Muon Collider Detector New Instrumentation Possibilities Alan Bross
Muon Collider Detector “New Instrumentation Possibilities Alan Bross BNL Muon Collider Design Workshop December 4, 2007 1
MCD – Progress Since Snowmass 96 · None · But an enormous amount of Detector R&D has been done LHC u ILC u CLIC (really ILC’) u · What is Most relevant to the MCD are developments for ILC detectors u Same physics, especially when we compare to CLIC Alan Bross BNL Muon Collider Design Workshop December 4, 2007 2
Major Issues · Event Rates u Not really an issue, LHC detector/electronics developments can easily handle MC event rates · Backgrounds, radiation-damage effects are what need to addressed Alan Bross BNL Muon Collider Design Workshop December 4, 2007 3
Snowmass 96 Background Calculations Alan Bross BNL Muon Collider Design Workshop December 4, 2007 4
Snowmass 96 Background Calculations II Longitudinal Radial Alan Bross BNL Muon Collider Design Workshop December 4, 2007 5
Snowmass 96 Detector Alan Bross BNL Muon Collider Design Workshop December 4, 2007 6
Snowmass 96 Detector II · Central Magnet u u 2 T » 8 X 15 m · SVD u u 300 mm cell 300 mm thick · TPC Central Tracker · Cal u u LAr Scintillator Tiles · Muon u Alan Bross BNL Muon Collider Design Workshop Cathode Strips/pads December 4, 2007 7
Snowmass 96 Detector II · Central Magnet u OK s s CMS, 4 T, 6 X 12. 5 m MCD, 2 T, 8 X 15 m (probably 4 T is affordable) · SVD u u Occupancy OK – 3% (» 30 hits/cm 2) Radiation damage (n) – Marginally OK (conservative – Not OK) s · TPC u 1 Year lifetime (1014 n/cm 2) Probably would work even with 1996 technology · Calorimetry u Also Probably OK · Muon System u OK Alan Bross BNL Muon Collider Design Workshop December 4, 2007 8
Developments in last 10 Years · Lets Look at ILC/CLIC · CLIC IR: Beamstrahlung (d. B) » 20% » 6 o Alan Bross BNL Muon Collider Design Workshop December 4, 2007 9
CLIC Detector Performance Criteria/Goals Alan Bross BNL Muon Collider Design Workshop December 4, 2007 10
Track Density at CLIC Snowmass 96 MCD = 30/cm 2 @ 10 cm Alan Bross BNL Muon Collider Design Workshop December 4, 2007 11
ILC RDR – Volume 4: Detectors · 325 Participating Institutions! Alan Bross BNL Muon Collider Design Workshop December 4, 2007 12
ILC RDR – Volume 4: Detectors No one has been able to determine how many contributors Alan Bross BNL Muon Collider Design Workshop December 4, 2007 13
ILC Detector R&D · An enormous amount of work has been done over the last 10 years u Much is directly relevant to a MCD · What follows has been graciously supplied by Marcel Demarteau, the Fermilab ILC Detector R&D Leader Alan Bross BNL Muon Collider Design Workshop December 4, 2007 14
ILC Detector R&D Marcel Demarteau Fermilab TWEPP 07 – Prague, September 7, 2007
Some ILC Parameters l Time structure l l five trains of 2625 bunches per second bunch separation is 369. 2 ns (LEP: 22 s) 969 s l l l Once per train; time stamping sets time resolution Once per bunch Duty cycle (1 ms of data – 199 ms idle) allows for “power pulsing” l l 969 s Readout options driven by physics l l ~199 ms Switch power to quiescent mode during idle time Single IR with 14 mrad crossing angle Beam size: sx = 640 nm, sy = 6 nm 16
Specification for an ILC Detector l ILC detectors are precision detectors: fully reconstruct the final state over the full angular region l Identify each and every particle, with high efficiency and high purity, over the full angular range l l Differentiate between Z’s and W’s in their hadronic decay Differentiate between b- and c-quarks Differentiate between b- and anti-b quark Although these requirements are common drivers for all experiments, they are non-negotiable requirements for the ILC ! 17
The ILC Concept Detectors LDC Detector LDC GLD Si. D 4 th l GLD Si. D 4 th Premise Vertex Detector Tracking EM calorimeter Hadron calorimeter Solenoid Muon System PFA 5 -layer pixels TPC Gaseous Silicon. Tungsten Analogscintillator 4 Tesla Instrumented flux return PFA 6 -layer fine pixel ccd TPC Gaseous Scintillator. Tungsten Digital/Analog Pb-scintillator 3 Tesla Instrumented flux return PFA 5 -layer silicon pixel Silicon strips Silicon. Tungsten Digital Steel RPC 5 Tesla Instrumented flux return Dual Readout 5 -layer silicon pixel TPC Gaseous 2/3 -readouts Crystal 2/3 -readouts Tungsten-fiber 3. 5 Tesla Iron free dual solenoid Requirements: l l l Impact parameter resolution: Momentum resolution: Jet energy resolution: 18
Calorimetry
l Goal: s(E)/E ~ 3 -4% l l Paradigms: l l l Dual or Triple Readout Particle Flow Algorithm (PFA) Enabling Technologies: l l l Ability to separate Z → qq from W → qq’ (Ejet) (Ge. V) Calorimetry H 1 ATLAS ALEPH * Goal for PFA-ILC Ejet ( Ge. V) New generation of Photon Detectors Highly integrated microelectronics Strategies: l Digital versus Analogue readout 20
Multiple Readout Calorimetry l Dual-Readout: measure every shower twice l l Scintillation light: from all charged particles Čerenkov light: b=1 particles, mainly EM DREAM 200 Ge. V p By measuring separately both components can determine e/h fraction and correct the response (set e/h=1) Approaches: l Scintillating and quartz fibers embedded in Cu (DREAM); l l l no longitudinal segmentation Leadglass-Scintillator sampling Doped crystals 21
Particle Flow Algorithm l The other paradigm to obtain better energy resolution: PFA l PFA: Reconstruct momenta of individual particles in jet; avoid double counting l l Measure photons in the ECAL Measure charged particles in the tracking system Subtract calorimeter energy associated with charged hadrons Measure neutral hadrons in the HCAL (+ ECAL) l PFA: a brilliant idea ! Novelty is in reducing the role of the hadron calorimeter – and thus the hadron energy resolution – to the measurement of neutral hadrons only l Implications for the calorimetry l l Granularity, longitudinal and transverse ! Sampling of the hadron calorimeter Digital or analog readout Imaging calorimeter 22
Calorimeter Architectures l One of the main drivers for imaging calorimeters is granularity l Need to separate energy deposits from different particles Electromagnetic Active element Analogue Digital Silicon k. PIX SKIRoc MAPS Cells ~0. 5 x 0. 5 cm 2 Cells ~50 x 50 m 22 PPD readout – Scintillator Gas – – Hadronic Analogue Digital Too expensive PPD readout – – RPC GEM Micro. Megas Cells ~3 x 3 cm 2 Cells ~1 x 1 cm 2 23
Analogue Electromagnetic Calorimeter l Silicon-Tungsten sampling calorimeter l l Total Si area (incl. endcaps) ~2000 m 2 Total number of channels up to 80× 106 Average dissipated power 1 -4 μW/mm 2 LDC approach: l l l Sensitive silicon layers are on PCBs l 1 x 1 cm 2 pads, ~1. 5 m long × 30 cm wide Pad readout digitized to ~16 bits by VFE ASIC Si. D approach: l l 6” hexagonal wafers with 1024 13 mm 2 pixels Readout with one ASIC, connected to readout cable l Scintillator-Tungsten sampling calorimeter l GLD approach: l l Tile and strip configuration WLS fiber readout with Photo-detector 24
Digital Electromagnetic Calorimeter l EM calorimeter based on Monolithic Active Pixel Sensors l l l Intrinsic high granularity through wafer processing CMOS process cheaper than high resistivity pure silicon ECAL MAPS design l l Binary readout, threshold adjustment for each pixel Pixels 50μm× 50μm, 4 diodes for Charge Collection l l l 50 mm With ~100 particles/mm 2 in the shower core and 1% prob. of double hit the pixel size should be ~40 μm× 40 μm Prototype device with two types of readout Time Stamping with 13 bits (8192 bunches) Hit buffering for entire train, readout between trains Capability to mask individual pixels Total number of ECAL pixels around 8× 1011: Terapixels l Device being simulated l l Signal to Noise > 15 for 1. 8 µm Diode Size Critical issue for Terapixel system 25
Analogue Hadron Calorimeter l Planes of scintillator and absorber: l Very high granularity l l l GLD: z/x/T; LDC: tiles only 4 x 4 cm 2 x 5 mm; 1 x 20 cm 2 x 5 mm (GLD) 3 x 3 / 6 x 6 / 12 x 12 cm 2 tiles (Calice) Each element read out separately Massive number of readout channels ~50 M channels Photon detection of scintillator light l l l Collection through WLS fiber Direct coupling of detector on scintillator Enabling technology: Geiger-mode Avalanche Photo Diodes 26
Geiger-mode Avalanche Photo Diode The technology that enables this high granularity is Geiger-mode Avalanche Photo Diodes (MRS, MPPC, Si. PM, PPD) l l Array of pixels connected to a single output Signal = Sum of all cells fired; binary device ! If probability to hit a single cell < 1 Signal proportional to # photons Characteristics: l Pros l l l l Cons l l l Very compact High PDE (15~20% for 1600 pix) Insensitive to magnetic field High gain (105~106) Operational at Vbias=70~80 V Good timing resolution IRST Thermal noise rate (100 k. Hz~300 k. Hz @ 0. 5 pe) Response is non-linear due to limited number of pixels (saturation effect) Sensitive to temperature change Cross-talk and after-pulsing Vendors l 1 mm l Hamamatsu, Sens. L, IRST, Mephi, Pulsar, CPTA/Photonique, Dubna/Mikron, Kotura, a. Peak, … 3 x 3 x 0. 5 cm 3 UNIPLAST 1 mm WLS Kuraray fiber Y 11(300) 27
Digital Hadron Calorimetry l Three technologies: l l l Signal Pad(s) Mylar Glass Glue HV Channel Fishing line Resistive paint Glass Separate drift and amplif. gap Aiming at ~1 x 1 cm 2 readout Mylar Gas volume Micro MEsh GAseous Structure l l Single gap Coated glass as resistive plates Avalanche mode Readout pads ~1 x 1 cm 2 Gas Electron Multiplier (GEM) l l Signal path Resistive Plate Chamber (RPC) l l RPC Fine mesh separates 3 mm drift and 0. 1 mm amplification gaps GEM Padboard R&D l Performance metrics l l l MIP detection efficiency uniformity Readout multiplicity Noise rate, rate capability Gain experience in large scale and long-term operation and production Identify critical operational issues 28
Tracking
Tracking l Goal: l l Superb momentum resolution Robust pattern recognition and good two track separation Tolerant to high machine background Paradigms: l Silicon Tracking l l l Time Projection Chamber (TPC) l l l • superb position resolution • compact tracker many space points (~200) Two track resolution <2/5 -10 mm (r, )/(r, z) Enabling Technologies: l l Advances in Si processing Precision TPC readout 30
TPC Tracking l ILC TPC l l l ALICE TPC l l dp/p 1%, B=0. 4 T Material 3. 5% X 0 near = 0 MWPC readout, ~500 k cathode pads, pad sizes 4 x 7. 5, 6 x 10, 6 x 15 mm 2 hit resol. 800 … 1250 m r , z l dp/p 0. 1%, B=4 T Material <3% X 0 near = 0 <30%X 0 endcap pads per endcap > 106, pad size about 1 x 6 mm 2 hit resol. 100, 500 m r , z @ 4 T Readout l l l GEM Micro. Megas CMOS Pixels 31
TPC Readout l GEM l Micro. Megas Anode l two copper foils separated by polyimide l l uses 2 or more stages for safer operation high electric field inside the holes, in which multiplication takes place 50 m amplification region is displaced from the anode Micro. Megas, 2 x 6 mm 2 pads B=1 T micromesh sustained by pillars l l l amplification between mesh and pads/strip plane single stage 50 m amplification region includes the anode Now “Bulk Micromegas” can be obtained by lamination of a woven grid on an anode with a photoimageable film The ILC-TPC resolution goal, ~100 µm for all tracks, appears feasible. 32
TPC CMOS Readout l Use bare CMOS chip as anode to directly collect signals from GEMs or Micromegas: Medi. Pix chip l l l Currently: l l l 3 rd coordinate (time) being added: Time. Pix chip Integration of GEM/Micromegas grid and CMOS sensor through wafer processing (In. Grid) Prospects: l l GEM foil integrated on chip Charge collection with granularity matching primary ionization cluster spread On-chip processing of signals l Ionization cluster counting is possible to improve part. id. performance Potential for large improvements in pattern recognition and d. E/dx “Digital Bubble Chamber” 33
TPC R&D l Many prototype TPC’s built l l Interchangeable gas-amplification Wide range of studies: l l l l Gas and resolution studies Candidate gas amplification devices l Direct comparison of triple-GEM and Bulk Micromegas Ion/electron transmission studies Ion feedback measurements Plan for large prototype TPC l Cornell/Purdue small prototype 60 cm drift length, 80 cm diameter Interchangeable gas-amplification modules designed to directly compare gas-amplification technologies Need for large bore high magnetic field! Large TPC D=80 cm R&D synergistic with T 2 K l T 2 K will have 3 TPCs l l l 72 Micromegas modules Total area ~ 9 m 2 124416 readout channels 34
Silicon Tracker l All silicon tracking, Si. D l l “Power-pulsing” allows for gas cooling Hybrid-less design l l l Si. D 100 x 100 mm 2 sensor from 6” wafer with 1840 (3679) readout (interm. ) strips Integration of pitch adapter through 2 nd metal layer in sensor for signal routing Sensor (1840 channels) read out with two asics (k. Pix) Power and clock routed over the sensor ! ECT SIT Silicon as “ intermediate layers” l Double-sided layers to act as tracker l l SET FTD d-s silicon R&D actively being pursued in Korea Single-sided layers to “link” subdetectors l Long-ladders with associated FE Asic LDC 35
Vertexing
Vertexing l Goal: l l Superb impact parameter resolution Minimal material budget: < 0. 1%X 0 / layer l l l Equivalent to 100 m of Silicon Minimal power consumption (<50 W) Ability to determine quark charge Tolerant to high machine background Paradigms: l l Readout during the train Readout in-between trains 37
Silicon Technologies 38
ILC Candidate Technologies l CCD’s l l l l l ISIS Mimosa series (Ires) INFN LDRD 1 -3 (LBNL) CAP 1 -4 (Hawaii) Chronopixel (Oregon/Yale) LBL-LDRD 3 MIIMOSA-n SOI l l l CPC 2 CMOS Active Pixels l l Column Parallel (UK) Fine Pixel (Japan) ISIS (UK) Split Column (SLAC) 3 D American Semiconductor/FNAL LDRD-SOI (LBNL) CAP 5 (Hawaii) OKI/KEK CAPS 4 VIP (FNAL) DEPFET (Munich) l l 3 D DEPFET 39
Sensor Architectures l An incomplete attempt at listing some of the current architectures design for ILC pixel detectors CMOS MAPS CCD Rolling Shutter Mimosa 1 -N LDRD 1, 2 Normal CCD Column Parallel Mimosa 8 LDRD 3 CP-CCD SC-CCD Pipelined Storage Mimosa-12 CAP ISIS Time Stamp Chronopixel l DEPFET SOI 3 D LDRD-SOI DEPFET/ CURO CAP-5 ASI SBIR VIP-1 With apologies to all other technologies, I will only mention three: CP-CCD, Mimosa, 3 D 40
Column Parallel CCD CP-CCD: read out a vector instead of a matrix l l l Readout time shortened by orders of magnitude But every column needs its own amplifier and ADC: readout chip Need to operate at 50 MHz to meet ILC readout rate spec. Driving of CP-CCD is a major challenge M l 2 nd generation large area sensors : CPC 2 l l l N l Dedicated readout chip l l Column Parallel CCD Readout time = N/fout Devices with 2 -level metal clock distribution 25 μm and 50 μm epi layers Reaches 45 MHz operation (designed for 50 MHz) CPR 2, bump bonded at VTT to CPC 2 Dedicated clock drive chip l CPD 1, requirement of 2 Vpk-pk at 50 MHz over 40 n. F CPR 2 CPC 2 41
Mimosa l Mimosa-16 being developed as beamline telescope for DESY (and Fermilab) testbeam: l l l Final geometry: l l l l 1024 columns of 512 pixels, 20 µm pitch Expected hit resolution < 2. 5 µm Sensitive area = 20. 48 x 10. 24 mm 2 pixels with integrated CDS sensor with integrated 4/5 -bit ADC possibly zero-suppression Read-out speed l l Column parallel readout 32 // columns of 128 pixels (pitch: 25 µm) ~11– 16 µm epitaxy on-pixel CDS default tr. o. = 512 lines / 5 MHz ~ 100 µs Possible variant l 1280 columns of 640 pixels, 16 µm pitch with binary readout 42
Vertical Integration – 3 D l l A 3 D device is a chip comprised of 2 or more layers of semiconductor devices which have been thinned, bonded, and interconnected to form a monolithic circuit Advantages of 3 D l l l Increased circuit density due to multiple tiers of electronics Fully active sensor area Independent control of substrate materials for each of the tiers l l l Process optimization for each layer Ability to mate various technologies in a monolithic assembly Technology driven by industry l l Reduce R, L, C for higher speed Reduce chip I/O pads Provide increased functionality Reduce interconnect power, crosstalk l Critical issue are: l l Layer thinning to < 10 m Precision alignment (< 1 m) Bonding of the layers Through-wafer via formation 43
VIP Chip l 3 D chip Vertical Integrated Pixel (VIP) chip submitted by Fermilab to DARPA funded MIT-LL 0. 18 m 3 D process l Chips due to arrive in a couple of weeks; key features: l Analog pulse height, sparse readout, high resolution time stamp, front-end power ~ 1875 W/mm 2 (before cycling), 175 transistors in 20 x 20 µm 2 pixel. 3 D Via Tier 3 8. 2 µm Tier 2 7. 8 µm oxide-oxide bond Tier 1 6. 0 µm 2000 ohm-cm p-type substrate Buried Oxide (BOX) 400 nm thick 44
Sensor Technology l Device thinning is becoming very common l l CCD’s are regularly thinned to 20 m LBL has thinned over 15 Mimosa CMOS MAPS chips down to 40 mm l l Yield of functional chips ~90% Studies of charge collection and S/N before/after back-thinning l Some evidence of small signal loss after thinning Sensors will be used in Fermilab beam telescope Fermilab has thinned BTe. V Fpix chips/wafers to 15/20 m with ~75% yield Detector bias l 20 Thinned Edgeless Sensors l l Sensors sensitive to the edge can be fabricated by a combination of trench etching, thinning, and laser annealing Fermilab producing a set of detectors thinned to 50 -100 m at MIT-LL for beam and probe tests m Diode implants Trench on detector edge filled with poly and connected to bottom implant To other pixels Implant with laser annealing Detector Cross section near one detector edge 45
Conclusions · For the most part, currently available or developing technology will meet the performance criteria as stated in Snowmass 96 for u u u Muon System Calorimetry Central Tracking (r>20 -50 cm) Non-Silicon s However, neutrons could still be a problem for readout electronics. For a TPC option, for example, front-end electronics at the end planes would have to be shielded from low-energy neutrons (longitudinal fluence) · Inner Tracking (Vertexing) presents problems due to the large n fluence Alan Bross BNL Muon Collider Design Workshop December 4, 2007 46
Conclusions II · Vertexing example: CMS Pixels (@ r=4. 3 cm) u » 2 X 1014 n for 5 years of running @1034 cm-2 s-1 s Lifetime limit · Options u Thinner detectors s u u 40 mm vs. 300 (X 8) NIEL Non-Ionizing Energy Loss Amorphous Si CVD Diamond detectors s ~ X 100 hardness over Si · LARP Detector R&D u Prototype Diamond detector system (pixel luminosity telescope (PLT) for CMS Alan Bross BNL Muon Collider Design Workshop December 4, 2007 47
Conclusions III · Ongoing detector R&D (ILC, LARP) is addressing many detector issues for the MCD · First order of business - Need u u Next iteration on interaction region design Next iteration on collider ring design s May ameliorate some of the radiation background problems · The outcome of these design studies can then be used as input to a new round of radiation background studies for the MCD u Lead to directions for dedicated detector R&D for the MCD Alan Bross BNL Muon Collider Design Workshop December 4, 2007 48
- Slides: 48