Direct Dark Matter Search Experiments

Transcription

1 Direct Dark Matter Search Experiments Kamioka Observatory, ICRR, University of Tokyo 2011/12/18 Particle Cosmology at Nagoya University

2 Outline Introduction Detection of WIMP Review of Direct Dark Matter Search Experiments XMASS Experiment at Kamioka Summary

3 Variety Observation Bullet clusters CMB NASA/WMAP SDSS D. Clowe et al Rotation curve dark matter M. Blanton and the SDSS Those result encourage the direct detection of dark matter

4 Weakly Interacting Massive Particle Dark Matter is required to be Neutral can not see... Non-baryonic weakly interacting Cold (non-relativistic) large scale structure New Particle? neutralino, Kaluza-Klein particle, axion gravitino... SUSY One of the favored scenario:! astrophysics particle physics dark matter distribution Springel et al The lightest SUSY particle is stable and likely becomes a dark matter candidate! Linear combination of SUSY particles χ 1 0 = α 1! B + α 2!W + α 3! Hu 0 + α 4! Hd 0

5 Sensitivity and SUSY Parameter SI Log[σ (pb)] p cross section Roszkowski, Ruiz & Trotta (2007) CMSSM, µ > 0 ZEPLIN I EDELWEISS I CDMS II m χ (TeV) XENON100 68% 1 event/kg/yr 1 event/100kg/yr CMSSM in 2007! hep-ph v1! Roszkowski et al. Focus Point region near future CDMS-II, XENON100, XMASS, COUPP, CRESST-II, EDELWEISS-II, ZEPLIN-III,... coannihilation region Future experiments

6 Cold Thermal Relics Goal is to detect the undiscovered particles in our galaxy. H:expansion rate Γ: interaction rate when T>mx χχ ff (f:sm particle) T<mx χχ ff H>Γ freeze out Production=Annihilation (T>mχ) Production suppressed (T<mχ) Freeze out (H>Γ) density of WIMP E.W. Kolb and M.S. Turner.

7 Approaches to Dark Matter Detection q q q q q - -q direct indirect! SK, PAMELA... Colliders! LHC, ILC...

8 WIMP in the Galactic Halo halo disk solar WIMP Deposit energy! in the detector on the earth bulge We need...! - local density and velocity of dark matter! - characteristic of target nuclei (Xe, Na, Ge, I...)

9 The Milky Way We are here! halo disk bulge 8kpc Sun KLYPIN et al. APJ 2002 NASA/JPL-Caltech/R. Hurt (SSC/Caltech) Normally, we take ρdm ~0.3 GeV/cm 3, vearth ~ 230km/sec

10 WIMPs around us ρ dm = 0.3 GeV/cm 3 A few WIMPs in 1 Liter of volume around us if we assume 100 GeV mass through your hand every second. only < 1 event/day will interact with your body. How can we detect them?

11 Direct Detection Principle WIMPs elastically scatter off nuclei in targets, producing nuclear recoils. Dark Matter (WIMP) Deposit Energy E recoil ~ <100 kev χ + N χ + N Log(Rate) energy

12 Deposit Energy E R = r r = (1 cos ) 2 E W (center of mass) 4M w M N (M w + M N ) 2 WIMP θ θ nuclei For example, assuming Mw = 100 GeV/c2, M T = 100 GeV/c2, r = 1 WIMP velocity: v~ 0.75 X 10-3 = 220 km/sec E R = 1 2 M W 2 c 2 = 1/2 x 100 x GeV/c 2 (0.75 x 10-3 ) c 2 = 30 kev Log(Rate) energy

13 Differential Rate dr = R 0F 2 (E R ) de R E 0 r k 0 k 1 2πv 0 vmax v min 1 v f(v, v E)d 3 v motion dynamics Spin independent Spin dependent σ 0 = A 2 µ2 T µ 2 p σ χ p σ 0 = (λ2 N,Z J(J+1))Nuclear (λ 2 p,z J(J+1))proton µ 2 T µ 2 p σ χ p

14 Energy spectrum Energy threshold Event Rate For 100 GeV! WIMPs Xe Ar I Na Ge O Recoil Energy (kev)

15 Energy spectrum Energy threshold Event Rate Xe I For 10 GeV WIMPs Na Ge Ar O Recoil Energy (kev)

16 Direct Detection

17 Underground lab At the surface, a few cosmic rays go though your hand every second.! It reduced to 1/10 5 if you are in Kamioka mine.

18 Direct Dark Matter Search in the World SNOLAB DEAP-CLEAN Picasso COUPP Boulby NaIAD ZEPLIN I/II/III DRIFT I/II Soudan CDMS II CoGENT Frejus EDELWEISS I/II CanFranc IGEX ROSEBUD ANAIS ArDM Gran Sasso DAMA/LIBRA CRESST I/II XENON WARP NIT(?) JINPING PANDA-X CDEX YangYang KIMS Oto PICOLON Kamioka XMASS NEWAGE

19 Techniques for Detector Various Targets: Ge, Xe, Ar, Ne and so on.! Two Signals are used to particle identification to distinguish btw Nuclear Recoil and gamma or beta. CDMS CRESST EDELWEISS Phonon Ionization Erecoil Light XMASS,CLEAN ZEPLIN, XENON! WARP, LUX,ArDM

20 Current status M2 M1

21 XENON at Gran Sasso, Italy XENON100 Goal (compared to XENON10): increase target 10 reduce gamma background 100 material selection & screening detector design Quick Facts: 161 kg LXe TPC (mass: 10 Xe10 ) 62 kg in target volume active LXe veto ( 4 cm) 242 PMTs (Hamamatsu R8520) passive shield (Pb, Poly, Cu, H2O, N2 purge) arxiv: M. Schumann (U Zürich) XENON E. Aprile, ICRR, Univ of Tokyo

22 Leff proportional Gas Xe Liquid Xe direct Gamma γ Xe Event Discrimination: Electron or Nuclear Recoil PMT Array e- E g E d WIMP WIMP Gamma e- (S2/S1) wimp << (S2/S1) gamma bottom PMT array 0.35 Arneodo 2000 Bernabei Akimov 2002 Aprile 2005 Chepel Aprile 2009 Manzur Plante 2011 S1 S1 drift time drift time S2 S2 nuclea r Gamma or Electron gamma/e rejection -> ~1/ Energy [kevnr] WIMP or Neutron Nuclear Recoil Equivalent Ener Nuclear recoil equivalent energy E nr is obtained from the S1 signal E nr = S1 1 L y,er L eff (E nr ) S er S nr L y,er, light yield of electron recoils electro from 122 kev γ rays n S er, S nr, scintillation light quenching due to drift field Relative scintillation efficiency L eff L eff (E nr )= L y,nr (E nr ) L y,er (E ee = 122 kev) L eff The uncertainty in L eff at low energies is the, ICRR, Univ of Tokyo

23 Result of XENON100 S1 [PE] 4 ground model of the Profile Likelihood analysis employs same data and background assumptions. 0.4 After unblinding the pre-defined WIMP search region, 0.2 a population of events was observed that passed the S1 0 coincidence requirement only because of correlated electronic noise thattaken is picked fromhalf an external 100 khz -data inupfirst of source, as verified by inspection of the digitized PMT sig life days -0.4 nals. These events are mostly found below the S1 analysis -48 kg3 events fiducial out leaking of 62 into kg -0.6 threshold, with fromvolume this population the WIMP search region close to the 4 PE lower bound. -data blinded in ROI -0.8 This population can be identified and rejected with a cut -1 on the S1 PMT coincidence level, that takes into account -1.2 correlated pick-up noise, and by cutting on the width N100: Limit Energy [kevnr] of the S1 candidate. These post-unblinding cuts have a 10 Recoil Energy [kev] 50 combined acceptance of 99.75% for NRs while removing Expected Background -39FIG. 3: Observed event distribution using the discriminathe entire population of noise events. 10 subtracting the ER tion parameter log10 (S2/S1), flattened by Leakage: 0.48criXENON100 (2011) With Gaussian these additional cuts, 3 events 1.14 pass all±quality DAMA/Na band mean, as a function of NR equivalent energy (kevcl) nr ). observed limit (90% teria for Anomalous single-scatter NRs and fall in the.56± WIMP search -40 Leakage: All quality cuts, defined before and afterexpected unblinding, arerun:used. limit of this CoGeNT region, see Fig. 3. This observation remains unchanged 1 σ expected with an Gray points indicate the NR distribution as±measured DAMA/I Neutron Background: 0.11 ±of ± 2 σ expected for moderate variations in the definition of any the data CDMS (2011) AmBe neutron source. The WIMP search region (dashed) quality cuts. These events were observed on ± January 23, is defined by the energy window kevnr (4 30 PE) Total: events CDMS (2010) and the lower bound from the S2 median of the software February 12, and June 3, at 30.2 kevnr, 34.6 kevnr, 300 PE 13 PE). A second analysis adxenon10 (S2 (4 only, 2011) 10-42threshold S2 > and 12.1 kevnr, respectively. The event distribution ditionally uses the 99.75% rejection line from above and the XENON100 (2010) EDELWEISS (2011) throughout the TPC is shown in Fig. 4. Given the backobserve 3 events 3 contour of the NR distribution from below (13 30 PE). ground expectation (1.83±or0.6) events, the observation 10-43Three events fall into this WIMP search region (red circles), - likelihoodoffor more events is 28% of 3 events does not constitute evidence for dark matter, with (1.8 ± 0.6) events expected from background. - Profile Likelihood analysis also does not yield as the chance probability of the corresponding Poisson significant ->events calculate process to result in 3signal or more is 28%. limit Trotta et al. Therefore, a limit on the spin-independent WIMP(1.14 ± 0.48) events in the WIMPet al.search region, where Buchmueller 10-45the error is dominated by the statistical uncertainty in nucleon arxiv: elastic scattering cross-section is calculated us ing the Profile Likelihood method [15]. WIMPs are asmass [GeV/c ] the definition discrimination line. Non-Gaussian of the WIMP, ICRR, Univ of Tokyo sumed to be distributed in an isothermal halo with v0 = gamma rays WIMP-Nucleon Cross Section [cm2] 10 log (S2/S1)-ER mean 5 nuclear recoil 30 35

24 DAMA/LIBRA Gran Sasso in Italy! DAMA(~100 kg) + LIBRA (~250 kg) of NaI! Annual Modulation (DAMA 7 yrs + LIBRA 4yrs) 30 km/s December ~ 232 km/s km/s June v sun ~ 232 km/s (Sun velocity in the halo) v orb = 30 km/s (Earth velocity around the Sun)! γ = π/3! ω = 2π/T T = 1 year! t 0 = 2 nd June (when v is maximum) v (t) = v sun + v orb cosγcos[ω(t-t 0 )] dr S k [!(t)] = $ de de R " S 0,k + S cos[% (t & t )] m,k 0 R #E k

25 DAMA/LIBRA Result Dark Matter investigation by model-independent annual modulation signature DAMA/NaI (7 years) + DAMA/LIBRA (6 years). Total exposure: 1.17 ton yr (the largest exposure ever collected in this field) Experimental single-hit residuals rate vs time in 2-6 kev EPJC 56(2008)333, EPJC 67(2010)39 Power spectrum Principal mode d -1 1 y kev Modulation 6-14 kev Acos[ω(t-t 0 )]! continuous lines: t 0 = d, T = 1.00 y A=(0.0114±0.0013) cpd/kg/kev χ 2 /dof = 64.7/ σ C.L.!!! Absence of modulation? No χ 2 /dof=140/80 P(A=0) = fit with all the parameters free: A = ( ± ) cpd/kg/kev t 0 = (146±7) d T = (0.999±0.002) y 8.9σ C.L. No systematics or side reaction able to account for the measured modulation amplitude and to satisfy all the peculiarities of the signature Comparison between single hit residual rate (red points) and multiple hit residual rate (green points) for (DAMA/LIBRA 1-6); Clear modulation in the single hit events A=(0.0091±0.0014) cpd/kg/kev; No modulation in the residual rate of the multiple hit events A=-(0.0006±0.0004) cpd/kg/kev No Modulation Multiple hits events = Dark Matter particle switched off The data favor the presence of a modulated behaviour with all the proper features for DM particles in the galactic halo at about 9σ C.L. A.Incicchi 2-6 kev This result offers an additional strong support for the presence of DM particles in the galactic halo further excluding any side effect either from hardware or from software procedures or from background

26 What is the background? Energy distribution of the modulation amplitudes [ ( t )] ( t) = S0 + Sm cos ω t eret=2π/ω=1 yr and t 0 = day 0 DAMA/NaI (7 years) + DAMA/LIBRA (6 years) total exposure: kg day 1.17 ton yr ΔE = 0.5 kev bins Event rate, events/kg/day/kev 10 1 K-40 U-238 Th signal subtraction A clear modulation is present in the (2-6) kev energy interval, while S m values compatible with zero are present just above The S m values in the (6 20) kev energy interval have random fluctuations around zero with χ 2 equal to 27.5 for 28 degrees of freedom A.Incicchi Energy, kev J. Of Phys. Conf. Ser. 203(2009) DAMA recently upgrade to Higher QE PMTs.

27 CoGENT at Soudan (Coherent Germanium Neutrino Telescope) CoGeNT Dark Matter Experiment (in brief ) 440-gram high purity germanium ionization spectrometer ~0.5 kev energy threshold In low-background shield at Soudan Underground Lab Operated for dark matter direct-detection search Installed: Aug Data start: Dec Offline: Mar (Soudan Lab mine fire) Data Restart: July , ICRR, Univ of Tokyo

28 CoGENT (surface event cut) Surface events -> external gamma, poor charge collection but... bulk surface Charge Collection time modelled (small 100 ns correction) C.E. Aalseth et al. PRL 106, MAJORANA BEGe (ORNL simulation), ICRR, Univ of Tokyo

29 CoGENT result - Energy threshold 0.5 kev! kg x 442 days! - modulation hypothesis 2.8 sigma! ±3.8% amplitude! - 347±29 days period! - minimum in Oct 16±12 d solid: best-fit! dash: 7GeV/c 2 with Maxwellian to increase statistic: 440g -> C4 (1 kg x 4 modules), ICRR, Univ of Tokyo

30 CRESST CRESST Cryogenic Detectors Target crystals operated as cryogenic calorimeters (~10mK) energy deposition in the crystal: mainly phonons temperature rise detected with W-thermometers measurement of deposited energy (sub kev resolution at low energy) small fraction into scintillation light Separate cryogenic light detector to detect the light signal Detector module: 300gx8 modules! for analysis Simultaneous measurement of: deposited energy E in the crystal (independent of the type of particle) scintillation light L (characteristic of the type of particle) phonon detector light detector reflecting scintillating housing W thermometer Light absorber CaWO 4 target W thermometer Results from 730 kg days of the CRESST-II Dark Matter Search Federica Petricca on behalf of the CRESST collaboration, ICRR, Univ of Tokyo 3

31 CRESST Estimate the contribution of the backgrounds and investigate a possible excess by using likelihood. two maximum in Likelihood Ch51 M1 M2 e / events events neutron events Pb recoils signal events m WN GeV pb WIMP acceptance region 67 events accepted Fig. 3. Illustration of background events due to surface con-, ICRR, Univ of Tokyo

32 CRESST result M2 M1 New physics run in preparation with several improvements aimed at background reduction: Modification of the clamps holding the crystals to reduce and Pb-recoils backgrounds Installation of an additional internal neutron shielding to complement the present one Expected to start this year Results from 730 kg days of the CRESST-II Dark Matter Search Federica Petricca on behalf of the CRESST collaboration Masaki 18 Yamashita, ICRR, Univ of Tokyo

33 Activity In Japan NEWAGE (Kobe Kamioka CF4(TPC) NIT (Nagoya R&D Emulsion PICOLON(Tokushima NaI CANDLES(Osaka CaF XMASS Kamioka LXe Kr km/s (600keV) 600 nm Emulsion

34 XMASS Experiment at Kamioka

35 Location Kamioka Tokyo

36 Kamioka Observatory By courtesy of Dr. Miyoki 1000m under a mountain = 2700m water equiv. 360m above the sea Horizontal access Super- K for ν physics and other experiments in deep underground KamLAND (Tohoku U.)

37 XMASS Experiment Multi purpose low-background experiment with LXe. Xenon MASSive detector for Solar neutrino (pp/ 7 Be) Xenon neutrino MASS detector (double beta decay) Xenon detector for Weakly Interacting MASSive Particles (DM) Solar Neutrino Dark Matter Double Beta Decay, ICRR, Univ of Tokyo

38 Phase 100kg Prototype (FV:30kg ~30cm) 800kg Detector (FV:100kg 80cm) 20ton Detector (FV:10ton ~2.5m) R&D Completed Dark Matter 2007: Project was funded. Solar Neutrino Dark Matter

39 Why Liquid Xenon? High Atomic mass Xe (A~131) good for SI case (cross section A 2 ) Odd Isotope (Nat. abun: 48%, 129,131) with large SD enhancement factors High atomic number (Z=54) and density (ρ=3g/cc): compact, flexible and large mass detector. High photon yield (~ UV photons/mev at zero field) Easy to purify for both electro-negative and radioactive purity by recirculating Xe with getter for electro-negative Distillation for Kr removal

40 Self Shielding by LXe Blue : γ tracking Pink : whole liquid xenon Deep pink : fiducial volume γ into LXe sphere MC simulation Counts/keV/day/kg Goal is to achieve 10-4 kev

41 Design of 800 kg Detector pentakisdodecahedron Hexagonal PMT Hamamatsu R10789 QE 28-39% l 60 triangle in total! l about 10PMT/triangle 60 l Total: 642 PMTs l Photo coverage: 62%

42 Assembly of PMT holder and installacon of PMTs OFHC: brought into the mine <1month after electrorefining. to avoid activation by cosmic rays.

43 Joining two halves

44 In the Water Shield 20 inch PMT x 70 gas/liquid line

45 Radioactivity e.g. 40 K case 4000 Bq/human 15 Bq/Banana 0.01 Bq/PMT! 6.42 Bq/642 PMTs

46 XMASS PMT HISTORY PMT YEAR Model Prototype R8778 R10789 Material:Body glass Kovar Kovar QE 25% 25% 27-39% RI: U [mbq/pmt] 50 18± / Th [mbq/pmt] ± / ±20 < < ± / The goal is to achive 1/10 Reduccon of radioaccvity level of R8778 (except 60 Co). n<1.2x10-5 Based on V.Tomasello, et al., NIMA595(2008)431

47 Demonstracon of the detector performance Calibracon system Introduccon of radioaccve sources into the detector. <1mm accuracy along the Z axis. Thin wire source for some low energy γ rays to avoid shadowing effect. 57 Co, 241 Am, 109 Cd, 55 Fe, 137 Cs.. Stepping Motor Linear Motion Feedthrough ~5m Gate valve Source rod with a dummy source 0.15mmφ for 57 Co source 4mmφ Top photo tube

48 Photoelectron dist. at the center of the detector Total photo electron distribucon real data simulation Reconstructed energy distribucon real data simulation 59.3keV of W 122keV ~4% rms 136keV 57 Co source at the center of the detector gives a typical response of the detector. High p.e. yield 15.1+/- 1.2p.e./keV was obtained. The photo electron yield distribucon was reproduced by a simulacon.

49 Performance of the vertex reconstruccon Real Data Simulation Reconstructed vercces for various posiconing of the source overlayid. Posicon resolucon was as expected by a simulacon: 1.4cm RMS (0cm) 1cm RMS (+- 20cm) for 122keV γ rays

50 Sensitivity for SI case 1keV, 2keV, 5keV DAMA CoGENT DAMA 10-4 dru, 100 kg fiducial (flat bg assumed) CDMSII XENON100 XMASS 800 kg 1 year 1 year exposure σ χp =10-44 cm 2 50GeV WIMP Black:signal+BG Red:BG

51 Summary XENON100:! continuing to running and preparing XENON1T! DAMA/LIBRA:! Upgrade to higher QE.! CoGENT:! 0.4 kg -> 1kg Upgrade. (C4)! CRESST:! Upgrade to reduce background from crystal holder and! will add more shield for neutron.! XMASS:! Commissioning mode. Hopefully, we say something in Mach 2012.