Calorimetry at the ILC Detectors

Transcription

1 Calorimetry at the ILC Detectors Sergej Schuwalow, DESY Zeuthen X Int. Conference on Instrumentation for Colliding Beam Physics, Novosibirsk, 28 Feb Mar /03/2008

2 Contents ILC detectors performance goals Particle Flow Approach, PFA calorimetry Detector concepts ECAL options, R&D status HCAL: analog and digital options DREAM: multiple readout calorimetry (see talk of J.Hauptman) Very Forward Calorimeters Summary 2

3 The International Linear Collider ~30km Parameters: 500 GeV (1 TeV upgrade possible) 2 x 1034 cm-2sec-1 electron polarization ~80 % positron polarization ~30 % (60 %) beam sizes: σx 600nm, σy 6nm, σz = 300μm 3

4 Detector performance goals Charge lepton final states : momentum resolution ~10 times better than LEP and ~3 times better than CMS 1/ PT = Hadronic final states: jet energy resolution ~2 times better than SLC and LEP σe/e ~ 30%/ E Flavor tagging, vertex charge: vertex & impact parameter resolution ~3 times better than SLD r z 5 10 / p sin 3/ 2 m SUSY: hermeticity & high veto-capability for two-photon events 4

5 Particle Flow Algorithm (PFA) Jet energy resolution: each particle in a jet is measured separately: charged particles by the tracker, photons by the ECAL, neutral hadrons by the HCAL Cluster separation is the main concern of the PFA calorimetry Small Moliere radius, fine segmentation Large distance from IP, large magnetic field Minimal material budget before the ECAL 5

6 Baseline Detector(s) Design GLD, LDC and SiD concepts are based on PFA approach 4th concept innovative multiple readout calorimetry (DREAM collaboration) Precision silicon micro-vertex detector TPC+several Si-layers (silicon tracker for SiD) as a central tracker LumiCal, BeamCal, GamCal forward calorimetry Large solenoid (calorimeters are inside the coil) Muon identification system 6

7 Detector concepts SiD GLD PFA LDC ILD 4th TPC 7

8 CAlorimeter for the LInear Collider Experiment 13 countries, 45 laboratories (+5 in discussion) 225 physicists/engineers, High granularity calorimeters for precision physics Study of particle flow for σe/e ~ 30%/ E Validation of hadronic interaction models in MC 8

9 Electromagnetic Calorimeters Compact and fine-grained sandwich ECAL Tungsten or lead are used as absorber Sensor planes: Si pad diodes, monolithic active pixel sensors (MAPS), scintillator strip/tiles, combination of silicon/scint. Typical thickness 24 X0 Pad size ~1/3 Moliere radius (5x5 mm2) analog R/O, 50x50 m2 (MAPS, digital R/O) Dynamic range 15 bit (analog R/O) 9

10 Silicon Tungsten Sandwich ECAL CALICE design Frame Detector slab 5x5 mm2 pads FE ASICs P~100 W/channel Beamtests at DESY and CERN The CALICE EUDET module ~40k channels 2 e- showers, 20 GeV 10

11 ECAL Prototype m 1c cm 1st ECAL Module (module -1) Module HCAL 54 c m 5/8 of CMS ECAL ECAL 1er proto. EUDET number of channels Size (cm) 36 x x 54 Tungsten (kg) chip VFE external internal ECAL Final Detector 11

12 Si-W Calorimeter Concept Baseline configuration: transverse seg.: 13 mm2 pixels longitudinal seg: (20 x 5/7 X0) + (10 x 10/7 X0) 17%/sqrt(E) 1 mm readout gaps Currently optimized for the SiD 13 mm effective Moliere concept radius 12

13 Silicon Tungsten Sandwich ECAL-2 US-design ECAL layer structure W (grey), sensors (green) 1024 hex diodes 12 mm channel KpiX, 15-bit ADC, BX tag to be used also for Si-tracker & HCAL 6-inch wafers Full depth ECAL prototype module: 30 layers of hexagonal planes + KpiX, 2.5 mm thick W absorbers. 13

14 The Scintillator-Tungsten ECAL Scintillator strip (4.5 x 1 x 0.3 cm) WLS fiber A PFA calorimeter designed for the GLD detector. Sandwich structure with Scintillator(2mm)Tungsten(3mm) layers. Adopt well-understood plastic scintillator technique. Scintillator strip structure (1 x 4.5 cm) Aiming to reduce number of readout channels while keeping granularity of 1 x 1 cm. Utilize extruded scintillator technique to reduce production cost. Strip clustering is a key issue. Full MPPC (Multi-Pixel Photon Counter) readout. Number of readout ~ 10 M channels 1600 pixel MPPC

15 The Scintillator-Tungsten ECAL WLSF readout Direct readout WLSF R/O from extruded strips GLD-CALICE design DESY testbeam 2007 Grooves 20x20 cm2 FNAL Beamtest August 2008 MPPC R/O 15

16 Monolithic Active Pixel Digital ECAL RAL-CALICE design MAPS with pixel size of 40x40 m2 Binary readout Total number of channels 8x

17 Digital ECAL: MAPS Larger ASIC2 to be designed and submitted in Mid 2008 Concern: Power consumption: 40 W/mm2 DAQ needs: 400 Gbit/s (noise, beam background) 17

18 Digital ECAL: Shower Imaging 18

19 Hadronic Calorimeters Compact and fine-grained sandwich HCAL Stainless steel or lead are used as absorber Sensor planes: scintillator tiles/strips (SiPM or MPPC R/O); GEM, MEGAS, RPC Typical thickness ~5 Pad size: 3x3 cm2-20x20 cm2 analog R/O, ~1 cm2 digital/semi-digital R/O Dynamic range 12 bit (analog R/O) 19

20 Analog HCAL CALICE 1 m3 prototype 2006/2007 Testbeam at CERN Scintillator tiles/wlsf/sipm Stainless steel absorber O(100 M) events collected For details see talk by V.Rusinov 20

21 AHCAL : technical prototype Goal: A compact and realistic (i.e. scaleable) scintillator HCAL structure with embedded electronics Integration issues Readout architecture Ultra-low power ASICs Calibration system Tile and SiPM integration Absorber mechanics with minimal cracks Feed-back from test beam essential Calibration concept Overall detector optimization 21

22 Pb-Sci AHCAL Hardware compensation Pb:Sc = 9.1:2 Strip/tile sizes are to be optimized Strips Better position resolution for same channel count Potential degradation of pattern recognition due to ghost hits 22

23 Digital HCAL Why the digital solution? Going from analog readout to 1:2-bit readout electronics: One can increase detector granularity and hence PFA performance while reducing cost. Cheap, robust detectors suitable for the digital version exist and are very attractive: GRPC, µmegas, GEM Does the digital option mean energy measurement degradation? 23

24 Digital HCAL: Energy Resolution 1-bit digital solution is better at low energy Analog solution is favored at high energy due to high number of particles in the central region. But what about the 2-bit readout solution? The study of KEK group for the GLD HCAL using : 2-bit, 3 thresholds (.5, 10,100 MIPs) associated to 1X1 cm2 tile size shows : Similar energy resolution with respect to the analog readout version for single particle Better energy resolution for jets 24

25 Digital HCAL: GEM Pads could be very small 80% Ar / 20% CO works well 2 30 x 30 cm2 GEM chamber 100 x 30 cm2 chamber for a full size test beam module ->

26 Digital HCAL: MEGAS Fine segmentation High rate counting Total thickness < 6 mm Charge : few-500 fc or pads 50 x50 cm2 MEGAS chamber ~1 m2 plane in

27 Digital HCAL: RPC GRPC : Total thickness (including elect. <6 mm) Fine segmentation (1X1 cm2 pads) Signal Avalanche mode allows high rate Charge: pc Efficiency >90% for gas mixture (TFE-Isobutane-SF6:93-5-2) Pick-up pads Graphite HV Gas Resistive plates Detector dimensions : 8X8, 8X32, 100X100 1cm2-pad, 3.2 mm thick: already produced and tested FNAL-2007 test beam see J.Repond, ALCPG07 Still to be done : Large size detectors up to 100x300 1cm2 pad: New gas mixtures to be found (Isobutane -> CO2, SF6->freonless)

28 Analysis of hadronic showers About 60% of the hadrons will interact in the ECAL The pattern of hadronic showers has to be efficient also in the ECAL Not only in the HCAL. Therefore the choice of technology for the ECAL has also to be based on PFA performances on jet (not only photons) CERN test beam 2007, pions interactions CERN test beam 2007, pions interactions The design of the ECAL must also allows a good pattern of the hadronic shower 28

29 Multiple Readout Calorimetry (4th) Separate measurements of the hadronic shower components Fiber section EM -clear fibers, Č light Charged scintillation fibers Neutrons e.g. Li- or B- loaded glass Fine-grained spatial sampling Excellent energy resolution Gaussian response Crystal EM section (PbWO4) Very good linearity (calibration!) Plans (LCRD): +dual-readout EM section (PbWO4), W absorber, MPPC as a photoconverter, ILC-like module DREAM homepage: 29

30 Design of the Forward Region GamCal ~185m BeamCal LumiCal ILC RDR 30

31 Tasks of the Forward Region ECal and Very Forward Tracker acceptance region. Precise measurement of the integrated luminosity (ΔL/L ~ 10-4) Provide 2-photon veto ad r 0m 5 ~1 ad r m ~40 5mrad IP Provide 2-photon veto Serve the beamdiagnostics using beamstrahlung pairs Serve the beamdiagnostics using beamstrahlung photons Challenges: High precision, high occupancy, high radiation dose, fast read-out! 31

32 Precise Measurement of the Luminosity Required precision is: ΔL/L ~ 10-4 (GigaZ 109/year) ΔL/L < 10-3 (e+e- W+W- 106/year) ΔL/L < 10-3 (e+e- q+q- 106/year) BeamCal LumiCal Bhabha scattering ee->ee(γ) is the gauge process: Count Bhabha events in a well known acceptance region => L = N/σ High statistics at low angles => NBhabha ~ 1/θ3 Well known electromagnetic process: the current limit on the theoretical cross section error is at ~

33 Physics Background and Beam-Beam Effect 2-photon rejection 2-photon events are the main background. BHSE We determined an efficient set of cuts to reduce the background to the level of The Bhabha Suppression Effect (BHSE) is due to the EM deflection and energy loss by beamstrahlung of the Bhabhas. Correction needs precise knowledge of beam parameters. C.Rimbault et al. JINST 2:O

34 LumiCal Design Si/W sandwich calorimeter, 2 half barrels, layers laser position monitoring system Single detector layer 48 azimuthal sectors, each sector subdivided into radial pads of about 1 mrad Each layer consists of 3.5mm tungsten absorber, 300μm silicon sensor and readout. 34

35 BeamCal Design Compact em calorimeter with sandwich structure: 30 layers of 1 X0 3.5mm W and 0.3mm sensor Angular coverage from ~5mrad to ~45 mrad Moliére radius RM 1cm BeamCal LumiCal Segmentation between 0.5 and 0.8 x RM W absorber layers Radiation hard sensors with thin readout planes BeamCal Space for readout electronics 35

36 The Challenges for BeamCal e.g. Breit-Wheeler process Creation of beamstrahlung at the ILC ee- e+ γ e- e+e- pairs from beamstrahlung are deflected into the BeamCal e- e+ γ 1 MGy/a e+e- per BX => TeV total energy dep. ~ 10 MGy per year strongly dependent on the beam and magnetic field configuration => radiation hard sensors Detect the signature of single high energetic particles on top of the background. => 5 MGy/a high dynamic range/linearity 36

37 BeamCal Mechanics Total outer radius: 220 mm Graphite shield of 10 cm Total weight: about 200 kg Upper part of the shielding tube must be removable Crane operation necessary for assembly/disassembly Additional space in front of and behind BeamCal is needed for electronics, cooling etc.. 37

38 Sensor Materials under Investigation pcvd diamonds: (courtesy of IAF) polycrystalline CVD diamond radiation hardness under investigation (e.g. LHC pixel detectors) advantageous properties like: high mobility, low εr = 5.7, thermal conductivity availability on wafer scale GaAs: semi-insulating GaAs, doped with Sn and compensated by Cr produced by the Siberian Institute of Technology available on (small) wafer scale GaAs SC CVD diamonds: available in sizes of mm2 CVD: Chemical Vapor Deposition Single crystal CVD diamond 38

39 GamCal Design 39

40 Summary The requirements on the ILC calorimeters are physics driven. Example: jet energy resolution σe/e ~ 30%/ E is needed Potentially many technologies may match these requirements. The majority of present designs are based on PFA approach. Numerous test beam studies/full-system tests are needed. CALICE tested ECAL & HCAL prototypes at the beam. DREAM made successful prove-of-principle beam test. FCAL: R&D on calorimeters for the very forward region. LumiCal for precise luminosity measurement + hermeticity. BeamCal for 2- events veto and fast beamdiagnostics. Intensive R&D activity on radiation hard sensors is ongoing. GamCal for fast luminosity optimization is still under design. 40

41 Backup 41

42 Material budget before ECAL Precision physics at ILC is incompatible with this CMS ATLAS 20-40% of the photons are converted before the ECAL 40-80% of the electrons start showering before ECAL 5-20% of the pions start had. shower before calo. Totally inconsistent with PFA approach for jets 42

43 Analysis of e.m. showers TB & analysis MIP signal 46 ADC counts/mip ECAL W-Si Electron data Linearity 1% (DESY/CERN) Noise 6 ADC counts From pedestal width Longitudinal eshower profiles Resolution E/E=17.13/ (E/GeV) 0.54% 43 43

44 Future plan : The FNAL beam test in Aug 2008 Establish the Scintillator-strip ECAL Test linearity of the full-mppc readout calorimeter with high energy beam. Evaluate all the necessary performances using various beams (π,k,e,µ.) with wider energy range Make the SCECAL ready for the engineering design Combined test with the Analog HCAL Test π0 -> 2γ reconstruction The 2nd prototype will be 4 than the DESY BT Measure hadron shower totimes test larger simulation model module. Compare the result with various models (20 x 20 cm, ~30 layers) Precise hadron simulation will help study of PFA Fully adopt the extruded scintillators. Expect > 2000 readout channels. MPPC R/O

45 Summary Study of the scintillator-ecal is steadily ongoing. R&D of the photon sensor is underway collaborating with Hamamatsu. Current 1600-pixel MPPC sample already shows almost satisfactory performance. We have proven that scintillator-strip calorimeter with full MPPC readout works. Study of the extruded scintillator production is ongoing in Korea. From results of the KEK beam test, we understand how to improve the performance of extruded scintillators. The Scintillator-ECAL technology will be established and all tested at the next FNAL beam test in this year. There are still some concerns, however almost of them can be solved. Dynamic range MPPC improvement ongoing. Granularity up to 1 x 1 cm possible Strip clustering work ongoing. Cost extrusion method will reduce scintillator cost. MPPC cost is another key issue. Detector calibration will use MIPs in jets, study ongoing.

46 courtesy from H.Matsunaga 46

47 courtesy from H.Matsunaga 47

48 Electronics DHCAL for ILC will have >25 million channels. So the electronics should be : Tiny, embedded electronics Low consumption (<10 µw/ch) Semi-digital (2-3 thresholds) Fast (< ns) Capable to store the events during data train Low cost

49 Electronics HARDROC 64 channels, 16mm² Digital/analog output. 2 thresholds low consumption, power pulsing (< 10 µw/ch) Digital memory able de store up to 128 evts. Large gain range Adequate for GRPC* (threshold > 10 pc) Another chip is currently under development to reduce the threshold down to 2 fc for µmegas

50 Perspectives 2008: 90X90 cm2 GRPC/µMEGAS pcb-connector DHCAL detectors fully equipped: Detectors: GRPC produced µmegas not yet chips: produced PCB : designed PCB-connectors under study DIF: under study PCB-detector assembling under study 1- To be completed by September Completely funded 50

51 2009 : Perspectives Technological prototype = 40 detector+absorber planes with full electronics readout : 40X (6+20 mm) Funded essentially by ANR-France ( ) In order to make this prototype as close as possible to the ILC module we need to determine : Detectors dimension Global mechanical structure 51

52 DHCAL architecture (H.Videau) Gas, H.V acess In addition : No cracks: Each particle crossing the same number of detectors no problem concerning particles produced with the θ = π/2 Easy access to each element of the DHCAL from the outside 52