JLC CCD Vertex JLC CCD Vertex Detector R&DDetector R&D

Y. SugimotoY. SugimotoKEKKEK

2003. 8. 192003. 8. 19

MembersMembers KEKKEK A. Miyamoto, K. Nakayoshi, Y. Sugimoto, A. Miyamoto, K. Nakayoshi, Y. Sugimoto, H. YamaokH. Yamaok

aa Niigata U.Niigata U. K. Fujiwara,K. Fujiwara, G. Iwai, G. Iwai, Y. Onuki,Y. Onuki, N. Tamura, N. Tamura, H. TH. T

akayama akayama Saga U.Saga U. T. Kuniya, K. Nakamura, T. Kuniya, K. Nakamura, K.D. Stefanov, T. TsukaK.D. Stefanov, T. Tsuka

motomoto Tohoku U.Tohoku U. T. Nagamine, T. Nagamine, Y. ShirasakiY. Shirasaki Tohoku Gakuin U.Tohoku Gakuin U. K. AbeK. Abe Toyama CollegeToyama College T. AsoT. Aso

of Maritime Tech.of Maritime Tech.

Red names: Graduated or left ( 4 Ms and 1 D)Red names: Graduated or left ( 4 Ms and 1 D)

ContentsContents IntroductionIntroduction

Possible Options Possible Options Why CCD? Why CCD?

R&D ProgramR&D Program What has been achieved?What has been achieved? What has been left to be done?What has been left to be done?

Future Prospects Future Prospects

Possible OptionsPossible Options Candidates for Vertex Detectors at Candidates for Vertex Detectors at

LCLC Silicon Strip Detector --- OccupancySilicon Strip Detector --- Occupancy Hybrid Active Pixel Sensor --- ThicknessHybrid Active Pixel Sensor --- Thickness Charge Coupled Device (CCD)Charge Coupled Device (CCD) Monolithic Active Pixel Sensor (CMOS)Monolithic Active Pixel Sensor (CMOS) Other New Ideas (DEPFET, Other New Ideas (DEPFET, SOI, etcSOI, etc.).)

DEPFETDEPFET

p+

p+ n+

rear contact

drain bulksource

p

sym

met

ry a

xis

n+

ninternal gate

top gate clear

n -

n+p+

--

++

++

-

MIP

50 µ

m

- -- ---~1µm

Why CCD?Why CCD? Mission: Show a design by the end of 2000 (ACFA RepMission: Show a design by the end of 2000 (ACFA Rep

ort)ort) Structure of CCDStructure of CCD

Diffusion of electrons in epitaxial layerDiffusion of electrons in epitaxial layer Key of excellent spatial resolution Key of excellent spatial resolution for CCD & CMOS pixel sensorsfor CCD & CMOS pixel sensors Takes time to diffuseTakes time to diffuse d = sqrt(Dt) ~ 6d = sqrt(Dt) ~ 6m @ t=10nsm @ t=10ns OK for JLC/NLC OK for JLC/NLC (Fully depleted CCD at TESLA)(Fully depleted CCD at TESLA)

CCD has simple structureCCD has simple structure Large area sensorLarge area sensor High yieldHigh yield

CCD is the most feasible optionCCD is the most feasible option

　　　　　　 CCD MAPS HAPS DECCD MAPS HAPS DEPFETPFET

ResolutionResolution AAAAAA AAA A A AAA A AThin material Thin material AAAAAA AAA C AA AAA C AARad. Hardness A(?) AA AAA AAARad. Hardness A(?) AA AAA AAA

(?)(?)Large waferLarge wafer AAA ? ? ?AAA ? ? ?

R&D ProgramR&D Program Design Criteria : Design Criteria : “ “The Highest Vertex Resolution with Technical The Highest Vertex Resolution with Technical

Feasibility”Feasibility”

High spatial resolution of the sensorHigh spatial resolution of the sensor Minimize multiple scattering Minimize multiple scattering Thin wafer Thin wafer Close to the IP Close to the IP Radiation Hardness Radiation Hardness Room temperature operation, if possibleRoom temperature operation, if possible

Spatial ResolutionSpatial Resolution Beam Tests in ’97 and ’98Beam Tests in ’97 and ’98

KEK PS T1 beam lineKEK PS T1 beam line 0.5 – 2.0 GeV/c pion0.5 – 2.0 GeV/c pion 4-CCD Telescope4-CCD Telescope CCD Samples: CCD Samples:

HPK 24 HPK 24 mm22 10/50 10/50 m epm epi. EEV 22 i. EEV 22 mm2 2 20 20 m epi.m epi.

Resolution better than Resolution better than 33m(r.m.s) was obtainedm(r.m.s) was obtained

Excellent spatial resolution of Excellent spatial resolution of CCD has been demonstratedCCD has been demonstrated..

Spatial Resolution Spatial Resolution (Cont.)(Cont.)

Resolution Study Resolution Study with Laser Beam with Laser Beam Scanner (Niigata U.)Scanner (Niigata U.) Beam spot size: 2Beam spot size: 2mm 532 nm / 1064 nm532 nm / 1064 nm IR(1064nm) Laser IR(1064nm) Laser simulates MIPsimulates MIP Quick study possibleQuick study possible Study of charge Study of charge

spread spread

Laser ScannerLaser Scanner 　　532 532 nmnm

1064 1064 nmnm

EEV V-Scan(532nm, normalized)

-0.2

0

0.2

0.4

0.6

0.8

1

1.2

0 22 44 66

Irradiated Position (um)

Sign

al N

orma

lized

by

Clus

ter(r

atio

)

727374

EEV V-Scan(1064,normalized)

-0.2

0

0.2

0.4

0.6

0.8

1

1.2

0 22 44 66

Irradiated Position (um)Si

gnal

Nor

maliz

ed b

y Cl

uste

r(rat

io)

727374

Thin WaferThin Wafer CCD has sensitive thickness ( = epitaxial lCCD has sensitive thickness ( = epitaxial l

ayer thickness) of ~20ayer thickness) of ~20mm

Can be thinned down to 20Can be thinned down to 20m m if mechanically OKif mechanically OK Several ideas: Several ideas:

Thin wafer stretched by tensionThin wafer stretched by tension Thin wafer glued on Be supportThin wafer glued on Be support Partially thinned wafer --- Our studyPartially thinned wafer --- Our study

Partially Thinned WaferPartially Thinned Wafer Picture Frame Picture Frame

TypeType Sample wafer : Sample wafer : Back illumination Back illumination

CCDCCD System for flatness System for flatness

measurement measurement constructedconstructed

Non-flatness has Non-flatness has been measuredbeen measured Poor FlatnessPoor Flatness

Honeycomb & Grid TypeHoneycomb & Grid Type Average thicknessAverage thickness = 76 = 76 mm = 100 = 100 m (including edge)m (including edge) ~0.1% X~0.1% X00

ANSYS analysis:ANSYS analysis:material~1/3material~1/3 rigidity~1/3rigidity~1/3

Simple plate:Simple plate:thickness 1/3thickness 1/3 rigidity 1/27rigidity 1/27

Models for ANSYS

Radiation Hardness of CCDsRadiation Hardness of CCDs

Radiation Damage on CCDsRadiation Damage on CCDs Surface Damage: Charge build-up in SiOSurface Damage: Charge build-up in SiO22

and SiOand SiO22-Si interface by dE/dx-Si interface by dE/dx Increase of surface dark currentIncrease of surface dark current Shift of operation voltage (Flat-band Voltage Shift of operation voltage (Flat-band Voltage

Shift)Shift) Bulk Damage: Displacement in latticeBulk Damage: Displacement in lattice

Increase of bulk dark currentIncrease of bulk dark current Charge Transfer In-efficiency (CTI)Charge Transfer In-efficiency (CTI)

Dark Current and Flat-band Voltage ShiftDark Current and Flat-band Voltage Shift HPK S5466 irradiated with 10mCi Sr-90 HPK S5466 irradiated with 10mCi Sr-90 --

source source

No bias during irradiation Biased during No bias during irradiation Biased during irradiationirradiation

Study of CTIStudy of CTI HPK S5466 and EEV CCD02-06 iHPK S5466 and EEV CCD02-06 i

rradiated with Sr-90 rradiated with Sr-90 -source an-source and Cf-242 n-sourced Cf-242 n-source

Read-out cycle = 3 sec (250 kHz)Read-out cycle = 3 sec (250 kHz) CTI looks decreasing at higher teCTI looks decreasing at higher te

mperature because of increase omperature because of increase of dark current which fill-up the traf dark current which fill-up the traps.ps.

(EEV CCD showed much worse (EEV CCD showed much worse CTI due to less dark current)CTI due to less dark current) NOT expected at JLC where NOT expected at JLC where Tcyc=6ms and much less dark Tcyc=6ms and much less dark currentcurrent Fat-zero charge injection Fat-zero charge injection (~1000 e) is desirable(~1000 e) is desirable

HPK S5466

Other CTI ImprovementsOther CTI Improvements Notch Channel CCDNotch Channel CCD High speed readout : Horizontal CTI is expected High speed readout : Horizontal CTI is expected

prop. to 1/f prop. to 1/f

Conclusion from Radiation Damage StudyConclusion from Radiation Damage Study Surface damage NOT problem in MPP mode Surface damage NOT problem in MPP mode

operation and 6ms cycle timeoperation and 6ms cycle time CTI study + Beam Background SimulationCTI study + Beam Background Simulation CCD can be used for 3 years withCCD can be used for 3 years with - B=2T, R=24mm- B=2T, R=24mm - JLC A-Option- JLC A-Option - Notch channel- Notch channel - Fat-zero charge injection- Fat-zero charge injection - assuming that effect of H.E. electrons is 10 - assuming that effect of H.E. electrons is 10

times stronger than Sr-90 times stronger than Sr-90 -source -source BUT large ambiguity in E-dependence of electron BUT large ambiguity in E-dependence of electron damage and neutron background level.damage and neutron background level.

Model Calculation of NIELModel Calculation of NIEL Bulk damage is thought to be Bulk damage is thought to be

proportional to Non-Ionizing Energy Loss proportional to Non-Ionizing Energy Loss (NIEL) (NIEL)

Sr-90 LC pair background

R&D Items left to be R&D Items left to be donedone

Spatial ResolutionSpatial Resolution Study of resolution of radiation-damaged CCDStudy of resolution of radiation-damaged CCD Study of charge diffusion in epi. layerStudy of charge diffusion in epi. layer

Thin WaferThin Wafer Try to get sample wafers of Honeycomb/Grid tTry to get sample wafers of Honeycomb/Grid t

ypeype

R&D Items left to be done R&D Items left to be done (Cont.)(Cont.)

Radiation Hardness StudyRadiation Hardness Study Study of energy dependence of bulk damageStudy of energy dependence of bulk damage

High energy (150MeV) electron irradiation at Tohoku Univ. High energy (150MeV) electron irradiation at Tohoku Univ. Study of characteristics of irradiated CCDsStudy of characteristics of irradiated CCDs

IIdd vs. Temp vs. Temp Flat-band Voltage ShiftFlat-band Voltage Shift CTI vs. TempCTI vs. Temp CTI vs. Readout frequency CTI vs. Readout frequency cPCI DAQ System cPCI DAQ System CTI vs. Fat-zero charge: Injection of controlled amount of chargeCTI vs. Fat-zero charge: Injection of controlled amount of charge CTI vs. clock pulse width/heightCTI vs. clock pulse width/height Annealing/anti-annealingAnnealing/anti-annealing

R&D Items left to be done R&D Items left to be done (Cont.)(Cont.)

Simulation studies concerning Vertex det.Simulation studies concerning Vertex det. Background study using Full Simulator (JIM, JUBackground study using Full Simulator (JIM, JU

PITER) PITER) Crossing angle: 7 mrad Crossing angle: 7 mrad 20 mrad 20 mrad

Physics study using Quick SimulatorPhysics study using Quick Simulator Physics and Detector study using Full SimulatoPhysics and Detector study using Full Simulato

rr

Future ProspectsFuture Prospects FY2003-FY2004FY2003-FY2004

Continue jobs left to be doneContinue jobs left to be done Find out the best design and operating Find out the best design and operating

condition of CCD vertex detectorcondition of CCD vertex detector Prepare for the next stepPrepare for the next step

Conceptual design of prototype ladder (with Conceptual design of prototype ladder (with HPK)HPK)

Find out the financial source Find out the financial source Japan-US, KAKENHI, or KEK GAISAN-YOUKYU ? Japan-US, KAKENHI, or KEK GAISAN-YOUKYU ?

FY2005- The Next StepFY2005- The Next Step Construction of prototype ladderConstruction of prototype ladder

Future plan in FY2005~Future plan in FY2005~ Custom made CCDs withCustom made CCDs with

Reduced material ( honeycomb type? )Reduced material ( honeycomb type? ) > 20MHz readout speed> 20MHz readout speed Multiple readout nodesMultiple readout nodes Notch structureNotch structure Charge injection capabilityCharge injection capability Readout by ASIC with multi-channel CDSs, AmplifierReadout by ASIC with multi-channel CDSs, Amplifier

s, ADCs, and a Multiplexers, ADCs, and a Multiplexer

Multi-Thread CCDMulti-Thread CCD Normal CCD: Normal CCD: Many V-shifts Many V-shifts Sig. Loss Sig. Loss CPCCD:CPCCD: Limited space for r.o.elec.Limited space for r.o.elec.

Multi-port CCD with few teMulti-port CCD with few tens of V-shifts : MTCCDns of V-shifts : MTCCD

Can be used as a high spCan be used as a high speed CCD cameraeed CCD camera

HPK says “Challenging but HPK says “Challenging but not impossible”not impossible”

ConclusionConclusion Feasibility of the baseline design of a CCD Vertex Feasibility of the baseline design of a CCD Vertex

Detector has been established.Detector has been established. R=24, 36, 48, 60 mmR=24, 36, 48, 60 mm < 4 < 4 mm Thickness = 300 Thickness = 300 m /layerm /layer bb = 7 + 20/(p = 7 + 20/(psinsin3/23/2mm

To get better performance, studies to getTo get better performance, studies to get Rin < 24 mm (Rin < 24 mm ( Radiation hardness) Radiation hardness) Thickness << 300 Thickness << 300 mmwill be continued. A milestone is will be continued. A milestone is bb = 5 + 10/(p = 5 + 10/(psinsin3/23/2mm

Eventually, we have to make a prototype ladder to Eventually, we have to make a prototype ladder to demonstrate the required performance. (demonstrate the required performance. ( need \) need \)

AppendixAppendix Situation in EuropeSituation in Europe

LCFI Group (UK) : R&D for Column Parallel CCDLCFI Group (UK) : R&D for Column Parallel CCD 2.26M£ from PPARC (UK): 2002, 2003, 2004 (3y)2.26M£ from PPARC (UK): 2002, 2003, 2004 (3y) Approved as DESY PRC R&D 01/01Approved as DESY PRC R&D 01/01

MAPS Group (CMOS)MAPS Group (CMOS) DESY PRC R&D 01/04DESY PRC R&D 01/04

DEPFETDEPFET DESY PRC R&D 03/01DESY PRC R&D 03/01

SiLC, CALICE, TPC, -----, submitted proposals tSiLC, CALICE, TPC, -----, submitted proposals to DESY PRCo DESY PRC