Post on 26-Sep-2020
R. Budnik, Columbia University 1SSI 2012
Dark Matter detection with the XENON Experiment
Ranny BudnikColumbia University
On behalf of the XENON100 collaboration
R. Budnik, Columbia University 2SSI 2012
N
χ
χNumber of targets
WIMPmass
WIMPdensity
Interactioncross section
NuclearForm factor
WIMPvelocity distribution
Only those WIMPs with velocity above threshold will contribute to signalFor Spin Independent interactions the cross section is enhanced by a factor A2 (coherent scattering)
Dark Matter Direct Detection
Expected interaction rate
Goal: Observe WIMP interactions with some target material
Reminder:
R. Budnik, Columbia University 3SSI 2012
Liquid Xenon for Dark Matter Search
● Large atomic number A~131 best for SI interactions (σ~A2).Need low threshold.
● ~50% odd isotopes: SD interactionsIf DM detected: probe physics with the same detector using isotopically enriched media.
● No# long-lived Xe isotopes.But control Kr-85, Rn-222. #Xe-136 2νββ
● High Z (54) and density: compact & self-shielding
● Scalability to large mass.
● “Easy” cryogenics (-100°C).
● Efficient and fast scintillator.
● Good ionization medium, long drift.
● Background discrimination in TPC.
– Ionization/Scintillation
– 3D imaging of TPC
WIMP Scattering Rate
R. Budnik, Columbia University 4SSI 2012
XENON100
The Liquid Xenon Dual Phase TPC Ionization + Scintillation
● WIMP recoil on Xe nucleus in dense liquid (2.9 g/cm3) → Ionization + UV Scintillation
● Detection of primary scintillation light (S1) with PMTs.
● Charge drift towards liquid/gas interface.
● Charge extraction liquid/gas at high field between ground mesh (liquid) and anode (gas)
● Charge produces proportional scintillation signal (S2) in the gas phase (12 kV/cm)
● 3D position measurement
– X/Y from S2 signal. Resolution few mm.
– Z from electron drift time (~0.3 mm).
R. Budnik, Columbia University 5SSI 2012
TimeA
mpl
itude
S1 S2
S1
S1
S2
S2
neg HVE~1 kV/cm
pos HVE~10 kV/cm
E
TPC = time projection chamber
Dual Phase TPC
R. Budnik, Columbia University 6SSI 2012
Dual Phase TPC
151 µs
PR
L 107, 131 302 (2011)
Astropart. Phys. 35, 573 (2012)
NR
ER
ER
NR
3d Vertex Reconstruction Signal/Background Discrimination
R. Budnik, Columbia University 7SSI 2012
The XENON program
Columbia Rice UCLA U Zürich Coimbra LNGS Mainz SJTU
Bologna MPIK NIKHEF Purdue Subatech Münster WIS
2005-2007:XENON10
XENON R&D
XENON: A phased WIMP search program 2010-2015: XENON1T
2008-201x: XENON100
R. Budnik, Columbia University 8SSI 2012
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
● passive shield (Pb, Poly, Cu, H2O, N2 purge)
Astropart. Phys. 35, 573 (2012)
161 kg LXe, 62 kg in target242 1'' x1'' PMTs
R. Budnik, Columbia UniversitySSI 2012
Laboratori Nazionali del Gran Sasso (LNGS)
LNGS: 1.4km rock LNGS: 1.4km rock (3700 mwe)(3700 mwe)
R. Budnik, Columbia University 10SSI 2012
(spin-independent) WIMP Limit 2011
PRL 107, 131302 (2011)
XENON100 sets the most sensitivelimit over a large WIMP mass range
Challenges the CoGeNT, DAMA, CRESST-II signals as being due to light mass WIMPs
already cited 362x
Limit derived with Profile Likelihood methodPRD 84, 052003 (2011)
CRESST (2011)
SIMPLE (2011)
CRESST (2007, reanalysis)
R. Budnik, Columbia University 11SSI 2012
(spin-independent) WIMP Limit 2011
PRL 107, 131302 (2011)
XENON100 sets the most sensitivelimit over a large WIMP mass range
Challenges the CoGeNT, DAMA, CRESST-II signals as being due to light mass WIMPs
already cited 362x
Limit derived with Profile Likelihood methodPRD 84, 052003 (2011)
CRESST (2011)
SIMPLE (2011)
CRESST (2007, reanalysis)
O U T D A T E D !!!
R. Budnik, Columbia University 12SSI 2012
XENON100 :
new results for 2012
R. Budnik, Columbia University 13SSI 2012
Data Taking
run_08
Data Collection Stability
Data taking over 13 monthsfrom Feb 28, 2011 to March 31, 2012→ full annual cycle
3 interruptions for maintenance
224.56 live days of dark matter data
To our knowledge, no largeLXe detector has ever been operated under such stable conditions for that long
R. Budnik, Columbia University 14SSI 2012
Improvements● Exposure more than doubled
● Lower thresholdS2>150 PE, S1>3 PE (6.6 keVr)
● Lower Background
● Much more calibration data 35x more ER calibration in ROI AmBe before and after run
● Higher LXe purity → smaller corrections
R. Budnik, Columbia University 15SSI 2012
Selected CalibrationsPosition dependent Corrections: Cs137, AmBe inelastic (40 keV), Xe* (164 keV)Kr83m (planned)
Agreement better than 3%
Electron Lifetime: Cs-137
375 – 610 µs (average 514 µs)
Electron Recoil Band (Background):Co60, Th232
Nuclear Recoil Band (Signal):Neutrons: AmBe
definition of WIMP search region, discrimination
P R E L I M I N A R Y
arXiv:1107.2155
R. Budnik, Columbia University 16SSI 2012
ER/NR Discrimination
Discrimination comparable to previous runs: ~99.5% ER rejection @ 50% NR accpetance
R. Budnik, Columbia University 17SSI 2012
Background of this Run
R. Budnik, Columbia University 18SSI 2012
Data Analysis: All data
More information on XENON100 data analysis in arXiv:1207.3458
Blinded ROI
R. Budnik, Columbia University 19SSI 2012
Basic Quality Cuts
R. Budnik, Columbia University 20SSI 2012
Single Scatter Selection
R. Budnik, Columbia University 21SSI 2012
Threshold and Fiducial Volume
R. Budnik, Columbia University 22SSI 2012
Consistency Cuts
S2 Width:
R. Budnik, Columbia University 23SSI 2012
Select Energy Range
R. Budnik, Columbia University 24SSI 2012
ER Rejection
99.75%
R. Budnik, Columbia University 25SSI 2012
WIMPs are Nuclear Recoil-like
„Benchmark Region“
R. Budnik, Columbia University 26SSI 2012
Background Prediction
Neutron background:- (,n)+sf and muon induced neutrons- MC simulation using the exact XENON100 geometry and measured contaminations
Expect: (0.17 +0.12 –0.07) events
ER background:- activity of the detector and shield- intrinsic radioactivity in the LXe (→ lowered this run)- use ER calibration to model background by scaling it to the observable DM data
Expect: (0.79 +-0.16) events
Sum: (1.0 +- 0.2) events
The same background model is implemented in the PL analysis
R. Budnik, Columbia University 27SSI 2012
R. Budnik, Columbia University 28SSI 2012
During ...
R. Budnik, Columbia University 29SSI 2012
… and after ...
R. Budnik, Columbia University 30SSI 2012
… Unblinding
(1.0 0.2) events expected 2 events observed
→ 26.4% probability that background fluctuated to 2 events→ PL analysis cannot reject the background only hypothesis
No significant excess due to a signal seen in XENON100 data.
R. Budnik, Columbia University 31SSI 2012
The new XENON100 Limit
2.0 x 10 – 4 5 cm² @ 55 GeV/c²Theory region combined from:Strege et al., JCAP 1203, 030(2012)Fowlie et al., arXiv:1206.0264Buchmueller et al., arXiv:1112.3564
ArXiv:1207.5988 PRL submitted
R. Budnik, Columbia University 32SSI 2012
LNGS, Italy
XENON1T(2011-2015)
● Liquid xenon TPC to explore σ ~ 2×10-47 cm2
● Detector size: ~ 1 m3, ~ 3 t LXe, ~ 1 t fiducial mass
● Water Cerenkov Muon Veto● Approved by INFN.
● Funded.● Construction start: fall 2012.
XENON1T in Hall B (next to Icarus) @ LNGS
Next on the agenda:
R. Budnik, Columbia University 33SSI 2012
The Future of Direct Dark Matter Searches(next ~5 years, the XENON perspective...)
Spin-independent sensitivity
R. Budnik, Columbia University 34SSI 2012
… but we hope for a detection
For a WIMPwith 10-45 cm2
~100 events
(Assuming standard isothermal halo, 220 km/s, escape vel. 540 km/s)
R. Budnik, Columbia University 35SSI 2012
XENON1T design challenges
Background* Xe purity (e- lifetime)
HVTotal Rn/Kr
XENON100 ~5∙10-3 dru (events/kg/keV/day)
Kr: ~20 pptRn ~65 μBq/kg
160 kg @ ~400 µs In several months
30 cm @0.53 kV/cm
XENON1T essentials
~5∙10-5 dru(Events/kg/keV/day)
Kr: 0.5 pptRn: ~1 μBq/kg
~3 tons @ ~1000 μsIn ~2 months
100 cm @ 1 kv/cm
By how much should we improve?
X 100Kr: X 40Rn: X 50
X 3 (purity)X 50 (purification
speed)X 6
* In FV, including Veto, before discrimination
R. Budnik, Columbia University 36SSI 2012
Summary● XENON100 was continuously running for more than 1
year in reduced background conditions
● Observed 2 events with a background expectation of 1, set a 90% CL limit in the WIMP-nucleon cross section of 2.0x10-45cm2 for a 50GeV WIMP
● Next on the agenda for XENON100:
– More analyses of the data, e.g. SD, annual modulation and more
– Continue running XENON100 with lower Kr and Rn
● We begin construction of XENON1T and reach a sensitivity of ~2x10-47 cm2 by 2017
R. Budnik, Columbia University 37SSI 2012
Backup
38SSI 2012
Impact of Leff● As an excersise, we have computed the same limit with the
approximation that Leff is 0 below 3 keVnr (red line in the figure)
● The impact on the limit is below 5% for all the relevant mass range
R. Budnik, Columbia University 39SSI 2012
Energy Scale
Enr=S1
Ly⋅Leff
⋅Se
Sr
Measured signal
Light quenching due to electric field
for gammas @122keV
Scintillation eff.for nr at 0 field
Light quenching due to electric
fieldFor nuclear
recoilsLight yieldfor gamma @122 keV
40SSI 2012
Low mass WIMP in XENON100
m = 8 GeV/c 2 σ=1.0x10 – 40 cm2
41SSI 2012
CRESST-like WIMP in XENON100
m = 25 GeV/c 2 σ=1.6 x10 – 40 cm2
R. Budnik, Columbia University 42SSI 2012
Events in Benchmark Region● visual inspection:
valid waveforms
● at 7.1 keVr and 7.8 keVrboth events between 3 and 4 PE
● rather low wrt the NR calibration data
● no low S2/S2-events below threshold