1 / 23

Rick Gaitskell Department of Physics Brown University Source at gaitskell.brown

XENON Experiment - SAGENAP Factors Affecting Detector Performance Goals and Alternative Photo-detectors. Rick Gaitskell Department of Physics Brown University Source at http:// gaitskell.brown.edu. SAGENAP Questions (020313).

spike
Download Presentation

Rick Gaitskell Department of Physics Brown University Source at gaitskell.brown

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. XENON Experiment - SAGENAPFactors Affecting Detector Performance Goals and Alternative Photo-detectors Rick Gaitskell Department of PhysicsBrown University Source at http://gaitskell.brown.edu

  2. SAGENAP Questions (020313) • What are the FTE commitments of each of the of the people listed in the final table over the two-year R&D? For each person, estimate also the FTE commitment to each of the other projects, and note faculty teaching loads as necessary. • What do you estimate to be the construction costs for the 100 kg experiment? Including shielding, readout, and support equipment costs (Xe purification, cryogenics). • We’d like to understand better the nature of the competition. Specifically, what is driving your two-year timescale? Have there been discussions for the collaboration with the other Xe TPC groups in the world, and what were the results? Would a merging of efforts be possible after the first phase of this project? • What is the expected trigger rate and event dead time? • The sensitivity plots on p14 of the presentation and the Zeplin IV Feb Status Report Presentation appear different by a factor ~50 (both are 1 tonne). Can you explain the reasons? • Please make a list of the deliverables from this 2-year R&D program and the person responsible for each. What design decisions will be made at the end of year 1 and year 2?

  3. SAGENAP Q3 • Why 2 years? • LXeGRIT is a prototype “-1” that already demonstrates much of base engineering. In this proposal we are focusing on a number of key technologies that will enable us to realise radioactive background, signal-to-noise and discrimination goals. The timeline of 2 years to construct 10 kg dark matter chamber is necessary, in order to contribute meaningfully to “world-wide” Xe beauty contest (100 kg+1 tonne) that we expect will occur in ~2004. • Collaboration Discussions • Japanese • Columbia has close collaboration with Waseda, which will continue. • UK • Collaboration discussed. They requested our direct commitment of 100 kg and 1 tonne phases to Boulby site, which is not necessarily consistent with large US involvement. • UCLA • Collaboration was discussed, but declined. • Future Merger of Xe Efforts • Definitely! - Given scale (manpower resource estimate ~25 scientists) of final projects this will be mandatory at US level • Probable that maximum of 2 Xe experiments will be operated worldwide

  4. Q4 Trigger Rate/Dead Time • We trigger on “large” light signal: - Direct light on CsI photocathode (>20 p.e.) amplified by Proportional Scintillation (~1000). - Requirement ≥ 2 PMT - Coincident with direct light on PMTs (>4 p.e.) exactly 150µs earlier • Rate above ground - Typical 104 /keV/kg/day, integration over spectrum ~ 20 Hz/kg, 10 kg mass -> 200 Hz - Cosmics (charged) ~few additional Hz • Rate below ground - Depdendent on shield: of order 1 Hz (conservatively) • Dead Time - Gas drift time: 150 µs - Post triggered FADC read out with buffer - Signals ~1 µs => FADC 5-10 Msample/sec - 37 FADC channels (for 100 kg) - 32 ADC for PMT in Xe shield 150 µs (300 mm)

  5. Q5 Sensitivity Projections

  6. Q2 Construction Costs • What do you estimate to be the construction costs for the 100 kg exp? • Include shielding, readout and support equipment costs (Xe, Purification, Cryo) • [ $0.32M ] Xe: 100 kg Active Target + ~100 kg Active Shield • $1.6/g ($6/g CDMS cryo-detector grade) • 1 module $320k of Xe • (1 tonne active Xe -> $1.6m) • [ ~$1M ] Xe Purification + Gas System / Handling / Circulation • [ ~$0.5M] Kr Removal • [ ~$1M ] Design + Construction of 1x100 kg module • [ ~$0.2M ] Clean Room Class 1000 • [ ~$1M ] Readout • [ $0.6M ] Shielding • [$4.6M] Total

  7. Goal Why is good detection/discrimination performance required down to 16 keVrecoil (4 keV electron equivalent)?

  8. Xe Eth=16 keVr gives 1 event/kg/day Very Typical WIMP Signal • Low Thresholds Vital • Graph shows integrated event rates for E>Er for Xe (green), Ge (red) and S (blue) • Large nuclei enhanced by nuclear coherence, however, in reality <<A2 … Xe WIMP rate for Er > 16 keVr is within factor 2 of maximum achievable rate (Er>0) equivalent kg/kg to low threshold Ge detector 5x better kg/kg than light nucleus (e.g. S in CS2) Example cross-section shown is at current (90%) exclusion limits of existing experiments

  9. Form Factor Suppression • Form Factor makes very significant modification to naïve ~A2 rate • … due to loss of coherence (since qr>>1) Dashed lines show ~A2 before considering q>0 Form Factor suppression Note Rapidly Falling Rate

  10. “Acceptable trade off” Good Performance Must Be Established at “Threshold” • Low threshold vital, since rate falls rapidly with energy • 10% of signal @ Recoil Energy >35 keVr (assuming 100 GeV WIMP) • Assuming 25% Quenching Factor this is equivalent to <8.8 keVee • ~45% of signal @ Recoil Energy >16 keVr • Equivalent to 4 keVee • Factor 2x sacrifice in “effective detector mass” relative to zero threshold rate • Need to maximise performance in low detection signal regime • Ensure that WIMP identification/background discrimination is working well at ~4 keVee

  11. e e e e e e Available Signal: UV Photons & Electrons • Focus on two types of messengers from primary interaction site • UV Photons (178 nm) from Xe scintillation • Consider energy required to create photons • Will not consider details of generation mechanism • Note that UVg generated via both Xe* and Xe+ mediated channels • No re-adsorption term to consider • “Free” electrons separated from Xe+ ions • Consider energy required to create electron-ion pairs • Need to consider loss due to local recombination in densely ionised region Summarise existing data from liquid Xe detector studies… • Electron Recoils from 1 keVee (electron equivalent) Gamma Events • Nuclear Recoils from 1 keVr (recoil) WIMPs/Neutrons

  12. Available Signal in Liq. Xe • Summary • The ranges shown reflect spread in existing experimental measurements • Note that the table considers signal from either 1 keV gamma or nuclear recoil event • 60 excitations / keV is equivalent to ~16 eV / excitation • Zero field electron-ion #’s in [ ] are inferred, but are signal is not measured (extracted) directly

  13. 40% 90% Available Signal in Liq. Xe (2) • Gamma Event • UV Photons • w ~13-15 eV / photon for zero field • As soon as field is applied (>0.2 kV/cm) electron-ions no longer recombine and this route (~50%-60%) for generation of photons disappears • Electrons • Also w ~13-15 eV / electron, Note that for zero field electrons are not measured directly since no drifting occurs • >~90% of electrons are extracted in high field

  14. ~100% 3-8% Available Signal in Liq. Xe (3) • Nuclear Recoil Event • UV Photons • w ~50-70 eV / photon, (Lindhard) Quenching Factor measured as 20-25% • Ionisation density is very much higher for nuclear recoil so even with high applied field most electron-ions recombine • Electrons • Lindhard Quenching Factor also applies to initial generation of electron-ions • Extraction of electrons from densely ionised region is very inefficient. • Literature quotes extraction in range (0.5-1.0%)/kV of applied field (in this case use 8 kV/cm so 4-8%) ( Note: Bernabei (DAMA) use Quenching Factor of 40% which has not been confirmed elsewhere )

  15. Summary - High Field Operation • Detection of primary scintillation light is a challenge • ~12 UV photons / keV recoil energy • Extraction of electron(s) from nuclear recoil densely ionised region is big challenge • We require observation of this signal to ensure correct identification of nuclear recoil event • ~0.4-1.2 electrons / keV recoil energy • Note once electron extracted from liquid to gas, significant gain ~1000 UVg / electron makes signal easy to observe

  16. Baseline - Simulation Results 16 keV recoil threshold event • Assumes 25% QE for 37 phototubes, and 31% for CsI cathode • A 16 keV (true) nuclear recoil gives ~ 24 photoelectrons. The CsI readout contributes the largest fraction of them • Multiplication in the gas phase gives a strong secondary scintillation pulse for triggering on 2-3 PMTs. • Coincidence of direct PMTs sum signal and amplified light signal from CsI • Main Trigger is the last signal in time sequence post-triggered digitizer read out Trigger threshold can be set very low because of low event rate and small number of signals to digitize. PMTs at low temperature low noise • Even w/o CsI (replaced by reflector) we still expect ~6 pe. Several ways to improve light collection possible

  17. Nuclear Recoil Event ~Threshold 16 keVr • Detection of electrons (drifted) • 0.5-1.0% / (kV applied field) extraction from dense ionised region (avoiding self recombination) • 4-18 electrons drifted toward liquid surface • In high field once electrons start drifting ~100% extraction from liquid • Gas Gain • ~1000 UVg from each electron in gas • Signal is localised to xy position of original interaction • Large signal in PMT • Even considering PMT/geometry efficiency this gives a large signal • Nuclear Recoil of 16 keVr (Threshold) • QF 25% -> 4 keVee • 300 UVg into 4π • Detection in Phototubes • Nominal Geometric Efficiency ~6% • Tubes have a active fill factor of ~50% at top of detector • Photons lost in windows (T=80%) and by wires (T=80%) giving ~60% • Total Internal Reflection(TIR) at liquid surface (n~1.65), acceptance ~20% • Ignore Teflon losses for this calc. • Tube photocathode Quantum Efficiency ~30% • 300g x 2% = 6 photoelectrons • Generation of electrons in CsI photocathode • Nominal Geometric Efficiency ~20-60% • CsI covers entire bottom surface • Due to TIR and Teflon this value is high • Strong position sensitivity, poor energy resolution • CsI cathode Quantum Efficiency ~30% • 300g x 6-18% = 20-60 photoelectrons These are ball-park numbers - Full simulation actually traces rays and includes all scattering

  18. Why is photodetector performance critical? • A factor 2 increase in threshold 16 keVr -> 32 keVr • Factor 5 loss in effective mass of detector for WIMP search • A factor 2 decrease in threshold 16 keVr -> 8 keVr • Factor <2 improvement in effective mass of detector for WIMP search • However, lower threshold will, of course, improve background identification/rejection

  19. Existing Photodetector Summary • Hamamatsu Low Temperature/Liquid Tube (6041) • Baseline design for XENON • Metal construction that has been shown to work in liquid Xe • Not Low Background: Could be made low background • Low Quantum Efficiency~10-15% • New Hamamatsu Low Background Tube (R7281) • Being tested by Xmass Collaboration • Room temperature tests only so far • Metal construction, and giving lower backgrounds • ~500 per day (XENON baseline target is 100 per tube per day) • Higher Quantum Efficiency~27-30% • Uses longer optics which give better focusing (could be accommdated in XENON)

  20. New Photodetectors • Micro-channel Plate • Burle 85001 • ~30% Quantum Efficiency (since photocathode can be selected separately) • Promising for low temperature operation • Large area (5x5 cm2) and compact design (few feed-throughs) • Investigate radioactive background situation • Large Area Avalanche Photodiodes • Advance Photonix / Hamamatsu • 100% Quantum Efficiency demonstrated at UV 178 nm (windowless) • Operation in liquid Xe has been demonstrated • “Large Area” 0.5-2 cm2 device available • Silicon construction is intrinsically low background/investigate packaging • Recent progress in device fabrication • leakage currents (dark noise) has been reduced significantly & benefits considerably from low temperature operation (<1 pA/cm2) (idark)170K~ 10-4 (idark)RT

  21. Effective Quantum Efficiency - LAAPD (Windowless) physics/0203011 demonstrate ~100% QE at 178 nm Advancedphotonix see also recent paper from Coimbra (Portugal) Policarpo Group physics/0203011

  22. XENON TPC Signals Time Structure • Three distinct signals associated with typical event. Amplification of primary scintillation light with CsI photocathode important for low threshold and for triggering. • Event depth of interaction (Z) from timing and XY-location from center of gravity of secondary light signals on PMTs array. • Effective background rejection direct consequence of 3D event localization (TPC) t~45 ns 150 µs (300 mm)

  23. Operation of LAAPD Array in Geiger Mode • Operation of sensor large pixellated array in “binary” mode • High voltage bias regime • Single photon causes flip - readout hit time only (not proportional mode) • Device recovery based on either passive (resistor) or active control of bias voltage • Dark Matter experiment is most concerned with few photon regime • Primary scintillation detection is starved of signal • Investigate Hamamatsu 32-channel APD array (S8550)

More Related