ILC Detector R&Ds and Design

ILC Detector R&Ds and Design

Toward detectors and collaborations that realize and maximize the physics output of ILC

Hitoshi YamamotoTohoku University

ICFA seminar, Daegu, Sept. 29, 2005

ILC Parameters

■ 1st stage Energy 200→500 GeV 500 fb-1in first 4 years + 500 fb-1in next 2 years

■ 2nd stage Energy upgrade to ~1TeV 1000 fb-1in 3-4 years

■ Energy scan + e polarization■ Options

eeeGiga-Z, e+ polarization

(http://www.fnal.gov/directorate/icfa/LC_parameters.pdf)

ILC Physics

e.g. Higgs coupling measurements

SM Higgs : coupling mass

Higgs Couplings : Deviations from SM(By S. Yamashita)

SUSY (2 Higgs Doulet Model)

Extra dimension(Higgs-radion mixing)

ILC Detector Performance Goals

■ Vertexing ~1/5 rbeampipe,~1/30 pixel size (wrt LHC)

■ Tracking ~1/6 material, ~1/10 resolution (wrt LHC)

■ Jet energy (quark reconstruction) ~1/2 resolution (wrt LHC)

€

σ ip = 5μm ⊕10μm / psin3 / 2 θ

€

σ(1/ p) = 5 ×10−5 /GeV

€

σE / E = 0.3/ E(GeV)

(http://blueox.uoregon.edu/~lc/randd.pdf)

€

(h → bb ,cc ,τ +τ −)

€

(e+e− → Zh → l +l −X; incl. h → nothing)

b, c tagging by vertexing

Pixel vertex detector

4-layer 0.3 % X0/ layer rbp = 2 cm conservative design 5-layer 0.1 % X0/ layer rbp = 1 cm agressive design (~goal resolution)

e+e → ZH Recoil mass resolution

■ Good momentum resolution of ~5x10-5 is required (not a luxuary). Not limited by the beam energy spread.

Only Z→l+l- detected : Higgs decay independent

Jet(quark) reconstruction

■ With , Z/Wjj can be reconstructed and separated

€

σE / E = 0.6 / E(GeV)

€

σE / E = 0.3/ E(GeV)

€

e+e− → νν WW ,νν ZZ

€

W /Z → jj

€

σE / E = 0.3/ E

(Strong EWSB)

PFA (Particle Flow Algorithm)

■ Many other important modes have 4 or more jets : e.g.

Higgs self-coupling : 6 jets

Top Yukawa coupling : 8 jets

WW* branching fraction of Higgs : 4 jets+missing

■ How to achieve for jet ?■ Basic idea : PFA

Use trackers for charged particles Use ECAL for photon The rest is assumed to be neutral hadrons (ECAL+HCAL)

€

e+e− → Zhh → (qq )(qq )(qq )

€

σE / E = 0.3/ E€

e+e− → tt h → (bqq )(b qq )

€

e+e− → Zh → (qq )(qq )(l ν )

Red : pionYellow : gammaBlue : neutron

e+

e-

Z→qq (by T. Yoshioka)

- Gamma Finding

Red : pionYellow : gammaBlue : neutron

gamma

- Track Matching

Red : pionYellow : gammaBlue : neutron

Remaining hits are assumedto be neutral hadrons.

Red : pionYellow : gammaBlue : neutron

PFA : major soruce = confusion

■ Using typical values

■ ... and ignoring confusion,

■ Confusion is dominant even for the goal of

■ → fine segmentation , large radius : cost!

€

σ jet2 = σ ch

2 + σ γ2 + σ nh

2 + σ confusion2

€

σ ch << σ γ ,nh , σ γ / Eγ =11% / Eγ , σ nh / Enh = 34% / Enh

€

σ jet / E jet =12% / E jet

€

σE / E = 30% / E

Beampipe radius

■ Stay-clear for the soft e+e- pair background

R ~ 1/B1/2

■ Larger ECAL radius → larger solenoid radiu

s → lower B (cost!) → larger beampipe R → worse vertexing

■ Where is the optimum?IP

Major Detector Concept Studies(the parameters are the current defaults - may change)

■ SiD (American origin) Silicon tracker, 5T field SiW ECAL 4 ‘coordinators’ (2 Americans, 1 Asian, 1 European)

■ LDC (European origin) TPC, 4T field SiW ECAL (“medium” radius) 6 ‘contact persons’: (2 Americans, 2 Asians, 2 Europeans)

■ GLD (Asian origin) TPC (+Silicon IT), 3T field W/Scintillator ECAL (“large” radius) 6 ‘contact persons’: (2 Americans, 2 Asians, 2 Europeans)

+ vertexing near IP

ECAL/HCAL inside coil

Detector Concepts

■ 4th concept proposed at Snowmass 05 Based on dual-readout compensating cal.

■ Requests from WWS for new concept (as of 2005,9)

Contact person(s) Provide representatives for panels (R&D panel, MDI panel, Costing panel) Produce “detector outline document” by end Feb. 2006

WWS (Worldwide Study)

■ Started in 1998 (Vancouver ICHEP)■ 6 committee members from each of 3 regions■ 3 co-chairs - now members of GDE

C. Baltay → J. Brau D. Miller → F. Richard S. Komamiya → H. Yamamoto

■ Tasks (in short) Recognize and coordinate detector concept studies Register and coordinate detector R&Ds Interface with GDE Organize LCWS (1 per year now)

Detector Outline Document

■ Document that precedes CDR■ Contents (~100 pages total)

Introduction Description of the concept Expected performances for benchmark modes Subsystem technology selections Status of on-going studies List of R&Ds needed Costing Conclusion

Detector Timeline

(2005 end) Acc. Configuration Document

Detector R&D report

(2006,2 end) “Detector outline documents” (one for each detector concept)

(2006 end) Acc. Reference Design Report

Detector CDR (one document)

(~2008) LC site selection Collaborations form

~Site selection + 1yr Global lab selects experiments.

Accelerator Detector

#BDS (beam delivery system) and crossing angles

20mrad xing simpler and better understood now Two BDSs →More constraints on linac One BDS with 10-12mrad xing? Machine simulation : more background for 2mrad Detector simulation : more background for 20mrad Baseline configuration to be determined

#IR, #detectors (at ILC startup)?■ Roughly in rising/falling order of preference for acc./det. p

eople, (iIR: instrumented IR, nIR: non-instrumented IR)

2 iIRs/ 2 detectors     1 iIR/ 2 detectors (push-pull) + 1 nIR 1 iIR/ 2 detectors (push-pull) 1 iIR/ 1 detector (push-pull capability) 1 iIR/ 1 detector + 1 nIR 1 iIR/ 1 detector

■ #det panel of WWS (chair: J. Brau) Produced a report (http://blueox.uoregon.edu/~lc/wwstud

y)

WWS Panels

WWS

parameter

R&D

MDI

benchmark

costing

software

........

done

~done

R&D Panel■ Charge:

Survey and prioritize R&Ds needed for ILC experiments (NOT individual proposals)

Inputs are from R&D collaborations and concept studies

Register and facilitate regional review processes■ Chair: C. Damerell ■ Outputs:

Web links to R&Ds https://wiki.lepp.cornell.edu/wws/bin/view/Projects/WebHo

me Detector R&D report (end 2005)

Horizontal and Vertical collaborationsIt is something like this : (detail may not be accurate)

Vertexing 1 train = ~3000 bunches in 1ms, 5 Hz Typical pixel size ~ (20m)2 → occupancy is too high if integrate

over 1 train. No solution to bunch id each hit so far. Then what?

■ Readout during train ( ~20 times) Standard pixel size - MAPS, CPCCD, DEPFET, SOI

■ Readout between train Standard pixel size ( ~20 time slices stored on-pixel)

◆ Store in CCD - ISIS◆ Store in capacitors - FAPS

Fine pixel size (~1/20 standard)◆ No Bunch id - FPCCD ◆ Bunch id - CMOS (double pixel sensor)

No demonstrated solution yet. (apology for not covering all...)

CPCCD (column-parallel CCD)

■ RAL■ Readout each column separately■ 50MHz would readout 5cm 20

times per train■ Diffusion : multi hit while shifting

→ fully depleted CCD?■ Prototype sensor (CPC1) tested w/

>25 MHz readout.■ Clock drive is challenging.■ Readout chip made (CPR1)

Operation verified (w/bugs to fix)■ New sensor/readout fabricated

(CPC2/CPR2) and under tests.

MAPS (Monolithic Active Pixel Sensor)

■ IReS,GSI,CEA (+SUCIMA coll.)■ Use the epi-layer of commercia

l processes - small signal (a few 10s e)

■ 1Mrad OK (SUCCESOR1)■ 1012n/cm2 OK, 1013e/cm2 OK (MIMOSA9)■ 3 sensors thinned to 50m

■ CP,CDS works(MIMOSA8), but not fast - readout transversely.

■ Also try FAPS-like scheme (MIMOSA12)

5mm 2mm

Inner layer

sensor ADC/clusterng

ADC count 55Fe

Before&after 1Mrad

ISIS (In-situ Storage Image Sensor)

Small CCD on each pixel (~20 cells) - charge is

shifted into it 20 times per trainImmune to EMITechnology exists as ultra-high-speed cameraPrototype now being made (E2V)

To column load

Source followerReset transistor Row select transistor

p+ shielding implant

n+buried channel (n)

storage

pixel #1

storage

pixel #20 sense node (n+)

Charge collection

row select

reset gate

VDD

p+ well

reflected charge

reflected charge

photogate

transfer

gate

output

gate

High resistivity epitaxial layer (p)

FAPS (Flexible Active Pixel Sensor)

Pixels 20x20 m2

10 storage cells per pixel

(20 in the real sensor)First prototypes in 2004Source test done

FPCCD (KEK)

■ Fine-pixel CCD (5m)2 pixel Fully-depleted to suppress

diffusion Immune to EMI CCD is an established technology Baseline for GLD

Fully-depleted CCD exists (Hamamatsu : astrophys.)

Background hits can be furhter reduced by hit pattern (~1/20)

No known problems now Want to produce prototype in 2

006 (Funding!)

CMOS (double pixel sensor)

■ Yale, Oregon■ 2 pixel sensors on top of each ot

her - 5x5m2 (micro) and 50x50m2 (macro)

■ Macro pixel triggers and times (bunch id) hits - up to 4 hits stored on pixel.

■ Micro pixels store analog signal.■ Time and ADC data are read out

between trains. ■ Only micro pixels under hit macr

o pixels are queried.■ Two sensors in one silicon, or bump-bonded.■ Conceptual design being worked

with Sarnoff.50m

Trackers

■ Two main candidates TPC - central tracker for GLD, LDC

◆ ~200 hits/track σm/hit Silicon strip - central tracker for SiD

◆ ~5 hits/track with much better σ◆ Also used as

◆ Inner/forward tracker for GLD, LDC◆ Endcap tracker for GLD◆ Outer tracker (of TPC) for LDC

TPC■ Endplate detectors

Wires - conventional◆ Amplification at wires only◆ Signal is induced on pads - slow collection◆ Strong frame needed - endplate material◆ Wires can break

MPGD (Multi-pixel Gas Detector) - R&D items◆ Amplification where drift electrons hit (w/i ~100m)◆ Directly detect amplified electrons on pads - fast◆ Ion feeback suppressed

◆ GEM (Gas Electron Multiplier)◆ 2-3 stages possible - discharge-safer(?)

◆ MicroMEGAS (Micro Mesh Gas detector)◆ 1 stage only - simpler

MicroMEGAS

■ Micromesh with pitch~50m■ Pillar height ~ 50-100m■ Amplification between mesh an

d pads/strips■ Most ions return to mesh.

S1

S2

σ

~50m

MicroMEGAS

■ Micromesh with pitch~50m■ Pillar height ~ 50-100m■ Amplification between mesh an

d pads/strips■ Most ions return to mesh.

S1

S2

σ

~50m

GEM■ Two copper foils on both sides

of kapton layer of ~50m thick■ Amplification at the holes■ Gain~104 for 500V■ Can be used multi-staged■ Natural broadening can help ce

nter-of-gravity technique.

p~140m

p~60m

ILC TPC R&D groups~70 active people worldwide

DESY

Aachen

Victoria

MPIKEK

Sacley-Orsay

KerlsruheBerkeleyNovosibirskCarletonCornell.....

Interconnected

TPC R&D results

• Now 3 years of MPGD experience gathered. MPGDs compared with wire

• Gas properties rather well understood (dirft velocity, diffusion effect ~ MC)

• Diffusion-limited resolution seems feasible

• Resistive foil charge-spreading demonstrated

• CMOS RO chip demonstrated• Design work starting for the

Large Prototype (funded by EUDET)

GEM vs wire

Charge spreading by resistive foil

Silicon Tracker R&Ds■ DSSD in-house fabrication in Kor

ea Characterized. S/N = 25 Radiation test in progress Hybrid is produced

■ Long-ladder R&D (SantaCruz) Readout chip LSTFE for long and

spaced bunch train. Being tested.

Backend architecture defined Long ladders being assembled

■ SILC collaboration 10-60cm strip length S/N = 20-30 for 28cm (Sr90), O

K New front end chip being tested ~OK. Next : power cycling Ladder assembly prototype soon

Calorimeters

E

%40~

■ Critical part of PFA

■ ‘Realistic’ PFA Full shower simulation Clustering Photon finding Track matching Achieved ~40%/E1/2 for the 3 concepts

■ Starting to be useful for detector optimization

Analog vs digital HCAL readout Segmentation However, not quite mature yet to be

conclusive

■ Large international collaboration : CALICE Jet energy resolution at Z→qq

ECAL■ Silicon/W

High granularity (~1cm2 or less) and stable gain. Cost : $2-3/cm2 for Si. How far can it go down?

CALICE prototype (1cm2 cell) beam test SLAC/Oregon/UCDavis/BNL silicon wafer (4x4mm2)

ECAL■ Scintillator/W

Cheaper and larger granurarity (3x3 - 5x5cm2) Scintillator strips may be cost-effective way for granurarity (1cm x Ycm) Read out by fibre + PMT or SiPM/MPC

Japan/Korea/Russia Colorado : staggered cells (5x5cm2)

■ SiPM (invented in Russia) ~100 cells in 1mm2

Limited Geiger mode High B field (5T) OK Gain ~ 106 ; no preamp Fast σ~ 50ps Quite cheap Noisy? Temperature dependence Steep bias valtage dependenc

e

HAMAMATSU MPC(Multipixel Photon Counter)Sees ~60 pe’s at room temp.

HCAL

■ Analog : Scintillator (CALICE) Modest granurarity (3x3cm2 u

p) SiPM readout MINICAL prototype tested with

100 SiPM - Same resolution as PMT

2 cm steel

0.5 cm active

HCAL■ Digital (CALICE)

Fine granurarity (~1x1cm2) 1 bit readout GEM and RPC w/ pad readout Common readout electronics Understood well - ready for 1m3

prototype

Signal PadMylar sheet

Mylar sheet Aluminum foil

1.1mm Glass sheet

1.1mm Glass sheet

1.2mm gas gap

-HV

GND

GEMRPC

Calorimeter R&Ds

■ Si-Scintillator hybrid for ECAL Cost-performance optimization

■ Crystal for ECAL Focus on energy resolution

■ DREAM Dual readout of dE/dx (scintillat

or) and Cerenkov (quartz fibre) Ideal compensation to obtain ve

ry good hadron energy resolution Basis for the 4-th concept Challenge : ILC implementation

Other subsystems

■ Muon system is probably easy in concept but difficult in practice (large system - support, etc.)

■ Solenoid and compensation coil (DID - for large xing angle) : non-trivial problem to realize, and DID is a problem to solve for trackers and bkg.

■ Forward regions (endcap regions) are important for t-channel productions such as

■ Very forward regions (FCAL, BCAL) are critical for tagging electrons for SUSY pair creations.

■ With the long train, DAQ is not a trivial problem

■ Beam instumentations such as pair background detector play important roles in machine operation/tuning

Just as importnat as what has been shown

€

e+e− → νν h

Concluding Remarks

■ Too many R&Ds too cover : apology for those not covered. Refered to the R&D report to be produced ~ end 2005.

■ Resolutions much better than past is not luxuries, but required for balanced investment in ILC.

■ With EUDET ($7M over 4 years), detecor R&D in Europe is now reasonably funded (only for ‘infrastructures), but severely underfunded in Americas and Asia.