Current and Future of WIMP Search in DarkSide

Transcription

1 Current and Future of WIMP Search in DarkSide Masayuki Wada Princeton University on behalf of the DarkSide Collaboration March nd Rencontres de Moriond Conference 1

2 DarkSide Program Direct detection search for WIMP dark matter Based on a two-phase argon time projection chamber (TPC) Design philosophy based on having very low background levels that can be further reduced through active suppression, for background-free operation from backgrounds (both from neutrons and β/γ s) 2

3 Gran Sasso 3800 m w. e. Deep underground location at LNGS, Italy.

4 DarkSide 50 Radon-free (Rn levels < 5 mbq/m 3 ) Clean Room 1,000-tonne Water-based Cherenkov Cosmic Ray Veto 30-tonne Liquid Scintillator Neutron and γ s Veto Inner detector TPC 4

5 Two Phase Argon TPC S1 A recoil excites and ionizes the liquid argon, producing scintillation light (S1) that is detected by the photomultipliers The electrons are extracted into the gas region, where they induce electroluminescence (S2) e - S1 Drift Time S2 S2 light fraction The time between the S1 and S2 signals gives the vertical position. x-y position of events are reconstructed from fraction of S2 in each PMT. Electron drift lifetime > 5 ms, compared to max. drift time of ~ 375 μs. Electron drift speed = 0.93 mm/μs 5

6 Features of Noble Liquid Detectors Dense and easy to purify (good scalability) High scintillation & ionization (low energy threshold) Transparent to own scintillation For TPC High electron mobility and low diffusion Discrimination electron/nuclear recoils (ER/NR) via ionization/scintillation ratio Liquid Xenon Denser & Radio pure Lower energy threshold Higher sensitivity at low mass WIMP 6 Liquid Argon lower temperature (Rn purification is easier) Stronger ER discrimination Intrinsic ER BG from 39 Ar

7 Backgrounds [30-200] kevr ELECTRON NUCLEAR RECOILS RECOILS 39 Ar n n μ ~30 evt/m 2 /day ~9x 4 evt/kg/day α 39 Ar μ Radiogenic n γ γ ~6x -4 evt/kg/day ~1x 2 evt/kg/day γ ~ evt/m 2 /day 0 GeV, -45 cm 2 WIMP Rate ~ -4 evt/kg/day 7

8 Underground Ar 39 Ar Reduction Intrinsic 39 Ar radioactivity in atmospheric argon is the primary background for argon-based detectors 39 Ar activity sets the dark matter detection threshold at low energies (where pulse shape discrimination is ineffective) 39 Ar is a cosmogenic isotope, and the activity in argon from underground sources can be significantly lower compared to atmospheric argon Rate/(Bq/keV) Energy Spectra Atmospheric Argon Underground Argon Depletion factor 150 or more underground argon were deployed in Results will be later in this talk Earlier study with a small detector Energy/keV Figure 7: The energy spectra recorded in the argon detector under different condi arxiv: [physics.ins-det] Red: underground argon data at surface; purple: underground argon data at surface 8 an active cosmic ray veto; blue: underground argon data at KURF; green: atmosph

9 Pulse Shape Discrimination Electron Recoil Discrimination Electron and nuclear recoils produce different excitation densities in the argon, leading to different ratios of singlet and triplet excitation states τsinglet ~ 7 ns τtriplet ~ 1500 ns Electron Recoil PSD parameter F90: Ratio of detected light in the first 90 ns, compared to the total signal ~ Fraction of singlet states NR band Nuclear Recoil ER band AmBe calibration data 9

10 Liquid Scintillator Veto Neutron Rejection Liquid scintillator allows Coincident veto of neutron (and γ) events in the TPC In situ measurement of the neutron background rate 4 m diameter sphere containing PC + TMB scintillator Instrumented with 1 8 PMTs Neutrons can be veto via neutron captures on B or thermalization signals.

11 Neutron Capture (Veto) Neutron Rejection n + B 7 Li (15 kev) + α (1775 kev) (6.4%) 7 Li* (839 kev) + α (1471 kev) (93.7%) 7 Li + γ (478 kev) J. Instrum. 11 P03016, 2016 Event Rate/PE AmBe Source data α + γ Veto efficiency from capture signal > 99.1% (from calibrations and simulations) ~7.7% of capture on 1 H; 2.2 MeV γ lost ~8% α ~0.05% of neutrons leave no signal in LSV at all Total Veto efficiency is larger due to additional thermalization signal Energy [PE] Due to low background, only ~0.9% acceptance loss with 1 PE threshold 11

12 Thermalization Signal (Veto) Neutron Rejection Measurement of veto efficiency for neutron thermalization signals is difficult coincidence γ from AmBe. AmC source has reduced coincident γ. Analysis in progress using new AmC source data (Dec 15 - Jan 16) Late thermalization NR TPC Rare in real BG n Early thermalization AmC LSV Charge in Veto [PE] AmC Late thermalization Similar to real BG NR n AmC Time [μs] TPC LSV 12

13 External Water Tank Cosmogenics Rejection 80 PMTs within water tank (11 m diameter x m height) Acts as a muon and cosmogenic veto (~ 99% efficiency) Provides passive γ s and neutron shielding 13

14 Radon-Free Clean Rooms Surface contamination Radon daughters plate out on surfaces of the detector causing dangerous alpha-induced nuclear recoils. Final preparation, cleaning, evaporation and assembly of all inner detector parts was carried out in radon-free clean rooms. Typical radon in air ~ 30 Bq/m 3 Cleanroom radon levels < 5 mbq/m 3 14

15 DS50 Commissioning (Oct. 2013) 15

16 CALIS - CALibration Insertion System Calibrate both TPC and Neutron veto Gamma sources: 57 Co (122 kev), 133 Ba (356 kev), 137 Cs (663 kev) Neutron source: AmBe and AmC NR band matches with the points extrapolated from SCENE. NR study (crosscheck of SCENE data) Test of the MC code DATA-MC comparison: 57 Co source next to the cryostat NR band s] Events/[20 PE Monte Carlo Data ER band S1 [PE] 16

17 Underground Ar 1. Extraction at Colorado (CO2 Well) Extract a crude argon gas mixture (Ar, N2, and He) Plant at Colorado 2. Purification at Fermilab Separate Ar from He and N2 3. Arrived at LNGS Ready to fill into DS-50 UAr bottles at LNGS 17 Distillation Column at Fermilab

18 Underground Ar 1. Extraction at Colorado (CO2 Well)! t r o Extract a crude argon gas mixture (Ar, N2, and He) s r a e Y f f E x i S Plant at Colorado, 2. Purification at Fermilab d e c Separate Ar from He and N u d o r p g k 5 Distillation Column at 15 Fermilab 2 3. Arrived at LNGS Ready to fill into DS-50 UAr bottles at LNGS 17

19 UAr Results AAr vs UAr. Live-time-normalized S1 pulse integral spectra at Zero field. 39 Ar reduction factor of ~1400! AAr 39 Ar 1/1400 LN2 (150L dewar + 300W CryoCooler); Ar condenser (300W latent heat); Radon Trap Unit (charcoal filter & heat exchanger); Gas control pannel (valves, MFCs/Ms); Nominal trigger rate ~1.5 Hz, predominantly ER. Low level of 39 Ar allows extension of DarkSide program to ton-scale detector. 18

20 UAr Results No background events in nuclear recoil (WIMP) region! WIMP Expected Region Drift 200 V/cm 71 live-days after all cuts. (2616±43) kg day exposure. Single-hit interactions in the TPC, no energy deposition in the veto. 19 Phys. Rev. D , 2016

21 UAr Results [cm 2 ] σ LN2 (150L dewar + 300W CryoCooler); 43 PandaX-I (2014) PICO (2015) WARP (2007) DarkSide-50 (AAr, 2014) DarkSide-50 (UAr, 2015) DarkSide-50 (combined) XENON0 (2012) Ar condenser (300W latent heat); CDMS (2015) Radon Trap Unit (charcoal filter & heat exchanger); Gas control pannel (valves, MFCs/Ms); 46 LUX (2015) 2 LUX (2016) Phys. Rev. D , Best limit to date, with argon target, third best limit behind LUX & Xenon0 at high mass WIMP range. Phys. Rev. D 93, 0811(R) (2016) PandaX-II (2016) 20 Best limit to date with argon target! 3 M χ 2 [GeV/c ] 4

22 Current Status of DS-50 The first WIMP search with UAr data set AmC calibration 83m Kr calibration 83m Kr calibration AmBe calibration Blind Data Set Accumulated more than 1 live-year blinded dataset. Blinded procedure nearly completion. Keep running to show stability and study rare BGs. 21

23 Blind Analysis Procedure 22

24 Blind Analysis Procedure At event reconstruction level, apply a blind box in F90 vs S1 space. 22

25 Blind Analysis Procedure At event reconstruction level, apply a blind box in F90 vs S1 space. Model BG events using calibration datasets and tuned MC. 22

26 Blind Analysis Procedure At event reconstruction level, apply a blind box in F90 vs S1 space. Model BG events using calibration datasets and tuned MC. Refine cuts based on leakage BG events ( 0.1 events total). 22

27 Blind Analysis Procedure At event reconstruction level, apply a blind box in F90 vs S1 space. Model BG events using calibration datasets and tuned MC. Refine cuts based on leakage BG events ( 0.1 events total). Test the BG models in test strips. 22

28 Blind Analysis Procedure At event reconstruction level, apply a blind box in F90 vs S1 space. Model BG events using calibration datasets and tuned MC. Refine cuts based on leakage BG events ( 0.1 events total). Test the BG models in test strips. 22

29 Blind Analysis Procedure At event reconstruction level, apply a blind box in F90 vs S1 space. Model BG events using calibration datasets and tuned MC. Refine cuts based on leakage BG events ( 0.1 events total). Test the BG models in test strips. Open the final blind box. 22

30 Blind Analysis Procedure At event reconstruction level, apply a blind box in F90 vs S1 space. Model BG events using calibration datasets and tuned MC. Refine cuts based on leakage BG events ( 0.1 events total). Test the BG models in test strips. Open the final blind box. If BG models are ready, we will open the blind box by summer. 22

31 Blind Analysis Procedure At event reconstruction level, apply a blind box in F90 vs S1 space. Model BG events using calibration datasets and tuned MC. Refine cuts based on leakage BG events ( 0.1 events total). Test the BG models in test strips. Open the final blind box. If BG models are ready, we will open the blind box by summer. Count radiogenic neutrons with Veto and compared with our prediction. 22

32 Future Detectors? DS-20k 30 tonne (20 tonne fiducial) detector ARGO 300 tonne (200 tonne fiducial) detector 23

33 Requirements for DS-20k Neutron Background: Cosmogenic: Veto system Radiogenic: radiopure SiPM & ultra-clean Titanium (TPC cryostat) β/γ background: 39 Ar: Underground Argon (Urania Project) & Depleted Argon (Aria Project) γ: SiPM & ultra-clean Titanium 24

34 Further Depletion of Ar Urania (Underground Argon):Towards DarkSide-20k DarkSide: URANIA& ARIA production facili DarkSide: URANIA& ARIA projects production faci Sub-projects of the DarkSide as Depleted ArCollaboration, : theclassified URANIA and ARIA infrastructures Ar: the URANIA and ARIA projects Depleted Sub-projects of the DarkSide Collaboration, classified as ExpansionLNGS of the argon extraction The plant Depleted Ar : the URANIA and ARIA projects URANIA: LNGS infrastructures Plant expansion The plant Replacement of the Ar extraction plant in Colorado to URANIA: plant in Cortez, CO, to reach reach capacity of 0 kg/day of UAr Plant expansion Replacement of the Ar extraction plant in Colorado to URANIA Cost: 3.2M reach capacity of 0 kg/day capacity of 0 kg/day ofof UAr Volatilità relative => MIUR/INFN (2.3M ) Cost: 3.2M Progetto Premiale 2013 Volatilità relative =>=>1.007 Volatilità relative Replacement of the Ar extraction plant in Colorado Underground Argon Valori tipici >1.5 NSF + other US sources (0.9M ) MIUR/INFN Progetto Premiale 2013 (2.3M ) to reach capacity of 0 kg/day of UAr Thousands of equilibrium stages reflects in a very tall column Valori tipici >1.5 Valori tipici >1.5 commissioning and test atthousands the Neutrino Platform the Numero diaria stadiproduction teorici => ordine delle ofof equilibrium stages reflects inina migliaia Thousands equilibrium stages reflects avery verytall tallcolumn column discussion with CERN towards possibile DarkSide: URANIA& facilities discussion withus CERN towards the possibile NSF + other sources (0.9M ) Aria (UAr Purification): commissioning and test at the Neutrino Platform Sub-projects of the DarkSide Collaboration, classified asurania& Numero diaria stadi teorici Numero di stadi teorici=>=>ordine ordinedelle dellemigliaia migliaia DarkSide: production facilities LNGS infrastructures 350 m tall distillation column in the Seruci mine in HETP = cm ARIA: URANIA: R&D Column of the DarkSide Collaboration, classified as Sardinia Sub-projects for chemical and isotopic purification of UAr LNGS infrastructures Replacement of the Ar extraction plant in Colorado to 350 m tall cm diameter distillation mine=in= HETP cm HETP cm30 R&D reach capacity of 0 column kg/day of UArin the Seruci Column URANIA: R&D Column Exploits finite vapor pressure difference between 39Ar/ Sardinia for chemical and isotopic purification of UAr Cost: 3.2M H= m Replacement of the Ar extraction plant in Colorado to 350 m height 3030 cm cmdiameter diameter 40Ar (39Ar reduction ofof2013 pass at the rate reach capacity offactor 0 Premiale kg/day UAr per MIUR/INFN Progetto (2.3M ) Exploits finite vapor pressure difference between 39Ar/ ARIA Cost: H= mm 350 NSF 3.2M + other US sources (0.9M ) of 0 kg/day) H= mmheight height 40Ar (39Ar factor ofthe2013 per pass at the rate reduction discussion MIUR/INFN Premiale (2.3M ) withprogetto CERN towards possibile Usuali = m Protocollo dicommissioning Intesa between INFN and Regione test at the Neutrino Platform NSF + other and US sources (0.9M ) of 0 kg/day) Very tall distillation column in Seruci mine (Sardinia) for ARIA: Sardegna discussion with CERN towards the possibile Usuali == mm Usuali Protocollo di Intesa INFN and 350 m commissioning and test at the Neutrino Platform tall distillationbetween column in the Seruci mine in Regione chemical and isotopic purification of UAr Fuori terra Sardinia for chemical and isotopic purification of UAr Cost: 12.5M ARIA: Sardegna Exploits finite vapor pressure difference between 39Ar/ 350 m tall distillation column in the Seruci mine in INFN 40Ar (39Ar reduction factor per pass at the rate (4M ) Fuori terra Fuori terra Sardinia for chemical andofisotopic purification of UAr Cost: 12.5M Seruci Wells Seruci SeruciWells Wells Very tall column in the Seruci mine in Sardinia, Italy, for high-volume chemical isotopic purification Exploits finite and vapor pressure difference between Ar/ NSF + other US sources (1.3M ) A sezioni separate INFN (4M ) Ar: Ar reduction factor of per pass at the rate of of 0 Underground sezioni separate sezioni separate CARBOSULCIS (4.5M ) NSF Argon + other US sources (1.3M ) A A kg/day m 325 m 325 m ARIA: 39 of 0 kg/day) Exploits finite vapor pressure difference between 39Ar/ Protocollo di Intesa between INFN and Regione 40Ar (39Ar reduction factor of per pass at the rate Sardegna of 0 kg/day) Cost: 12.5M di Intesa between INFN and Regione Protocollo Sardegna INFN (4M ) Regione Autonoma Sardegna Cost: CARBOSULCIS (4.5M ) NSF 12.5M + other US sources (1.3M ) (2.7M ) INFN (4M ) (4.5M ) CARBOSULCIS Regione Autonoma Sardegna (2.7M ) Sandro De Cecco 15 October 2015 NSF +Autonoma Regione other US sources (1.3M ) Sardegna (2.7M ) 15 October CARBOSULCIS (4.5M ) Regione Autonoma Sardegna (2.7M ) Sandro De Cecco DarkSide DarkSide DarkSide DarkSide

35 Expected sensitivity [cm 2 ] σ CDMS (2015) LUX (2015) WARP (2007) PandaX-I (2014) PICO (2015) DarkSide-50 (2015) XENON0 (2012) PandaX-II (2016) DarkSide-50 (3 yr proj.) LUX (2016) DarkSide-20k (0 t yr proj.) Argo (00 t yr proj.) Coherent neutrino-nucleon scattering floor M χ 2 [GeV/c ] 4

36 Summary DarkSide-50 performed the first ever WIMP dark matter search with Underground Ar. Concentration of 39 Ar in UAr is 1400 times lower than in AAr. DarkSide-50 has the strongest WIMP limit among Ar target experiments. Efficiency of the Neutron Veto is >99.1% with only the capture signals and is expected to be more with additional thermalization signals. More than 1 live-year blinded dataset is actively analyzed. Future detectors are planned and active R&D s are underway. 27

37 THE END 28

38 DS50 Timeline Oct. 2013: LArTPC, Neutron Veto and Muon Veto commissioned. TPC filled with atmospheric argon (AAr). Up to June 2014: data taken with high 14 C content in LSV live days (1422 kg day fiducial) for the first physics result. TMB ( 14 C) was removed to reduce the 14 C rate. Oct. to Dec. 2014: Calibration of TPC w/ radioactive sources. Jan. 2015: Add radiopure TMB at 5% concentration. Mar. to Apr. 2015: Fill with UAr and re-commissioning the detector. Apr. to Aug. 2015: Accumulate data with UAr for dark matter search. Since Aug. 2015: Continue dark matter search more than 1 live-year. 29

39 The First Physics Result from DS-50 Background-free exposure of 1422 ± 67 kg day f 90 No background events in nuclear recoil (WIMP) region! WIMP Expected Region % 65% 80% 90% 39 Ar BG(electron recoils) S1 [PE] Selected only single-hit interactions in the TPC fiducial volume (36.9 kg) with no energy deposition in the veto 30 Phys. Lett. B 743 (2015) 456 0

40 The First Physics Result from DS-50 Background-free exposure of 1422 ± 67 kg day f PE ~ 38 kevr (Nuclear quenching from 200 V/cm) F90 NR 50% Acceptance Curves from WIMP SCENE 200 Region V/cm 460 PE ~ 206 kevr % 65% 80% 90% From DS50 F90 model Ar BG(electron recoils) S1 [PE] Selected only single-hit interactions in the TPC fiducial volume (36.9 kg) with no energy deposition in the veto 30 Phys. Lett. B 743 (2015) 456 0

41 TPC: Electric Recoils calibration 39 Ar (565 kevee endpoint) present in AAr 83m Kr gas deployed into detector (41.5 kevee) Entries m Kr Half-life = 1.83 hours Edrift = 0 V/cm 83 Kr peak 41.5 kevee 39 Ar endpoint 565 kevee S1 [PE] Fits to 39 Ar and 83m Kr spectrum indicate LIGHT YIELD: 7.9 ± 0.4 PE/keVee at zero field and 7.0 ± 0.3 PE/keVee at 200 V/cm 31

42 For Nuclear Recoils SCENE (Scintillation Efficiency of Nuclear Recoils in Noble Elements) Proton Beam at University of Notre Dame Organic Liquid Scintillators (placed at precise deflection angles) Liquid Argon TPC (based on DarkSide design) 7 Li(p, n) 7 Be reaction produces low energy monoenergetic neutrons TOF measurement between target, LAr and organic scintillators allows clean identification of elastic neutron interactions of known energy 32

43 85 Kr Contamination E T Decay Events / [0 PE] escaping γ β+γ 514 kev (w/ 7.0 PE/keV) 687 kev (w/ 7.0 PE/keV) S1 β + S1 γ [PE] 99.57% decay via β (687 kev) 0.43% decay via β (173 kev) + delayed γ (514 kev) Measured meantime 1.64 μs in DS-50 Events / [0.1 µs] Decay time 1.64 ± 0.12 µs Rate in two branches (BR corrected) ± 2.2 cpd 33.1 ± 0.9 cpd t [µs] 33

44 Event Position in TPC z [cm] UAr z vs r 2 Top center Bottom r 2 [cm Entries ] The excess events at top and bottom are from γ induced ER events. Events within 1 cm from top are not reconstructed due to S1 and S2 overlap in waveform. 34

45 UAr First Results w/ S2/S1 cut f Energy [kev nr ] % 65% 80% 90% 95% 99% (S2/S1) Log ER S1 = PE NR AmBe calibration data S1 [PE] f 90 0 We have another discrimination power to suppress ERs (S2/S1 cut w/ 50% acceptance of NRs). 35

46 S2/S1 Electron Recoil Discrimination Electron and nuclear recoils produce different ionization densities that lead to different fractions of electrons that survive recombination Electron Recoil (S2/S1) Log NR ER Nuclear Recoil Am-Be Source f 90 0 Ratio of ionization and scintillation signal (S2/S1) can be used to distinguish between the two populations 36