(2-3):302{312,December2007.[4]A.H.Wuosmaa,J.P.ScB.B.Back,C.J.Lister,andK.E.Rehm.Asolenoidalspectrometerforreactionsininversekinematics.NuclearInstrumentsandMethodsinPhysicsResearchSectionA:Accelerators,Spectrometers,DetectorsandAssociatedEquipment,580(3):1290{1300,October2007.[5]R.Veenhof.adriftchambersimulationprogram.1998.[6]YuI.Davydov.OntheTownsendcotathighelectricd.arXivpreprintphysics/0409156,2004.[7]LorneM.ChaninandG.D.Rork.ExperimentaldeterminationsoftheTownsendionizationcotinhelium.PhysicalReview,133(4A):1005,1964.[8]B.Azmoun,W.Anderson,D.Crary,J.Durham,T.Hemmick,J.Kamin,G.Karagiorgi,K.Kearney,G.Keeler,E.Kornacki,andothers.AstudyofgainstabilityandcharginginGEMfoils.InNuclearScienceSymposiumConferenceRecord,2006.IEEE,volume6,pages3847{3851.IEEE,2006.77[9]M.Cortesi,J.Yurkon,andA.Stolz.OperationofaTHGEM-baseddetectorinlow-pressureHelium.JournalofInstrumentation,10(02):P02012,2015.[10]Th.GSchutandJ.ASmit.AdirectmeasurementoftheofthepPhysica,10(6):440{444,June1943.[11]FabioSauli.Gaseousradiationdetectors:fundamentalsandapplications.Cambridgemonographsonparticlephysics,nuclearphysicsandcosmology.CambridgeUniv.Press,Cambridge,2014.[12]FabioSauli.Principlesofoperationofmultiwireproportionalanddriftchambers.Lec-turesgivenintheacademictrainingprogrammofCERN1975-1979.CERN-Serviced'Informationscien-RD/233,Geneva,1977.[13]FabioSauli.Gaseousradiationdetectors:fundamentalsandapplications.Cambridgemonographsonparticlephysics,nuclearphysicsandcosmology.CambridgeUniv.Press,Cambridge,2014.[14]M.Cortesi,V.Peskov,G.Bartesaghi,J.Miyamoto,S.Cohen,R.Chechik,J.M.Maia,J.M.F.dosSantos,G.Gambarini,V.Dangendorf,andothers.THGEMoperationinNeandNe/CH4.JournalofInstrumentation,4(08):P08001,2009.[15]O.Bouianov,M.Bouianov,R.Orava,P.Semenov,andV.Tikhonov.ProgressinGEMsimulation.NuclearInstrumentsandMethodsinPhysicsResearchSectionA:Accelerators,Spectrometers,DetectorsandAssociatedEquipment,450(2):277{287,2000.[16]AnsoftCorporation.Ansoftmaxwell3dsimulator,version11.FourStationSquareSuite200Pittsburgh,PA15219,USA.[17]OAKRIDGENATIONALLABORATORY.Mcnpx2.4.0.LosAlamosNationalLabo-ratory,LosAlamos,NewMexico.[18]H.W.Ellis,R.Y.Pai,E.W.McDaniel,E.A.Mason,andL.A.Viehland.Transportpropertiesofgaseousionsoverawideenergyrange.AtomicDataandNuclearDataTables,17(3):177{210,March1976.[19]H.W.Ellis,E.W.McDaniel,D.L.Albritton,L.A.Viehland,S.L.Lin,andE.A.Mason.Transportpropertiesofgaseousionsoverawideenergyrange.PartII.AtomicDataandNuclearDataTables,22(3):179{217,September1978.[20]Bin-LongWang,QianLiu,Hong-BangLiu,Xiao-KangZhou,ShiChen,Dong-ShengGe,Wen-QianHuang,Yi-GangXie,Yang-HengZheng,YangDong,QiangZhang,Xin-78DaJiao,JingWang,MinLi,andJieChang.IonTransportationStudyforThickGasElectronMultipliers.ChinesePhysicsLetters,31(12):122901,December2014.[21]FranceLaboratoiredePhysiqueCorpusculairedeCaen.Fasterproject.website:http://faster.in2p3.fr/.[22]SimonWedlundetal.Comprehensivecalculationoftheenergyperionpairorwvaluesforemajorplanetaryupperatmospheres.AnnalesGeophysicae,29(1):187{195,January2011.79