Searching for Dark Matter From XENON10 to LUX. Luiz de Viveiros

Transcription

1 FCT - Universidade de Coimbra - Café com Física Searching for Dark Matter From XENON10 to LUX Luiz de Viveiros Brown University Physics Department De Viveiros - Brown University April 2009 v07 <1>

2 Summary Introduction to Dark Matter Detecting Dark Matter with liquid Xe The XENON10 Experiment Design and Deployment Results The LUX Experiment Building a better detector make it big! Increase sensitivity by ~100x Initial run at surface: LUX0.1 XENON10 22 kg LUX 350 kg De Viveiros - Brown University April 2009 v07 <2>

3 Composition of the Cosmos Dark Matter 23% Free H and He 4% Stars and Gas 0.5% Dark Energy 73% Heavy Elements (us) 0.03% ΛCDM Model De Viveiros - Brown University April 2009 v07 <3>

4 Dark Matter Evidence: Galactic Velocity Measurements Velocities of galaxies in galactic clusters Fritz Zwicky, 1936 measurement of radial velocities in Coma cluster Spiral Galaxies Bulk of luminous matter v ~ const data Using virial theorem, concluded that M stars ~ 0.5% M cluster bulge, disk & halo Galaxy rotation curves Expect v ~ r -1/2 (outside bulge) Observe v ~ constant => M(r) / r ~ constant => Dark Matter Halo bulge disk v ~ r 1/2 halo M (r ) light sun bulge disk r Coma cluster (image from SDSS) M(r) r const (r r core ) De Viveiros - Brown University April 2009 v07 <4>

5 Cosmic Microwave Background WMAP 5-year results (2009) (+SN+BAO) Best fit to Λ-CDM model gives density parameters: Ω total = 1.02 ± 0.02, Ω m = ± 0.015, Ω b = ± Matter (Ω m ) Baryonic Matter (Ω b ) Discrepancy Nonbaryonic Dark Matter T = K Best fit to Λ-CDM model Cosmic Microwave Temperature Fluctuations 5-year map (dipole + galaxy removed) Temperature Fluctuation by Angular Size De Viveiros - Brown University April 2009 v07 <5>

6 Dark Matter Candidate: WIMP WIMPs: Stable (or long lived) particles, relics from the Big Bang Supersymmetry independently predicts a massive, weakly interacting particle The Neutralino ( χ ): Lightest Supersymetric Particle (LSP) WIMPs (and neutrons) scatter elastically off nuclei Photons and electrons scatter off atomic electrons Recoil energy spectrum and rate dependent on local dark matter density ρ 0 and velocity distribution Galatic Dark Matter Relic Density ρ GeV/cm3 ( 3 WIMPs / liter for Mχ = 100GeV/c 2 ) Recoil energy few kev tens of kev (require detectors with low threshold) WIMPS and Neutrons nuclear recoils Photons and Electrons electron recoils De Viveiros - Brown University April 2009 v07 <6>

7 Typical WIMP Recoil Signal Calculated differential event rates for Xe and Ge targets m WIMP = 100 GeV/c 2 and σ WIMP = 7x10-46 cm 2 (estimated LUX sensitivity) Standard Halo parameters Assumes spin-independent interaction 1 dru = 1 event / kev / kg / day Ge Thresh at 5 kevr 1 event / 3,000 kg-days σ WIMP = 7x10-46 cm 2 (LUX extimated sensitivity) Xe De Viveiros - Brown University April 2009 v07 <7>

8 WIMP signals in Xe: Light and Charge Light (S1): UV Photons (175 nm) from Xe Scintillation W ph = 21.6 ev Enhanced by local recombination Charge (S2): Electrons separated from Xe+ ions: W ~ 15.6 ev Local recombination in densely ionized region suppressed with high electric fields Neutrons, WIMPs => Slow nuclear recoils => strong recombination => S1 preserved, but Ionization S2 strongly suppressed γ, e-, μ, (etc) => Fast electron recoils => Weaker S1, Stronger S2 Nuclear / Electron Recoil Singlet 3 ns e- e - e - e e - - e - e- Excitation Xe* Xe 2 * + Xe Triplet 27 ns Ionization Xe + Xe Xe e - Xe** + Xe 175 nm 175 nm Similar mechanism 2Xe 2Xe for all noble liquids. Ar: Singlet = 10 ns Triplet = 1500 ns λ = 128 nm De Viveiros - Brown University April 2009 v07 <8>

9 Dual Phase Xe Detector Dual Phase: Liquid and Gas Xe Primary Scintillation in liquid (S1) Ionization signal from nuclear recoil too small to be directly detected => extract charges from liquid to gas and detect much larger proportional scintillation signal (S2) Time Proportional ~1 µs width µs depending on depth Primary ~40 ns width Light Signal UV ~175 nm photons Gas Liquid S2 e- e- e- e- e- S1 Electron Drift ~2 mm/µs Anode E gas Grid E drift Cathode Time-Projection Chamber: Z-position of recoil proportional to drift time Incoming Particle Scale Exaggerated E gas > E drift De Viveiros - Brown University April 2009 v07 <9>

10 Electron Recoils vs. Nuclear Recoils Discrimination: Distinctly different S2 / S1 ratio for electron / nulcear recoils Electron Recombination for Nuclear Recoils is larger Light signal (S1) is preserved Ionization signal (S2) is smaller (S2/S1) NR < (S2/S1) ER Sensitive Volume (15 cm) Incoming Particle S2 e- e- e- e- e- S1 Electron Drift ~2 mm/µs Anode E gas Grid E drift Cathode S1 S2 Electron Recoil Event Large S2/S1 S1 S2 Nuclear Recoil Event Small S2/S1 De Viveiros - Brown University April 2009 v07 <10>

11 Dual Phase Xe Detector - Animation De Viveiros - Brown University April 2009 v07 <11>

12 XENON10 De Viveiros - Brown University April 2009 v07 <12>

13 The XENON10 Collaboration Columbia University, Brown University, Case Western Reserve University, RWTH-Aachen University, Yale University, Rice University, Lawrence Livermore National Laboratory, Laboratori Nazionali Gran Sasso (Italy), Universidade de Coimbra De Viveiros - Brown University April 2009 v07 <13>

14 Laboratori Nazionali Gran Sasso (Italy) 3100 m.w.e. µ flux reduced x10-6 compared to sea level 24 µ / m 2 / day De Viveiros - Brown University April 2009 v07 <14>

15 Deployment at Gran Sasso LdV on site for the commissioning, construction, deployment and operation of detector Unshielded detector installation March Laboratori Nazionali Gran Sasso (LNGS) Began detector calibration end of March Shield construction and detector installation: May August 2006 Calibrations and unblinded background data: Sept. Dec. 1st 2006 WIMP search runs: Oct Feb Unblind WIMP search data (~ 60 live-days) on April 8, 2007 XENON10 (August 2006) Detector installation in Shield Brown group De Viveiros - Brown University April 2009 v07 <15>

16 XENON10 Responsibilities Simulation of backgrounds Construction of the XENON10 background model Shield Design Advising on design detector Screening of materials at the SOLO counting facility High-Purity Ge detector at Soudan Mine DAQ and Trigger design, construction and operation Involved at all stages of detector construction, calibration and operation Analysis code development Data analysis Gamma and Neutron calibrations WIMP search data De Viveiros - Brown University April 2009 v07 <16>

17 Deployment at Gran Sasso De Viveiros - Brown University April 2009 v07 <17>

18 The XENON10 Detector PTR 22 kg of liquid xenon 15 kg active volume 20 cm diameter 15 cm drift PMTs Grids Teflon Can (Active Volume) 12 kv cathode E drift = 0.73 kv/cm E gas = ~ 9 kv/cm (S2) Liquid Xe maintained at T=180 K and P=2.2 atm. Cooling: Pulse Tube Refrigerator (PTR), 90W, coupled via cold finger (LN 2 for emergency) J. Angle (UFL) De Viveiros - Brown University April 2009 v07 <18>

19 The XENON10 PMTs 89 Hamamatsu R8520-AL PMTs (1 square) 48 Top + 41 Bottom Array Quantum efficiency > 175 nm ~2 x 10 6 gain x10 amp. De Viveiros - Brown University April 2009 v07 <19>

20 XENON10-88 channel DAQ / Trigger System Designed and constructed by LdV 100MHz Sampling, 14-bit Resolution 160 µs event length x 88 channels Sustained ~20 Hz, 45 MByte/sec (WIMP search trigger ~2.6 Hz, 93% live time) Multi-Event Mode and Dual Memory Bank for Dead-Time Reduction On-line event compression baseline suppression S1 trigger, S2 Trigger and S1+S2 trigger schemes tested S2 trigger used for all XENON10 results Struck VME ADC 100MHz, 14-bit De Viveiros - Brown University April 2009 v07 <20>

21 XENON10 Shield Construction - LNGS 40 Tonne Pb / 3.5 Tonne Poly 20 cm HDPE reduction in neutron flux 20 cm Pb 10-5 reduction in γ flux 2410 mm 20 cm HDPE 20 cm Pb 3500 mm Shield design and commissioning by the Brown group XENON10 De Viveiros - Brown University April 2009 v07 <21>

22 DRU XENON10 MC ER Backgrounds Depth [cm] log 10 DRU Gamma Background rates for XENON10 from Monte Carlo models Use of xyz cuts (Single Scatters in 5.4 kg Fiducial Volume) instead of LXe Outer Veto Main contribution: Stainless Steel Cryostat PMT radioactivities ( 238 U / 232 Th / 40 K / 60 Co) obtained from screening 70% of PMTs 10 Geant4 MC Geant4 model Steel Cryostat 137 Cs from Cryostat walls 85 Kr XENON10 data PMTs Pb shield Cs from PMTs Energy [kevee] radius [cm] De Viveiros - Brown University April 2009 v07 <22> -1

23 XENON10 MC NR Backgrounds Main Neutron Backgrounds PMT (α,n) / Fission subdominant (α,n) / Fission Neutrons from Rock Muon Induced Neutrons from Pb Shielding Monte Carlo event rates for neutrons are x1/3 below XENON10 background goal. 0.3 Nuclear Recoil events expected in XENON10 WIMP search run (59 days, 5.4 kg fiducial mass) Low Energy Neutrons are moderated by 20cm poly inside Pb shield Active Muon Shield Not Required (α,n) Neutron Flux: ~2 x 10-6 n s-1 cm -2 Neutron Yield in Pb: 4 x 10-3 n/(μ g cm -2 ) De Viveiros - Brown University April 2009 v07 <23>

24 The XENON10 Signal Primary Scintillation (S1) created by interaction in Lxe Std Pattern - spread evenly 20/80 top/bottom Secondary Scintillation (S2) following Ionization: e- are extracted and accelerated in Xe gas S2 signal Localized in XY - event position reconstructed from S2 Hit Pattern (σ XY 1 mm) Z-position proportional to drift time S2_time S1_time (σ Z 0.3 mm) Maximum Drift Length = 15 cm / 80 usec Single Scatters, Fiducial Volume (5.4 kg) Cuts Incident Particle S1 e- e- e- S1 e- e- γ 4.5 kevee E=1kV/cm S2 e- e- e- e- S2 Very clean signal! De Viveiros - Brown University April 2009 v07 <24>

25 Δlog 10 (S2/S1) Δlog 10 (S2/S1) Discrimination Gaussian Background Neutrons g Gamma Calibration ( 137 Cs) Δ log 10 ( S2 / S1 ) 2-12 kevee 99.6% 0.73 kv / cm Neutron Calibration (AmBe) Discrimination improves at low energies! S1 [kevee] (2.2 p.e. / kevee) S1 [kevee] (2.2 p.e. / kevee) De Viveiros - Brown University April 2009 v07 <25>

26 Scintillation Yield for Nuclear Recoils (1) Nuclear recoil light yield L eff = sets the energy scale for nuclear recoils Precise L eff necessary for high confidence on our threshold Neutron calibration data vs. MC determines L eff Good agreement (within 10%) between data and MC with L eff = 0.19 L eff = 0.19 ~ 1 p.e. / kevr E nr E-field quenching factors for ER and NR 1 Se 1 [ kevr] S1[p. e.] L S L eff Light Yield for 122keVee γ n y trigger roll-off MC (L eff = 0.19) Neutron Source Xe Conventional Neutron Calibration Xe Data Neutron Source (AmBe) Pb (5cm) XENON10 (LEFT) The spectrum of single scatter nuclear recoils from exposure to an AmBe neutron source (black line, with errors), and the spectrum (red line) from a detailed Monte Carlo of the experiment, obtained from the best-fit L eff curve shown at right (red line). The result of assuming a constant Leff=0.19 is also shown in blue. (RIGHT) Schematic of the setup for conventional neutron calibrations, and for the XENON10 neutron calibration De Viveiros - Brown University April 2009 v07 <26>

27 Scintillation Yield for Nuclear Recoils (2) Nuclear recoil light yield L eff = sets the energy scale for nuclear recoils Precise L eff necessary for high confidence on our threshold Neutron calibration data vs. MC determines L eff New best fit L eff curve from maximum likelihood comparison between Monte Carlo and AmBe neutron calibration data Results rule out sharp drop in L eff at low energy E nr E-field quenching factors for ER and NR 1 Se 1 [ kevr] S1[p. e.] L S L eff Light Yield for 122keVee γ n y MC (L eff = 0.19) L eff = 0.19 trigger roll-off MC (best fit L eff ) Data best fit L eff (LEFT) The spectrum of single scatter nuclear recoils from exposure to an AmBe neutron source (black line, with errors), and the spectrum (red line) from a detailed Monte Carlo of the experiment, obtained from the best-fit L eff curve shown at right (red line). The result of assuming a constant Leff=0.19 is also shown in blue. (RIGHT) The best fit L eff curves obtained from a maximum likelihood comparison (red). Also shown are data from [Aprile 2005] (triangles) and [Chepel 2006] (squares). De Viveiros - Brown University April 2009 v07 <27>

28 log( S2 / S1 ) XENON10 WIMP Search Data XENON10 Blind Analysis 58.6 days WIMP Box defined at ~50% acceptance of Nuclear Recoils (blue lines): [Centroid -3σ] 2-12keVee (2.2phe/keVee scale) 23 Events in the Nuclear Recoil Acceptance Window 13 events are removed from box by Gamma-X Cuts (+) 10 events in the box after all primary cuts (o) 5 of these are not consistent with Gaussian distribution of ER Background log ( S2 / S1) vs S1 Straightened Y Scale ER Band Centroid => 2.5 ER ( WIMPS? ) Non-Gaussian Background Leakage Events Gamma-X cuts NR S1 De Viveiros - Brown University April 2009 v07 <28>

29 Fake WIMPs - Gamma-X Events S1 signals from multiple scatters are indistinguishable too fast Scatters in the Reverse Field Region produce S1 light but very little S2 signal no information for scatters below cathode Fake WIMPs can occur for Multiple Scatter events with 1 scatter in the Sensitive Volume, 1 scatter in the Reverse Field Region Gamma-X : unknown component for scatters in Reverse Fiducial Region discrimination not possible Multiple Scatter Event Gamma-X Multiple Scatter Event Incoming gamma Anode Incoming gamma Anode Sensitive Volume (15 cm) S2 e- e- e- e- e- S1 S2 e- e- e- e- e- E gas Grid E drift Electron Drift ~2 mm/µs S2 e- e- e- e- e- S1 E gas Grid E drift Electron Drift ~2 mm/µs S1 Cathode Reverse Field Region (1.2 cm) e- S1 e- Cathode S1 S2 S2 S1 S2 no S2! De Viveiros - Brown University April 2009 v07 <29>

30 Gamma-X Event Rate Geant4 MC of ER background + Gamma-X events Ratio of Gamma-X events to Electron Recoils ~ 1 / 1000 at 10 kev ee Increases with energy: ~ 1 / 100 at 50 kev ee Multiple scatters boost S1 signal few Gamma-X multiple scatters at low energies Source: PMTs + Cryostat 2.2 phe / kevee ~ 1 phe / kevr De Viveiros - Brown University April 2009 v07 <30>

31 Identification of Gamma-X events - S1 Hit Pattern Internal Reflection: Asymmetry of the S1 light 20% Top / 80% Bottom Localization of S1 signal for large Z (bottom of detector) The hit pattern for events at the bottom of the detector tend to be more localized than events in the bulk, which have a more spread out hit pattern Scatters very close to bottom PMT array (<1cm) tend to deposit most of their light in a single, or a couple, of PMTs. S1 S1 X Y Z still above cathode! Event keVee γ Z = 8cm Event keVee γ Z=13cm De Viveiros - Brown University April 2009 v07 <31>

32 Identifying Anomalous Topologies Localization of Secondary Scatters (with no associated S2) point to specific anomalies Reverse Field Region Secondary Scatters Below Cathode have no Z information, but exhibit large degree of localization in single PMT and random XY distribution Gamma-X Events S1 Hit Pattern S1 = 18 p.e. 40 µs = 7.5 cm De Viveiros - Brown University April 2009 v07 <32>

33 log( S2 / S1 ) Applying the Gamma-X Cuts to XENON10 Data XENON10 Blind Analysis 58.6 days WIMP Box defined at ~50% acceptance of Nuclear Recoils (blue lines): [Centroid -3σ] 2-12keVee (2.2phe/keVee scale) 23 Events in the Nuclear Recoil Acceptance Window 13 events are removed from box by Gamma-X Cuts (+) 10 events in the box after all primary cuts (o) 5 of these are not consistent with Gaussian distribution of ER Background 1 event identified as a glitch (x) Coherent noise pick-up 4 remaining event consistent with being Gamma-X events (x) Appear preferentially at higher E Clustered at the outer bottom region of detector, where Gamma-X events are more likely Removed by more advanced Gamma-X cuts (not applied to the published blind analysis) log ( S2 / S1) vs S1 Straightened Y Scale ER Band Centroid => 2.5 S1 Non-Gaussian Background Leakage Events Gamma-X cuts Glitch Consistent with Gamma-X Gamma-X MC slide ER NR De Viveiros - Brown University April 2009 v07 <33>

34 Limit Plot Upper limits on the WIMPnucleon cross section derived with Yellin Maximal Gap Method (PRD 66, 2002) No Background Subtraction! Treats all 10 events as possible WIMP signal For a WIMP of mass 100 GeV/c cm 2 Max Gap Factor of 2.3 below best previous limit at 100 GeV/c 2 (CDMS-II ) Comparable to Zeplin-III (2008) at 100 GeV/c 2 CDMS-II XENON Zeplin-III 2008 CDMS-II 2008 Spin-independent coupling De Viveiros - Brown University April 2009 v07 <34>

35 LUX De Viveiros - Brown University April 2009 v07 <35>

36 The LUX Collaboration Brown University, Case Western Reserve University, Lawrence Livermore National Laboratory Lawrence Berkeley National Laboratory University of Maryland, Texas A&M, UC Davis University of Rochester, Yale University De Viveiros - Brown University April 2009 v07 <36>

37 Large Underground Xenon (LUX) Homestake Mine (South Dakota, US) 4300 m.w.e. µ flux = 4 µ / m 2 / day reduced x10-7 compared to sea level Thermosyphon Davis Cavern Water Shield LUX Homestake Mine Sanford Lab (SUSEL) Davis Lab at 4850L (~1.5 km deep) De Viveiros - Brown University April 2009 v07 <37>

38 Sanford Lab at Homestake Mine Sanford Lab at the Homestake Mine (South Dakota, US) LUX will be deployed in the Davis Cavern at 4850 feet level (~1.5 km deep) LUX Collaboration Meeting at Homestake, March 2009 De Viveiros - Brown University April 2009 v07 <38>

39 The LUX Detector 350kg Liquid Xe Detector (59cm height, 49cm diameter) 120 Hamamatsu R8778 PMTs (2 round): 60 on top, 60 on bottom Low-background Ti Cryostat Thermosyphon: >1 kw cooling power Teflon Can R8778 PMTs Titanium Cryostat PMTs (60 Top / 60 Bottom) De Viveiros - Brown University April 2009 v07 <39>

40 Flux Reduction LUX Water Shield Water Tank: d = 8 m, h = 6 m (300 Tonnes) 3.5m shield thickness on the sides Inverted steel pyramid (20 tons) under tank to increase shielding on top/bottom Ultra-low background facility Geant4 MC of LUX backgrounds Gamma event rate reduction: 2 x High-Energy Neutrons (> 10 MeV) flux reduction ~ 10-3 Flux Attenuation in Water (Geant4 MC) 2.75 m 3.5 m Water Shield (300T) µ-induced Neutrons (>10 MeV) 1.2 m Rock Neutrons (<10 MeV) Gammas Shield Thickness (m) Inverted Steel pyramid De Viveiros - Brown University April 2009 v07 <40>

41 Ratio Gamma-X / Gamma events LUX Background Studies - Gammas Background Model for LUX detector Monte Carlo simulations using Geant4 Gamma Backgrounds MC determined shape of detector (tall, not pancake ) All detector components are being screened for radioactivity at the SOLO and LNGS counting facilities External sources contribution < 10-4 than PMT contribution 85 Kr reduced to <2ppt ( < x1/2 of PMT background) Gammas from PMTs 390 udru r Dominated by PMTs: 390 µdru ee (PMT radioactivity for 238 U / 232 Th / 40 K / 60 Co measured at SOLO) 1 event in 45,000 kg-days at 99.4% discrimination 1e-1 Gamma-X Events optimal fiducial mass = 100kg Gamma-X backgrounds: Ratio: 500:1 Gamma to Gamma-X events in 0-30keVee Equivalent to 99.8% discrimination > 99.4% discrimination goal for LUX Rate further reduced by applying Hit Pattern cuts (same as developed for XENON10) 1e-2 1e-3 All events Fiducial volume events (100kg) Source: PMT gammas De Viveiros - Brown University April 2009 v07 <41>

42 LUX Background Studies - Neutrons Background Model for LUX detector Monte Carlo simulations using Geant4 Neutron Backgrounds External: High Energy µ-induced neutrons from rock and water tank < 200 ndru r (before Muon veto) Neutrons from PMTs All Events Internal: (α,n) neutrons from PMTs ~ 500 ndru r (conservative estimate for 5 N / year / PMT) 238 U / 232 Th alpha decays Muon veto can also veto neutron events, which are likely to scatter once in the detector and then in the water ~66% capture efficiency Achieve 1 NR event in 1,000 days with fiducial volume of 100 kg (5-25keVr) 500 ndru r Single Scatters Veto neutron capture in water (x3 reduction) optimal fiducial mass = 100kg De Viveiros - Brown University April 2009 v07 <42>

43 LUX Projected Goal ZEPLIN-III (2008) CDMS-II LUX XENON10 ZEPLIN-III (after PMT updates) SuperCDMS SNOLab (2013) XMASS (2009) LUX x20 bigger x100 better sensitivity 100 GeV WIMP 7 x cm 2 (XENON10: 8.8 x cm 2 for 100 GeV WIMP) Spin-independent coupling De Viveiros - Brown University April 2009 v07 <43>

44 Simulated signal in LUX 300 days acquisition, 100 kg fiducial mass ER Background ~390 µdru γ background ER NR band mean L eff = 0.19 Using same ER and NR bands as XENON10 NR band -3σ NR De Viveiros - Brown University April 2009 v07 <44>

45 Simulated WIMP signal in LUX 300 days acquisition, 100 kg fiducial mass ER Background ~390 µdru What will WIMPs look like in LUX? Example: m WIMP = 100 GeV/c 2 and σ WIMP = 2.1x10-45 cm 2 (3x the estimated LUX sensitivity) ER background ER NR band mean L eff = 0.19 WIMPs NR band -3σ NR Using same ER and NR bands as XENON10 De Viveiros - Brown University April 2009 v07 <45>

46 LUX PMT initial run of the LUX detector Detector is filled with 50kg of Liquid Xe + ~260kg Aluminum can (to be replaced with 350kg of Liquid Xe in LUX 1.0 run) 2 active Xe region Currently under operation at Case Western University (OH), and is being used to test all LUX subsystems LUX 0.1 PMT Assembly LUX 1.0 PMT Assembly De Viveiros - Brown University April 2009 v07 <46>

47 LUX 0.1 Installation Detector built and assembled at Case Spring kg of Al (to be replaced by 350kg of LXe) LUX 0.1 Steel Cryostat (to be replaced with Ti) De Viveiros - Brown University April 2009 v07 <47>

48 Detector built and operational at Case Detector filled with 50kg of Liquid Xe S1 and S2 pulses observed Subsystems tested and deployed at Case: Thermosyphon Cooling System Rapid (high power: >1kW) cooling system Recirculation System It will permit 50 slpm of LXe High Voltage Feedthroughs Slow Control and Safety Systems Data Acquisition System Pulse-Only Digitization mode successfully tested Custom built amplifiers and trigger system Digital Trigger with S1/S2 recognition, based on DDC-8 acquisition boards LUX Milestones De Viveiros - Brown University April 2009 v07 <48>

49 LUX 0.1 Brown at Case Brown presence at Case since March 2008 Assembly and deployment of detector Operations running the detector Development of Safety Protocols and Testing of Safety Systems Brown responsible for electronics subsystems and analysis software in LUX0.1 Data Analysis Software DAQ System Pulse Only Detection (POD) mode LED Calibration System PMTs De Viveiros - Brown University April 2009 v07 <49>

50 Larger Detectors Monte Carlo of larger detector masses Evolution of fiducial volume: more mass more self-shielding Larger fraction of low-background volume available detector rate / mass = constant LUX 350 kg 20T = 66% fiducial volume 3T 10T 20T LUX (350 kg) = 33% fiducial volume De Viveiros - Brown University April 2009 v07 <50>

51 LZ20 LUX-Zeplin Collaboration: 20 Tonnes liquid Xe detector Estimated Schedule for Construction and Operation: 2012 and 2015 CDMS-II XENON10 ZEPLIN-III (2008) ZEPLIN-III LUX LZ3 LZ20 LZ20 Baseline Design De Viveiros - Brown University April 2009 v07 <51>

52 Conclusions XENON10 Liquid Xe detectors work, and well Has delivered very competitive results Gamma-X backgrounds, although <1% effect, could become a problem if not accounted for in larger detector design LUX builds on established Xe technology Self-shielding = efficient background reduction Dominant background is from PMTs Screening show lower radioactivity (x1/3 ) than original estimates BG model predicts 1 ER and 1 NR event in energy window, for 100 kg and 1000 days Expect sensitivity to (100 GeV) WIMP dark matter of 7x10-46 cm 2 x100 times below current limits First stage of LUX detector is already running LUX deployment at the Homestake Mine - Summer 2009 LUX Underground operation by the end of 2009 De Viveiros - Brown University April 2009 v07 <52>

53 The End Thank You De Viveiros

54 De Viveiros Extra Slides

55 XENON10 Material Screening Dedicated Low Background Facility at Soudan: SOLO (operated by Brown) High Purity Ge detector: Diode-M 0.6 kg (Brown) and Gator 2 kg (UFL) Also use LNGS screening Facility (Laubenstein, LNGS) Screening of PMTs, electronic components, construction materials Currently screening materials for the LUX experiment 10 Pb X-rays Diode-M Background Diode-M Diode-M 1 DRU 40 K 1460 kev DRU = 1 event / kev / kg / day Energy [kev] De Viveiros - Brown University April 2009 v07 <55>

56 XENON10 Trigger Threshold S2 Trigger: Σ(34 top-center PMTs) Integrate with τ = 1 μs Threshold discriminator S2 trigger efficiency >99% at 4.5 kevr (= 2 kevee) Typical S2: ER (2 kevee): 2800 phe (~100 e-) NR(4.5 kevr): 1100 phe (~40 e-) Smallest NR S2 at 4.5keVr threshold: 300 phe (~12 e-) S2 Histogram (AmBe Data) Analysis Software Threshold (S1 n 2) De Viveiros - Brown University April 2009 v07 <56>

57 Scintillation XENON10 primary scintillation (S1) light yield in terms of PMT photoelectrons per kevee (Nphe/keVee) De Viveiros - Brown University April 2009 v07 <57>

58 Scintillation Yield for Nuclear Recoils (3) Independent Verification through the Ionization Yield, calculated with the Multiplicity Method Ratio of 1 / 2 / 3 scatters dependent on the threshold for individual scatters Compare Data (# of p.e.) to MC (kevr) 24±7 p.e. / e - E nr E-field quenching factors for ER and NR 1 Se 1 [ kevr] S1[p. e.] L S L eff Light Yield for 122keVee γ n y De Viveiros - Brown University April 2009 v07 <58>

59 XENON10 WIMP Search Run WIMP search run: 59 live-days, blinded (see next slide for flagged events) ~20 days of WIMP search data unblinded to test and optimize calibrations and cuts not blind background high stats γ-calibration blinded WIMP search neutron calibration De Viveiros - Brown University April 2009 v07 <59>

60 LUX DAQ Design and Upgrade 122 channels based on 14 bit VME Struck ADCs - 10 ns/sample, 700 µs event length Pulse-Only Digitization (POD) mode Software Developed by Brown and Struck to increase data throughput Struck ADC Boards (same as XENON10) => Firmware Upgrade Data is acquired only when below a given threshold Decreases the amount of data that needs to be transferred through the VME bus Saves Hard Disk Space Maximum Acquisition Rate for Classic DAQ: 5Hz 60cm Drift Length requires ~ 17 MB per event (122PMTs) Maximum Acquisition Rate for P.O.D. DAQ: 1300 Hz (P.O.D. rate allows for 500Hz spurious pulses per channel) mv ch 1 ch 2 ch 3 ch 4 S1 200 p.e. S p.e. P.O.D. Mode Baseline not acquired LUX0.1 Sample Event µs De Viveiros - Brown University April 2009 v07 <60>