1. TESLA Detector Markus Schumacher, University of Bonn American Linear Collider Workshop, Cornell, July 2003 « Requirements « Basic Concepts « Developments

2. Requirements from Physics  momentum: (1/p) = 7 x 10 -5 /GeV (1/10xLEP) e + e -  ZH  ll  X goal: M  <0.1x      dominated by beamstrahlung  impact parameter : d=510/p(GeV)m (1/3xSLD) excellent flavour tagging capabilities for charm and bottom quarks e.g. measurement of Higgs branching ratios  jet energy : E/E = 0.3/E(GeV) (<1/2xLEP) M Dijet ~  Z/W  e.g. separation between e + e - WW qqqq and e + e - ZZ qqqq LC LEP  reconstruction of multijet final states: e.g. e + e -  H + H -  tbtb  bqqb bqqb  hermetic down to 5 mrad missing energy topologies (e.g. SUSY and Higgs) Physics determines detector design

3. Requirements due to the accelerator design « Time Structure: « Event rates: Luminosity: 3.4x10 34 cm -2 s -1 (6000xLEP) e + e -  qq,WW,tt,HX 0.1 / train e + e -    X:~200 /Train Background from Beamstrahlung: 6x10 10 /BX 140000 e + e - /BX + secondary particles (n,) 5 Bunch Trains/s t bunch =337ns But still: 600 hits/BX in Vtx detector 6 tracks/BX in TPC E=12GeV/BX in calorimeters E 20TeV/BX in forward cals.  Large B field and shielding High granularity of detectors and fast readout for stable pattern recognition and event reconstruction

4. Basic TESLA Detector Concept No hardware trigger, dead time free continous readout for complete bunch train (1ms) Zero suppression, hit recognition and digitisation in FE electronics Large gaseous central tracking device (TPC) High granularity calorimeters High precision microvertex detector All inside magnetic field of 4 Tesla

5. Overview of tracking system Central region: Pixel vertex detector (VTX) Silicon strip detector (SIT) Time projection chamber (TPC) Forward region: Silicon disks (FTD) Forward tracking chambers (FCH) (e.g. straw tubes, silicon strips) Requirements: Efficient track reconstruction /good resolution down to small angles independent, robust track finding in TPC (200) and in VTX+SIT (7 points)  allows calibration, alignment excellent momentum resolution (1/p) < 7 x 10 -5 /GeV

6. Vertex Detector: Conceptual Design 5 Layer Silicon pixel detector Small R1: 15 mm (1/2 SLD) Pixel Size:20x20m 2   Point =3 m Layer Thickness: <0.1%X 0 suppression of  conversions – ID of decay electrons minimize multiple scattering 800 million readout cells Hit density: 0.03 /mm 2 /BX at R=15mm  pixel sensors Read out at both ladder ends in layer 1: frequency 50 MHz, 2500 pixel rows  complete readout in: 50s ~ 150BX <1% occupancy no problem for track reconstruction expected Impact parameter d ~R1  point

7. Vertex Detector: Technology Options Established Technology: CCDs Excellent experience at SLD (300 million channels) R&D: efficiency and stability of charge transfer readout speed, thinning of sensors, mechanics, radiation hardness „New“ Technologies: MAPS (Monolithtic Active Pixel Sensors), FAPS DEPFET (Depleted Field Effect Transistor) HAPS (Hybrid Active Pixel Sensors), SiO Each pixel can be adressed individually Only single row active per ladder  smaller power consumption First amplification in pixel  smaller noise R&D: above + building of large devices steering readout See Chris‘ talk for more details

8. Flavour Tagging LEP-c « Powerful flavour tagging techniques (from SLD and LEP)  M e.g. vertex mass ­ charm-ID: improvement by factor 3 w.r.t SLD Expected resolution in r,and r,z  ~ 4.2  4.0/p T (GeV) m

9. Flavour Tagging : Recent Studies  Inner layer at 1.5cm is very important, e.g. e + e -  Z *  ZH ZH  llbb,ZH  llcc,ZH  llgg « W/O inner layer: charm tagging degraded by 10% Double layer thickness small effect Quark  Antiquark discrimination via Vertex/Dipole Charge: bottom: p= 80% = 80% charm: p= 90% 35% « However: minimal amount of material important limited number of conversions, electron-id, reconstruction of vertex mass including  0, …

10. Gaseous or Silicon Central Tracking Detector? gaseous silicon

11. Motivation for a TPC « Large no. of 3D space points robust and efficient track reconstruction in high track density environment new heavy stable particles GMSB SUSY:    + G ~ ~  Minimal material little influence on calorimetry little multiple scattering small number of conversions « dE/dx  particle identification « Tracking up to large radii Reconstruction of V 0, Kink Tracks aid energy/particle flow + sensitivity to new physics

12. TPC Conceptual Design Radial space points: 200 Point res.: < 140 m (goal:100 m) Pad size: 6 (r) x 2 (phi) mm 2 Large lever arm: R I/A = 40/160 cm Little material: < 3% X 0 Gas choice: Ar:CO 2 :CH 4 = 93:2:5 % (CF 4 also investigated) Compromise between drift velocity ~ 5cm/s and neutron cross section Total Drift time 50 s = 160 BX  80000 hits in TPC (physics+BG) 8x10 8 readout cells (1.2MPads+20MHz)  0.1% occupancy No problem for pattern recognition/track reconstruction TPC: (1/p) = 2.0 x 10 -4 GeV -1 +VTX: (1/p) = 0.7 x 10 -4 GeV -1

13. Gas Electron Multipliers or MicroMEGAS « better instrinsic point resolution 2 dimensional readout symmetry electron signal read out Small hole separation  reduced ExB effects « natural supression of ion feedback  no wire tension  thin endplates Gas Amplification & Point Resolution - chevron pads - large number of small silicon pads - resistive or capacitive coupling of neighbouring pads - larger gap between GEMs and pad plane Small width of electron cloud (single pad)  improve point resolution by charge sharing (details see Ron’s and Dean’s talks)

14. Intermediate and Forward Tracking SIT: 2 Layers of Si-Strips  r = 10m FTD: 7 Disks 3 layers of Si-pixels 50x300m 2 4 layers of Si-strips  r = 90m FCH: 4 Layers Strawtubes or Silicon strips (double sided)  Forward tracking (e.g. e+e-  WW  qql recover mom. resolution at small angles 250 GeV  «Increase track matching from TPC to VTX by 4 %  Improve Momentum resolution: TPC+VTX: (1/p) = 0.7 x 10 -4 GeV -1 «V 0 -Reco. Eff. 73  86% (for r=6to11cm) track reconstruction efficieny: =98.4 (incl. Background hits) +SIT : (1/p) = 0.5 x 10 -4 GeV -1

15. Calorimetry ZHH  qqbbbb  Kinematic fits often not applicable – Beamstr., ISR,, LSP  Intrinsic jet energy resolution is of vital importance  Design optimised for Particle/Energy Flow Algorithm Excellent jet energy resolution much of LC physics depends on reconstruction of invariant masses from jets in hadronic final states Good energy and angular resolution for photons Reconstruction of non-pointing photons Hermeticity Requirements:

16. Particle / Energy Flow 60 % charged particles:30 %  :10 %K L,n The energy in a jet is: Reconstruct 4-vectors of individual particles avoiding double counting Charged particles in tracking chambers Photons in the ECAL Neutral hadrons in the HCAL (and possibly ECAL)  need to separate energy deposits from different particles small X 0 and R Moliere : compact showers high lateral granularity D ~ O(R Moliere ) large inner radius L and strong magnetic field granularity more important than energy resolution  K L,n  e  Discrimination between EM and hadronic showers small X 0 / had longitudinal segmentation

17. Calorimeter Conceptual Design « ECAL and HCAL inside coil « large inner radius L= 170 cm  good effective granularity ECAL: silicon-tungsten (SiW) calorimeter (preferred choice) Tungsten : X 0 / had = 1/25, R Moliere ~ 9mm (gaps between Tungsten increase effective R Moliere ) Lateral segmentation: 1cm 2 matched to R Moliere Longitudinal segmentation: 40 layers (24 X 0, 0.9 had ) Resolution:  E /E = 0.11/E(GeV)  0.01 x~BL 2 /(R M D) 1/p x distance between charged and neutral particle at ECAL entrance 2nd option: 45 layers of Pb(W)+scintillating plates+WLS + 3 layers of Si sensors (.9x.9 cmxcm)

18. Two Options: Tile HCAL (Analogue readout) Steel/Scintillator sandwich WLS + Photodetectors (WLS: different geometries) (APDs, SiPM on tiles,…) Lower lateral segmentation 5x5 to 25x25 cm 2 Longitudinal segmentation: 9-12 samples 4.5 – 6.2 had (limited by coil radius) Hadron Calorimeter HCAL ECAL Digital HCAL (digital readout) via RPCs,GEMS, small scint. tiles High lateral segmentation 1x1 cm 2 resolution: E/E =0.35/E(GeV)  0.03  seperation:   fake  =10 -3

19. Calorimeter Reconstruction `Tracking calorimeter’ – very different from previous detectors Requires new approach to reconstruction Already a lot of excellent work on powerful particle/energy flow algorithms Still room for new ideas/ approaches A lot of R&D activities: Continue evaluation of digital vs analog HCAL Calorimeter segmentation, HCAL active medium Simulation of hadronic showers  test beams jet energy:E/E = 0.3/E (GeV)  = 68mrad/E(GeV)  8mrad without vertex constraint for photons OPAL

20. Forward Calorimeters LCAL: Beam diagnostics and fast luminosity (28 to 5 mrad) ~10 4 e + e — pairs/BX 20 TeV/BX 2MGy/yr Need radiation hard technology: SiW, Diamond/W Calorimeter or Scintillator Crystals LAT : Luminosity measurement from Bhabhas (83 to 27 mrad) SiW Sampling Calorimeter aim for  L / L ~ 10 -4 require  = 1.4 rad TDR version of maskL* = 3 m Tasks: Shielding against background Hermeticity / veto

21. Recent Developments Shower leakage Difficulty in control of inner acceptance to ~1m  TDR version of LAT difficult for a precision lumi measurement ? New L* = 4-5 m version currently being studied.  Flat: better for Lumi. measurement  More Space for electronics etc.  inner radius LAT: 8cm  5cm  Hermetic to 3.9 mrad (was 5.5 with gaps)  less indirect background hits ?? Design in flux + very active R&D

22. Detector Optimization  Current detector concept essentially unchanged from TDR + OTHER/NEW IDEAS…… « Time to think again about optimizing detector design, consider the detector as a whole entity  Optimize design w.r.t. overall detector performance using key physics processes, e.g.  Need unbiased comparison Same/very similar reconstruction algorithms Common reconstruction framework Same Monte Carlo events  looking at TPC length, extra Si tracker between TPC and ECAL,… « something forgotten ?  devil’s advocate committee

23. Conclusions Precision physics determines the detector design Basic design almost unchanged compared to TDR Proof of principle for the best suited technologies to be provided by ongoing R&D Optimise Overall Detector Performance  in worldwide collaboration to find best detector concept for a future linear collider ! The Physics potential at a LC is excellent, the requirements to the detector are challenging High lumi  large statistics  small systematics  need best detector which can be build