The Assembly of the Belle II TOP Counter BoqunWang,OnbehalfoftheBelleIIPIDGroup DepartmentofPhysics,UniversityofCincinnati,Cincinnati,OH,USA UniversityofCincinnatipreprintUCHEP-14-01 Abstract Anewtypeofring-imagingCherenkovcounter,calledTOPcounter,hasbeendevelopedforparticleidentificationattheBelleII 5experimenttorunattheSuperKEKBacceleratorinKEK,Japan.Thedetectorconsistsof16identicalmodulesarrangedazimuthally 1aroundthebeamline. TheassemblyprocedureforaTOPmoduleisdescribed. Thisprocedureincludesacceptancetestingofthe 0 quartzmirror,prism,andquartzbarradiators. Theacceptancetestsincludeachipsearchandmeasurementsofbulktransmittance 2 andtotalinternalreflectance. Theprocessforaligningandgluingtheopticalcomponentstogetherisdescribed. n aKeywords: Cherenkovdetector,TOPcounter,BelleII J 3 11. Introduction ] ThetimeofpropagationdetectornamedTOPcounter[1,2, t e3, 4, 5, 6] is the particle identification detector developed for dtheBelleII/SuperKEKBexperiment[7]. AsshowninFigure1, - soneTOPmoduleconsists2.7mlongquartzbarradiatorwitha n sphericalmirrorononeendandaprismwedgeontheotherend. i .ThetimeofpropagationoftheCherenkovphotonstravelingin- s csidetheradiatorisdetectedbyanarrayofmicro-channel-plate iphotomultipliertubes(MCP-PMTs)[8,9]attachedtotheprism. s yThe Cherenkov ring image is reconstructed as the pattern of hthedetectedtimeandpositionoftheCherenkovphotons. The p quartz bar radiator must transmit the Cherenkov photons over [ longopticallengthwithalargenumberofinternalreflections. 1Therefore,highbulktransmittanceandtotalinternalreflectance v arerequired. Thequartzbarsshouldbeopticallyuniform,and 9 thenthelevelofdamageonthesurfacesandedgesmustbelow. 0 Figure1:OverviewoftheTOPcounter. Thelengthforonemoduleis270cm 0Thegoalbeingthatthephotondetectorsdetectasmanyphotons andthewidthis45cm. Itconsistsoftwo125cmlongquartzbars,one10cm 3aspossibletoreachthedesignK/πseparationpower. longmirrorwithfocallengthas5m,andone10cmlongprismwitha∼18◦ 0 Atotalof16identicalTOPmoduleswillbeinstalledonthe tiltedangle . 1Belle II detector. Therefore 32 quartz bars, 16 mirrors and 16 0prisms should be manufactured, tested and glued together. In face flatness, surface roughness, parallelism, perpendicularity 5thispaper,wewilldescribethefabricationprocessoftheTOP 1 and chamfer specs will be measured and guaranteed by ven- counter;theacceptancetestresultsforthequartzbars,mirrors : dors. The metrology report containing the properties above is vand prisms after their delivery, which includes the bulk trans- providedwiththeopticalcomponents. Ximittance and internal reflectance for bars, the tilted angle for The quartz bars will be delivered directly to KEK, Japan. prism and the focal length and reflectivity for mirror; and the r After delivery, the surface and edge quality will be roughly aalignmentandgluingproceduretoassembleallopticalcompo- checked by eyes then carefully checked by an automatic chip nentstogetherforthefollowingbarboxsealing. searchsystem.Ifthesizeandnumberofthechipsandscratches satisfythespecification,thebarwillbeaccepted.Thenthebulk 2. Fabricationprocess transmittance and internal reflectance of the bar will be mea- suredbytheautomaticmeasurementsystem.Themeasurement Theopticalcomponents,whichincludesquartzbars,prisms resultswillbeusedforMCsimulation. and mirrors, are manufactured by several vendors. The sur- The mirrors and prisms will be delivered to University of Cincinnati, USA. Their surface qualities will be checked first Emailaddress:[email protected](BoqunWang,Onbehalfof theBelleIIPIDGroup) like the quartz bars. The transmission and reflectivity will be PreprintsubmittedtoNuclearInstrumentsandMethodsA January14,2015 measured,too. Formirrors,additionalpropertiesliketheposi- tionofopticalaxis, focalpoint, focallength, sphericalaberra- tionandastigmatismwillbemeasured.Forprisms,theangleof thetiltedsurfaceisaveryimportantpropertyandwillbemea- sured carefully. After the above acceptance tests, mirrors and prismswillbepackedanddeliveredtoKEK,Japan. In KEK, after receiving all the optical components, includ- ing two bars, one mirror and one prism, they will be mounted onsomespeciallydesignedgluingstages. Thelevelandangle of these stages are controlled by some micrometers, so these componentscouldbealignedveryprecisely. Withautocollima- torandlaserdisplacementsensor,theangleandleveldifference betweentwocomponentscouldbeverysmall. Afterthealignment,thegluingprocesscanstart. Somespe- cial jigs are designed to help the gluing. This includes an air dispensingsystem,andamechanicalgluingtrolleytotranslate the syringe. The adhesive for gluing is NOA63, which is UV adhesive. TheassemblyroominKEKisshowninFigure2. Thisisa clean room with a class 5000 buffer room outside and a class 1000darkboothinside. Twoopticaltablesstayinsidethedark Figure3: ThequartzbarmadebyZygo. Thebarismountedonthegluing booth.Thelargeroneisusedforthetesting,aligningandgluing stagewithsixmicrometers.Thegluingstageismountedonhighprecisionrails of the optical components. The gluing stages are mounted on whicharefixedontheopticaltable. twohighprecisionrailswhicharefixedonthetable. Afterthe gluing, the completed TOP module will be moved to another optical table by a specially designed handling jig and crane. Then the module will be sealed in the quartz bar box (QBB). Finally,theTOPmodulewillbemovedoutsidethecleanroom formountingelectronicsandcosmicraytesting. 3. Acceptancetests By the end of February, 2014, two bars were delivered to KEK(onemadebyZygoandtheotheronemadebyAperture- Figure4: Theacceptancetestsystemforquartzbar. Twobarsareputonthe stagesandthex-ystageswithlaserandphoto-diodeareinstalledatthetwo Okamoto). Four prisms (all made by Zygo) and one mir- endsofthebars. ror (made by ITT) were delivered to University of Cincinnati. Some tests are performed with these components and in this sectionsomepreliminaryresultswillbeshown. Thebulktransmittanceτisobtainedfromtheequation I (1−R )τ(1−R )= I (1) 0 0 1 1 3.1. Quartzbar whereI istheintensityoftheincidentlaser,I istheintensity 0 1 ThequartzbarmadebyZygoisshowninFigure3. Theac- after the transmission through the quartz bar, and R and R 0 1 ceptance test system for quartz bar is shown in Figure 4. The are the reflectance at the air-to-bar and bar-to-air surfaces re- quartzbarsareputonthestagesandthemeasurementsystemis spectively,whichcouldbecalculatedbyFresnelequation. The installedaroundthem.Thelaserandphoto-diodeareonthetwo wavelengthofthelaserweusedinthisandthefollowingmea- endsofthequartzbars, andtheyareusedtomeasurethebulk surementsare405nm. transmittanceandinternalreflectanceofthequartzbars.Thepo- Theinternalreflectanceαisobtainedfromtheequation sitionofthelaserandphoto-diodeiscontrolledbyx-ystages, √ which are controlled by a notebook computer. The measure- I (1−R )αNe−L/Λ 1+(bN/L)2(1−R )= I (2) 0 0 1 1 mentsystemisautomatic,whichmeansaftersettingupthesys- tem, the measurement will be finished automatically without whereI ,I ,R andR havethesamedefinitionsasbulktrans- 0 1 0 1 humanintervention. Afteroptimization,thescanspeedisquite mittance. N is the number of bounces, Λ is the attenuation fast. Wecanfinishthescanningof∼400pointswithin5h. lengthdefinedasthepathlengthinthebulkregioncorrespond- The methods for bulk transmittance and internal reflectance ingtothephotonretainmentof1/ewhichcouldbecalculated measurement are shown in Figure 5. The scanning points are from the result of the bulk transmittance measurement, and L alsoshowninFigure5. and b are the length and the thickness of the bar, respectively. 2 Figure2: TheTOPmoduleassemblycleanroominKEK.Outsideisaclass5000bufferroomandinsideisaclass1000darkbooth. Therearetwoopticaltables insidethecleanroom: onefortesting,aligningandgluing,andtheotheroneforQBBassembly. Acranewithaspeciallydesignedhandlingjigcouldmovethe TOPmodulefromonetabletoanother. Theexponentialfactorindicatesthephotonlossduetothebulk 3.2. Mirrorandprism transmission. ThefirstmirrormadebyITTandfourprismsmadebyZygo have been delivered to University of Cincinnati, as shown in Figure 7. The surface quality is good, and only a few small chipsarefound. Figure5: Themethodsoftheacceptancetests. Onthetopisthemethodfor bulktransmittanceandthescanpointdistribution.Onthebottomisthemethod forinternalreflectancemeasurement. Figure7: Themirrorandprism. Themirrorismountedonstageandthefour prismsareontheopticaltable. The results of the quartz bar acceptance tests are shown in Figure6. Forthezygobar,thebulktransmittancemeasurement The mirror has been tested for reflectivity and focal length. result is (99.66±0.07)%/m, and the internal reflectance mea- TheresultofthemirrortestsisshowninFigure8. Thereflec- surement result is (99.971±0.013)%. Some scratches on the tivity is (88.51±0.20)% for the data taken at 11.7 mm from quartz surface may affect the result. The results for Aperture- surfaceS1,and(88.73±0.18)%forthedatatakenat15.0mm Okamotobarare(99.54±0.08)%/mforbulktransmittanceand fromsurfaceS1. Thefocallengthis(4965.35±11.84)mmfor (99.982±0.013)%forinternalreflectance. Theerrorscontain the x > 0halfofthemirrorand(4967.55±12.55)mmforthe both statistical and systematic errors, which include the con- x<0halfofthemirror. Thelaserwavelengthis532nm. tributionfrombulktransmittancemeasurement, laserstability, The testing of the prism is still on going. The preliminary anglemeasurementandlaserpolarization. result of the tilted angle is (18.085 ± 0.003)◦. This result is 3 Figure6:Theresultsofthebulktransmittance(left)andinternalreflectancewithN =21inEquation2(right). Fourdifferentcolorscorrespondtofourscanning linesinsidethequartzbar.Thelargeuncertaintynear10cmiscausedbythescratchonthesurface. Figure8:Thetestingresultofmirrorreflectivity(left)andfocallength(right). consistent with the specification which is (18.07±0.04)◦ and thevendor’smeasurementwhichis18.07◦. 4. Gluing After testing, all components will be mounted on specially designedgluingstageswithmicrometerswhichcancontrolthe levelandangleofthestagesforalignment. Autocollimatorand laserdisplacementsensorareusedtomonitorthelevelandan- Figure9: Thealignmentforgluing. Thisshowstheconditionforprism-bar gledifferencebetweenthem. gluing.Forbar-barandbar-mirrorgluingthesetupissimilar. Afterthealignment,theopticalcomponentswillbegluedto- getherwithadhesives. TheadhesiveforthegluingisNOA63, between the optical components. If the gap is too small, the which is UV adhesive. For one TOP module, there are three adhesive will flow very slowly and it is very sensitive to the kindsofgluing: bartobar,bartomirrorandbartoprism. For surface roughness and cleanliness. This may generate lots of allofthem,thegluingsetupissimilar,whichisshowninFig- bubbles. Moretestsareneededbeforetheassemblyofthefirst ure9. TOPcounter. Testsforthegluinghavebeentakingplace. Aspeciallyde- signedmechanicaltrolleywasusedandprovedtobeusefulfor 5. Summary easing up the gluing. A few tests succeeded with barely no visible bubbles and others ended with some large and small TheTOPcounterisimportantforBelleIIdetector’sparticle bubbles. We tested the gluing procedure with different gaps identification capability. The mass production of the optical 4 components has started and the basic assembly procedure has been fixed. After delivery, the four components (two quartz bars,oneprismandonemirror)willbetestedseparatelyinthe University of Cincinnati and KEK. Then they will be aligned andgluedtogetherinKEKandreadyforQBBassembly. Fornow,therearetwoquartzbars,fourprismsandonemir- ror which are manufactured and delivered by vendors. Their properties are measured and the preliminary results show that they satisfy our specification. The assembly of the first TOP counter will start soon. Before that, further tests of the gluing procedureareneeded. References [1] M.Akatsu,etal.,NuclearInstrumentsandMethodsinPhysicsResearch SectionA440(2000)124. [2] T.Ohshima,NuclearInstrumentsandMethodsinPhysicsResearchSec- tionA453(2000)331. [3] T.Ohshima,ICFAInstrumentationBulletin.20(2000). [4] Y.Enari, etal., NuclearInstrumentsandMethodsinPhysicsResearch SectionA494(2002)430. [5] K.Inami,NuclearInstrumentsandMethodsinPhysicsResearchSection A595(2008)96. [6] K.Matsuoka,NuclearInstrumentsandMethodsinPhysicsResearchSec- tionA732(119)(2013)357. [7] T.Abe,etal.,KEK-Report-2010-1(2010). [8] M.Akatsu,etal.,NuclearInstrumentsandMethodsinPhysicsResearch SectionA528(2004)763. [9] K.Inami,etal.,NuclearInstrumentsandMethodsinPhysicsResearch SectionA592(2008)247. 5