1 / 24

Linear Collider Flavour Identification Collaboration: Case for Support

Introduction to the ILC and LCFI: Physics at the ILC LCFI physics studies Sensor design and testing Mechanical studies Proposed LCFI programme: WP1 Simulation and physics studies WP2 Sensor development WP3 Readout and drive electronics WP4 External electronics WP5 Integration and testing

sissy
Download Presentation

Linear Collider Flavour Identification Collaboration: Case for Support

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Introduction to the ILC and LCFI: Physics at the ILC LCFI physics studies Sensor design and testing Mechanical studies Proposed LCFI programme: WP1 Simulation and physics studies WP2 Sensor development WP3 Readout and drive electronics WP4 External electronics WP5 Integration and testing WP6 Vertex detector mechanical studies WP7 Test-beam and EMI studies Linear Collider Flavour Identification Collaboration:Case for Support

  2. LCFI Collaboration P Allport3, D Bailey1, C Buttar2, D Cussans1, CJS Damerell3, J Fopma4, B Foster4, S Galagedera5, AR Gillman5, J Goldstein5, T Greenshaw3, R Halsall5, B Hawes4, K Hayrapetyan3, H Heath1, S Hillert4, D Jackson4,5, EL Johnson5, N Kundu4, AJ Lintern5, P Murray5 A Nichols5, A Nomerotski4, V O’Shea2, C Parkes2, C Perry4, KD Stefanov5, SL Thomas5, R Turchetta5, M Tyndel5, J Velthuis3, G Villani5, S Worm5, S Yang4. 1 Bristol University2 Glasgow University 3 Liverpool University 4 Oxford University 5 Rutherford Appleton Laboratory

  3. Standard Model of particle physics is clearly incomplete. From 2007, LHC experiments will study pp collisions √s = 14 TeV giving large mass reach for discovery of new physics. Precision measurement (of masses, branching ratios etc.) complicated by hadronic environment. International consensus: e+e- LC operating at up to √s ~ 1 TeV needed in parallel with the LHC, i.e. start-up in next decade. Detailed case presented by LHC/LC Study Group: hep-ph/0410364. International Technology Review Panel recommended in August 2004 that superconducting technology be used for accelerating cavities. Global effort now underway to design SC ILC, director Barry Barish. Timeline defined by ILC Steering Group foresees formation of experimental collaborations in 2008 and writing of Technical Design Reports in 2009. Agreement that vertex detector technology be chosen following “ladder” tests in 2010. The International Linear Collider

  4. Many of interesting measurements involve identification of heavy quarks. E.g. determination of branching ratios of Higgs boson. Are BRs compatible with the SM? Physics studies can also benefit from separation of E.g. Reduce combinatorial background. Allows determination of Higgs self-coupling. Flavour and quark charge identification at the ILC

  5. Increases sensitivity to new physics. E.g. effects of large extra dimensions on Study ALR = (sL – sR)/stot as a function of cosq. For muons, effects of ED not visible: Changes much more pronounced for c (and b) quarks: Requires efficient charge determination to large cosq. Quark charge identification

  6. Provides new tools for physics studies. E.g. measure top polarisation in decay Top decays before hadronisation. Anti-strange jet has 1 – cosq distribution w.r.t. top polarisation direction. Distinguish between t and by tagging b and c jets and determining quark charge for (at least) one of these jets. Example of physics made accessible using this technique: Determine tanb and tri-linear couplings Ab and At through measurements of top polarisation in Quark charge identification b t W+ c

  7. Average impact parameter d of B decay products ~ 300 mm, of charmed particles less than 100 mm. d resolution given by convolution of point precision, multiple scattering, lever arm and mechanical stability. Multiple scattering significant despite large √s at ILC as charged track momenta extend down to ~ 1 GeV. Resolve all tracks in dense jets. Cover largest possible solid angle: forward/backward events are of particular significance for studies with polarised beams. Stand-alone reconstruction desirable. Implies typically: Pixels ~ 20 x 20 mm2 Hit resolution better than 5 mm. First measurement at r ~ 15 mm. Five layers out to radius of about 60 mm, i.e. total ~ 109 pixels Material ~ 0.1% X0 per layer. Detector covers |cos q| < 0.96. B+ d1 d2 B- Vertex detector performance goals track 1 track 2

  8. Minimum beam pipe radius 14 mm. Pair background at this radius in ~ 4T field causes 0.03 (0.05) hits per BC and mm2 at √s = 500 (800) GeV. Bunch train structure: For 109 pixels of size 20 x 20 mm2, implies readout or storage of signals ~ 20 times during bunch train to obtain occupancy less than ~ 0.3 (0.9) %. Must withstand: Radiation dose due to pair background of ~ 20 krad p.a. Annual dose of neutrons from beam and beamstrahlung dumps ~ 1 x 109 1 MeV equiv. n/cm2. Must cope with operation in 4T field. Must be robust against beam-related RF pickup and noise from other detectors. 337 (189) ns 0.2 s 2820 (4500) 0.95 ms Constraints due to machine and detector

  9. Here using CCDs: VXD surrounded by ~ 2 mm thick Be support cylinder. Allows Be beam pipe to be of thickness of ~ 0.25 mm. Pixel size 20 x 20 mm2, implies about 109 pixels in total. Standalone tracking using outer 4 layers. Hits in first layer improve extrapolation of tracks to IP. Readout and drive connections routed along BP. Important that access to vertex detector possible. Conceptual vertex detector design

  10. Amount of material in active region minimized by locating electronics only at ends of ladders. Resulting material budget, assuming unsupported silicon sensors of thickness ~ 50 mm: Conceptual detector design Material of:beam pipe five CCD layerscryostatsupport shell

  11. Performance of vertex detector investigated and optimised using Monte Carlo simulations. E.g. study effect on impact parameter resolution of variations in beam pipe radius, material budget and number of layers in vertex detector. Observe moderate effects due to increase in material budget, severe degradation due to increase in beam pipe radius. Impact parameter resolution Vertex detector performance – impact parameter

  12. Simulate flavour ID inevents, here at Z0 pole. Feed information on impact parameters and vertices identified using Zvtop algorithm into neural net. Modest improvement in beauty tagging efficiency/purity over that achieved at SLD. Improvement by factor 2 to 3 in charm tagging efficiency at high purity. Charm tag with low uds background interesting e.g. for Higgs BR measurements. Efficiency and purity of tagging of beauty and charm jets: Flavour identification performance

  13. Must assign all charged tracks to correct vertex. Multiple scattering critical, lowest track momenta ~ 1 GeV. Sum charges associated with b vertex: Quark charge identification for neutral B requires “dipole” algorithm. Quark charge identification performance

  14. Standard CCDs cannot achieve necessary readout speed LCFI developed Column Parallel CCD with e2v technologies. Sensors for the vertex detector – CCDs

  15. First of these, CPC1, manufactured by e2v. Two phase, 400 (V)  750 (H) pixels of size 20  20 μm2. Metal strapping of clock gates. Two different implant levels. Wire/bump bond connections to readout chip and external electronics. Direct connections and 2-stage source followers: Direct connections and single stage source followers (20 mm pitch): Sensors – CPCCD

  16. Standalone CPC1 tests: Noise ~ 100 e- (60 e- after filter). Minimum clock potential ~1.9 V. Max clock frequency above 25 MHz (design 1 MHz). Limitation caused by asymm. clock signals due to single metal design. Marry with CMOS CPCCD readout ASIC, CPR1 (RAL): IBM 0.25 μm process. 250 parallel channels, 20 μm pitch. Designed for 50 MHz. Sensors – CPC1 and CPR1

  17. Bump bonding of CPC1 and CPR1done at VTT: CPR1 bump bonded to CPC1, signal from charge channels: Observe ~ 70 mV signal, expected 80 mV, good agreement. Sensors – CPC1 and CPR1

  18. Study CTI in CCD58 before and after irradiation (90Sr 30 krad). Measure decrease in charge from 55Fe X-rays as func. of number of pixels through which charge transferred. Compare data with simulations performed using ISE-TCAD. Extend to CPCCD. CCD radiation hardness tests

  19. In-situ storage image sensor. Signal collected on photogate. Transferred to small CCD register in pixel. Signal charge always buried in silicon until bunch train has passed. Column parallel readout at ~ 1 MHz sufficient to read out before arrival of next bunch train. ISIS1 being built by e2v. 40  160 μm2 cell containing 3-phase CCD with 5 pixels. Sensors – ISIS

  20. Monolithic Active Pixel Sensors developed within UK. Ongoing development for scientific applications by MI3 collaboration. Storage capacitors added to pixels to allow use at ILC, Flexible Active Pixel Sensors. Vreset Vdd Select Reset Out Sensors – FAPS

  21. Present design “proof of principle”. Pixels 20 x 20 mm2, 3 metal layers, 10 storage cells. Test of FAPS structure with LED: 106Ru b source tests: Signal to noise ratio between 14 and 17. MAPS demonstrated to tolerate radiation doses above those expected at ILC. Sensors – FAPS

  22. Thin ladder design. Stresses introduced when silicon is processed imply “unsupported” option requires Si thickness > 50mm. “Stretching” maintained longitudinal stability, but provided insufficient lateral support. Re-visit using thin corrugated carbon fibre to provide lateral support. Supporting CCD on thin Be substrate studied: Problems observed at low Tin FEA calculations. Confirmed by measurements in Lab. Adhesive 0.2mm Beryllium substrate (250 μm) Mechanical considerations

  23. Importance of good matching of coefficients of thermal expansion of silicon and substrate demonstrated in laboratory measurements: Now exploring use of silicon and reticulated vitreous carbon foam sandwich: Diamond structure also recently purchased from Element Six: Mechanical considerations

  24. Progress made in understanding physics accessible with a precise vertex detector at the ILC via: Flavour identification. Determination of b, c charge. Column Parallel CCD development progressing: LCFI will soon have sensors of scale close to that required for the ILC. A major remaining challenge is the construction of low mass CCD drive circuitry. Studies of ISIS and FAPS storage sensors initiated. Mechanical studies have demonstrated: Unsupported Si will not result in lowest mass sensors. Emphasis shifted to new materials. Milestones of previous proposal met or surpassed in last three years. Summary

More Related