1. A WIMP dark matter search with CsI(Tl) crystal SeungCheon Kim Department of Physics & Astronomy Seoul National University 1 Seminar@Fermilab (2012-07-31)

2. Outline A brief introduction about dark matter A Description about KIMS experiment A study about the background The recent results The prospect 2

3. 3 Dark matter ? Hubble image of coma cluster 1936 Zwicky first proposed the existence of dark matter, inferred from the motion of galaxies in coma cluster. => The total mass of the luminous matter is too small to explain the velocity of the galaxy in the cluster! 10 7 -10 8 K What is holding such hot gases ? The recent x-ray measurement of the cluster also requires the invisible matter!

4. 4 Big Bang Nucleosynthesis Gravitational Lensing Rotation curve for Galaxy Bullet cluster cosmic microwave background A pile of evidences for dark matter !!! 99 % of people believe the dark matter…

5. What we know about dark matter Dark (invisible)  no EM interaction rarely interacting very stable Massive  gravitationally evident non-relativistic (cold – most, not necessarily all of them) Influencing the structure formation of the Universe 5

6. What we know about dark matter Overwhelmingly abundant  the majority of the matter constituents in the universe Exotic  Not made of the known matter Need of the new theory other than SM (multifunctioning particle – not invented for DM – favored) 6

7. WIMP (Weakly Interacting Massive Particle) - a good dark matter candidate 7 Ω : energy density of the universe Ω = Ω DM + Ω Λ + Ω baryon = 1 for the flat universe Ω DM = 0.222 ± 0.026 from WMAP The weakly interacting relic particles can survive the annihilation and so exist in this level of the density. Supersymmetry and Univ- ersal extra dimension theories provide good candidates of this kind.  lightest neutralino lightest kluza-klein particle

8. WIMP (Weakly Interacting Massive Particle) - a good dark matter candidate 8 And, it is testable with the experiment, even though very challenging. It has been expected WIMP will deposit a few tens keV in the detector absorber. The event rate will be very small. Background reduction & discrimination is essential. σ χ-nucleon =10 -4 pb

9. KIMS (Korea Invisible Mass Search) A WIMP search experiment using CsI(Tl) scintillators It has been carried out at Yangyang underground lab(Y2L), South Korea. 9 Seoul National University: H.C.Bhang, J.H.Choi, K.W.Kim, S.C.Kim, S.K.Kim, J.H.Lee, J.I.Lee, J.K.Lee, M.J.Lee, S.J.Lee, J.Li, X.Li, S.S.Myung, S.L.Olsen, I.S.Seong Sejong University: U.G.Kang, Y.D.Kim Kyungpook National University: H.J.Kim, J.H.So, S.C.Yang Yonsei University: M.J.Hwang, Y.J.Kwon Ewha Womans University: I.S.Hahn Korea Research Institute of Standard Sciences: Y.H.Kim, K.B.Lee, M. Lee Tsinghua University : Y.Li, Q.Yue, J. Li

10. CsI (Tl) scintillator ? Widely used, well-known detector Easy handling Large scintillation yield : 60,000/MeV technique 10 nuclear recoil electron recoil Low hygroscopicity Pulse Shape Discrimination  Distinguishing nuclear recoil event at the statistical basis

11. CsI (Tl) scintillator ? Large atomic number, Cs (133), I (127) Good for A 2 scaling High spin expectation value for proton Sensitive to SD interaction 11

12. CsI (Tl) scintillator ? However, it has some inherent background. 137 Cs artificial radioisotope, half-life of 30 years beta decay & delayed (2.5min) gamma of 662 keV 134 Cs Produced by the capture of neutron from the cosmic ray muon, half-life of 2 years, beta decay & several prompt gamma rays 87 Rb It exists in the pollucite (ore for Cs). Natural abundance of 27.84 %, beta decay 12

13. CsI (Tl) scintillator ? It contains iodine (I) like DAMA/LIBRA experiment.  Cross-checking the WIMP-I interaction scenario for DAMA positive signal. So, it’s very good idea to search for WIMP using CsI (Tl) scintillators if one can manage its internal background. 13

14. One detector module : one CsI Crystal + 2 PMTs PMT : 3” PMT (9269QA), RbCs photo cathode (green extended) Crystal size: 8x8x30 cm 3 (8.7 kg) (Beijing Hamamatsu Photon Techniques Inc.) Background level: 2-3 counts/keV/kg/day (cpd) 14 Am241 calibration 13.9keV Np L  X-ray 17.8keV Np L  X-ray 20.8keV Np L  X-ray 26.35keV gamma Cs, I X –ray escape 59.54 gamma... Light yield:5-6 p.e/keV E th = 3 keV -data -simulation (GEANT4.9.5) CsI (Tl) detector for KIMS Event window is 40µs. Digitized with 400MHz FADC

15. 15 Detector array 3 x 4 detector array The total mass : 103.4 kg Multiple hit events  Vetoed for DM search But, references for calibration & event selection E of det10 Total E except det10 134 Cs 134 Ba 4+4+ 3+3+ 4+4+ 2+2+ 2+2+ 0+0+ 1969.87 1643.28 1400.55 1167.93 604.70 0.0 E [keV] β-β- 27.3% 2.50% 70.1% 0.033% 0.10% Decay for 134 Cs (T 1/2 =2.06yr) Measurement: 2.053±0.145yr

16. 16 Neutron shield & Muon Veto (30 cm mineral oil) Lead shield (15 cm) Polyethylene (5 cm) Copper shield (10 cm) 12 x CsI (Tl) crystal N 2 gas flow inside the Cu shield KIMS Detector system

17. Yangyang Pumped Storage Power Plant Experimental site: Yangyang Underground Laboratory (Y2L) Minimum depth : 700 m / Access to the lab by car (~2km) (Upper Dam) (Lower Dam) (Power Plant) KIMS (DM Search) AMoRE (Double Beta Decay Experiment) Seoul Y2L

18. The Experimental hall 18 Muon flux : 2.7 x 10 -7 /cm 2 /s (2000 m.w.e) Neutron flux : 8 ×10 −7 /cm 2 /s for 1.5-6 MeV neutrons Rn background in the air : 1-2 pCi/ l Clean room system Online-Monitoring of temperature, humidity, power stability, Rn background level

19. Background study 19 Multiple hit only α spectrum Crystal : 134,137 Cs, 87 Rb PMT : 238 U, 232 Th, 40 K Surface : progenies of 222 Rn

20. Background study pmt0 pmt1 20 Surface alpha (SA) background Scintillation at PMT body Clean acryl box PMT These events makes trouble in pulse shape discrimination. SA => Characterize its PDF PMT events  Find cuts to minimize its contamination.

21. γ & β background (electron recoil (ER) events) -main background Background level (cpd) & their origin Internal source~ 1 ( 134,137 Cs) ~ 0.1 ( 238 U, 232 Th) ~ 0.3 ( 87 Rb) PMT~ 0.5 ( 238 U, 232 Th, 40 K) Crystal surface (progenies of 222 Rn) 0.1 – 1 (?) (mainly, 210 Pb) 21 Background at the low energy A few tens of recoil energy will be measured as a few keV because quenching. The current background level below 10 keV : 2.8 counts/keV/kg/day (cpd) + PMT dark current Surface alpha (SA)~ 0.1 Scintillation at PMT body < 0.1

22. Rejection of events produced from PMT itself => PMT body scintillation & dark current (PMT events) PMT generates events by itself.  PMT body scintillation + After pulse + random coincidence + etc. => It can be problematic for the low energy, rare phenomena search. After pulse  A delayed (~1 μs) pulse following the primary pulse  The atom in the PMT vacuum ionized by the accelerating photo-electron drifts to the photo- cathode producing a big signal (spike-like). 22

23. Signal size => # of SPE Time span btw neighboring SPEs(us) pmt0pmt1 Afterpulses in PMT Dummy Detector Afterpulse by postive ion(H,He,N2...) in PMT 23 Clean acryl box PMT

24. PMT events rejection 24 Compton events From the PMT dummy detector data, We have developed the event selection condition to minimize the PMT events contamination. The Efficiency for nuclear recoil (NR) events (WIMP candidate) : ~30% tail events Clean acryl box PMT

25. Pulse Shape Discrimination (PSD) Discriminating the event type by the pulse shape Enabling the estimation of NR event rates statistically 25 PSD parameter Reference data for PSD Surface alpha (SA) : from the Radon-contaminated crystal Nuclear Recoil (NR) : by irradiating the test crystal with Am-Be neutron source Electron Recoil (ER) : compton scattering SA NR ER at 3 keV

26. Opening the data… Period: September 2009 – August 2010 => 317.15 live days Exposure: 24524.3 kgdays After applying the PMT noise selection cut, Assuming the data are composed of SA, ER and NR events Estimating the fraction NR event rates with PSD. Candidate events for WIMP : NR occurring in one detector 26

27. 27 The estimation of the NR events rate Pdf = f0 x F NR + f1 x F SA + (1-f0-f1) x F gamma The posterior pdf for f0 & f1 is obtained from Bayesian analysis method.

28. 28 The estimated NR event rates for 12 detectors 90 % limit 68 % interval SA The positive signals at higher energy bins seem correlated with SA contamination. The PSD PDF of NR at higher energy bins might be contaminated with gamma ray caused by the inelastic scattering. => need of further study For the conservative claim, in the combining all the detectors, the detectors with high SA contamination are excluded.

29. 29 The overall NR events rate det0, 8, 11 excluded Total exposure: 24524.3 kg days No meaningful excess of NR events is observed.

30. 30 KIMS NR event rates & DAMA/LIBRA annual modulation amplitude The annual modulation amplitude from DAMA  At 2-4keV, 0.0183±0.0022 counts/day/kg/keV  At the corresponding energy range (3.6-5.8 keV) in KIMS, (QF for DAMA =0.08 adopted) The 90 % C.L limit is 0.0098 counts/day/kg/keV This is well below the DAMA modulation amplitude. Any scenario with Iodine for DAMA is not consistent with KIMS results !!! DAMA/LIBRA The spectrum of the amplitude of the annual modulation

31. 31 KIMS & DAMA in iDM model For 70 GeV mass WIMP Cs I Expected spectrum for iDM for I & Cs target => M=70 GeV, δ=116keV, vesc=500 km/s, ve=235km/s, σ=0.93e-3 pb

32. 32 The new exclusion limits from KIMS Spin-Independent interaction Spin-dependent proton interaction Published in PRL 108, 181301 (2012)

33. Annual modulation study in KIMS The analysis of 2.5 years data is almost completed. PSD was not taken into account. There’s time-varying background from 134 Cs (2year half-life). The results will be reported very soon. 33 Sep. 2009 ~ Feb. 2012 3-6 keV

34. Prospect of KIMS experiments New PMTs are under test.  ~10 p.e/keV ? Decreasing the energy threshold  Reduced 238 U, 232 Th, 40 K Decreasing current background level by ~ 1 cpd The treatment of the surface background  Polishing or surrounding with active veto layer ? Adding low background NaI (Tl) scintillators is being planed. 34

35. Summary 12 CsI (Tl) scintillators (103.4kg) have been running at Yangyang underground lab. Most of backgrounds of the detector has been understood. The improved limit of WIMP-nucleon interaction has been derived based on 1 year data with PSD. Annual modulation study based on 2.5 years was almost done and will be reported very soon. New upgrade is being planed. 35

36. Back up 36

37. 37 Event rate of WIMP in the detector number of target nuclei number density of WIMP WIMP nucleon cross section Coherent enhancement along nucleus nuclear form factor The effect by the velocity Distribution of WIMP If σ n = 10 -5 pb, for 100 GeV WIMP and Iodine target, The expected event rate is 1-2 counts/keV/kg/day,considering quenching.

38. CsI (Tl) detector for KIMS One detector module : one CsI Crystal + 2 PMTs PMT : 3” PMT (9269QA), Electron tube RbCs photo cathode (green extended) Crystal size: 8x8x30 cm 3 (8.7 kg) (Beijing Hamamatsu Photon Techniques Inc.) Background level: 2-3 counts/keV/kg/day (cpd) 38 Am241 calibration 13.9keV Np L  X-ray 17.8keV Np L  X-ray 20.8keV Np L  X-ray 26.35keV gamma Cs, I X –ray escape 59.54 gamma 5-6 p.e /keV

39. Neutron shield(30cm mineral oil) Lead shield (15cm) Polyethylene(5cm) Copper (10cm) CsI(Tl) Scintillator Neutron detector Muon detector (Neutron sheild) KIMS Detector system N 2 gas flow inside the Cu shield

40. 40

41. 41 Annual Modulation Studies (w/o PSD) Sep. 2009 ~ Feb. 2012 Total DAQ rate is under 6Hz. 2.5 year data to see annual modulation ; 75.53 ton∙days The temperature of detector array is 20 - 21.6 O C depending on the position, and it is maintained stably with a maximum fluctuation of around 0.2 O C.

42. 42 3-6 keV Annual modulation amplitude is obtained including the exponential decay of 134 Cs. Annual modulation amplitude is consistent with null. The 90% upper limit of the amplitude is comparable to DAMA’s annual modulation signal (0.0189 cpd/kg/keV) We are trying to give final numbers shortly in this summer.