XMASS

2012 Shanghai Particle Physics and Cosmology Symposium

Hiroshi Ogawa (ICRR, Univ. of Tokyo)Sep. 16th , 2012

XMASS experiment

Dark matter

Double beta

Solar neutrino

2

pp/7Be chain: n+e n+e

136Xe 136Ba + 2e-

χ+Xe χ+Xe

Multi purpose low-background and low-energy threshold experiment with liquid Xenon

Large photon yield Low threshold Scintillation wavelength (175 nm, detected directly by PMT) High density (~3 g/cm3) Compact detector Large Z (=54) Self-Shielding ~-100 liquid : easy to use.℃ Purification (distillation) No long life radioactive isotope

Liquid Xenon :

Direct Detection Principle

• From the density of dark matter in the galaxy:

• Every liter of space: 10-100 WIMPs,

• moving at 1/1000 the speed of light

• Less than 1 WIMP/week will collide with an atom in 1kg material

WIMPs elastically scatter off nuclei in targets, producing nuclear recoils.

Xe (A=131) is one of the best target.

R0: Event rate F: Form Factorshould be calculated in each nuclei

Maxwellian distribution for DM velocity is assumed.v0:dispersionV :velocity onto target, VE: Earth’s motion around the Sun

Spin independent case:

Larger A is higher event rate

Dark Matter（WIMP）

Deposit Energy

• XMASS : single phase detector – Large volume and simple structure, op

eration. • 1 ton scale xenon detector, 100kg f

or fiducial volume.– Background reduction technique :

• Self shielding • Reconstruction by hit pattern of PM

Ts – High light yields & Large photon cover

age (15 pe/keV)• Low energy threshold (< 5 keVee ~

25 keVNR ) for fiducial volume• Lower energy threshold: 0.3 keV fo

r whole volume– Large Scalability, simple to construct.

Characteristics of XMASS

Self shielding

1ton 10ton

XMASS detector

5

XMASS detector : siteKamLAND

Super-KXMASS(Lab-C)

CANDLES

IPMU　Lab1 CLIONEWAGE

Lab2/EGadKamioka Mine

Tokyo

Kamioka Mine

10m

11m

Water tank

Refrigerator

72 20inch PMTs (veto)

Elec. hut

•72 20-inch PMTs are installed to veto cosmic-ray muon (<10-6 for thr-mu, 10-4 for stop-mu).•Water is active shield for muon induced neutron and also passive shield for gamma-ray and neutron from rock/wall.

Structure of XMASS detector• Detector

– Single phase (scintillation only) liquid Xenon detector

– Operated at -100oC and 0.065MPa– 100 kg fid. mass, [835 kg inner mass (0.8 m)]– Pentakis-dodecahedron

12 pentagonal with 5 triangle– 630 hexagonal & 12 round PMTs with 28-39%

Q.E.– photocathode coverage: > 62% inner surface

1.2m diameter

Detector construction 2010 Detector construction 2010 11stst application of WC tank for application of WC tank for WIMP search WIMP search

By fall, 2010

8

Detector performance

9

• Calibration– Source Rod (57Co, 241Am, 137

Cs, 109Cd, 55Fe)– External sources: 60Co, 137Cs,

232Th, Neutron• Normal Data taking (physics runs)• Change of the physical condition of

Xenon.– High/Low pressure run

• Change of the refractive index of Xe

– O2 runs: change of the absorption length

– Boiling runs: create convection inside of the detector

• PMT gain reduced run • Evaluate the signal from

radon daughter in xenon.

• Gas run- Important to identify the

surface BG• BG measurement of the

detector parts（ attach the material at the end of the calibration source rod） Al, GORE-TEX, Cu, Ni

plate

Data taking :commissioning run Nov.2010-June.2012

Detector response for a point-like source (~WIMPs)

• 57Co source @ center gives a typical response of the detector.

• 14.7p.e./keVee ( 2.2 for S1 in XENON100)

• The pe dist. well as vertex dist. were reproduced by a simulation well.

• Signals would be <150p.e. exp shape.

total photo electron

dataMC

122keV

136keV59.3keV of W

~4% rms

data MC

reconstructed vertex+15V

RI source with rod

RI in PMTActivity per 1PMT(mBq/PMT)

238U-chain 0.70+/-0.28

232Th-chain 1.51+/-0.31

40K-chain 9.10+/-2.15

60Co-chain 4.09+/-0.22

~20

0keV

Energy spectrum data and PMT MC~1

MeV

~50k

eV

MC: PMT rays

Observed data

~x100

Measured Spectrum (Whole Volume) and background evaluation

Gamma from PMTs

RI of PMT are measured by HPGe detector per PMT parts.(+40% for 60Co)Data has large excess in <200keV compared with data. ~x100 difference in ~5keV

Gamma from PMTs had been assumed as mainly background sources

~1M

eV

~50

keV

MC: PMT rays

MC: PMT Al238U & Pb210Surface 210Pb

Observed data

~1M

eV

Measured Spectrum (Whole Volume) and background evaluation

Aluminum Seal

RI from PMT aluminum seal

–Aluminum sealing used for the PMT between quarts window and metal body. It contains 238U-230Th and 210Pb-206Pb.

210Pb on detector surface

222Rn daughter put btw. construction work

These background are faced to liquid xenon.X-ray and beta from RI emit to Liquid xenon directly.

14

• GORE-TEX: between PMT and holder used for a light seal. It contains 06±3% of modern carbon.

– GORE-TEX might explain the background below 5keV

• But parameters ( ex. transparency of light inside of GORE-TEX) are not well known

• We will remove GORE-TEX in future detector refurbishment

Measured Spectrum (Whole Volume) and background evaluationBelow 5keV

14C in GORE-TEXMC: GORE-TEXModern C: 7.5%LXe inside scintillate0.3mm photon att.

~5ke

V

MC: GORE-TEXModern C: 7.5%LXe inside scintillate0.1mm photon att.

~5ke

V

mDM = 30GeVSI = 1.4x10-41cm2

Background estimates

Material Measured RI and activity Methods of the measurements

PMTs

(per PMT)

238U: 0.704 ± 0.282 mBq

232Th: 1.51 ± 0.31 mBq

60Co: 2.92 ± 0.16 mBq

40K: 9.10 ± 2.15 mBq

HPGe detector measurement for each parts and whole PMT

PMT aluminum

(210g)

238U-230Th: 1.5 ± 0.4 Bq

210Pb: 5.6 ± 2.3 Bq

232Th: 96 ± 18 mBq

235U: ~67 mBq

HPGe detector measurement.

By calculation

Detector surface 210Pb: ~40 mBq

Alpha candidates using FADC data

Surface: PMT window 59%, PMT Al 7.0% PMT rim 7.0%, GORETEX 3.7%, Cu 23.3% (surface 7.8%, wall 14.2%, bottom 1.3%)

GORE-TEX for PMTs

(120g)

14C: 0.4 ± 0.2 Bq

(6±3% of modern carbon)

210Pb: 26.5 ± 11.9 mBq

14C: modern carbon measurement.

210Pb: Ge measurement.

Internal RI in xenon

85Kr: <2.7 ppt

214Pb: 8.2 mBq

85Kr : API-MS measurement

214Pb : ~222Rn concentration in detector

Low background even with the surface BG

• Our BG is still quite low, even with the extra surface BG!• In principle, the surface BG can be eliminated by vertex

reconstruction. Optimization of the reconstruction program is on going to minimize a possible leakage to the inner volume.

E. Aprile, 2010 Princeton

XMASS full volume

• Our sensitivity for the low mass WIMP signals at low energy without reconstruction will be shown. Ev

ens/

kg/d

ay/k

eV

16

Physics analysis

17

Physics analysis(sensitivity study)

•Whole Volume Analysis (Large BG, Large target mass of 835 kg, low energy threshold, no reconstruction)

1) 4 hit threshold analysis• Low mass region search

2) 1keVee threshold analysis• Check DAMA modulation

3) Axion DM (super-WIMPs)4) Solar Axions

•Fiducial Volume analysis1) Standard WIMPs search (> 5 keV)

• Event reconstruction/reduction Standard WIMPs:Prefer heavy massLow BG

Light WIMPs & ALP:Low thresholdBG less important

Physics analysis:Physics analysis:Low energy, full volume analysis for low mass WIMPsLow energy, full volume analysis for low mass WIMPs• The dark matter signal rapidly incre

ase toward low energy end. The large p.e. yield enables us to see light WIMPs. Try to set absolute maxima of the cross section (predicted spectrum must not exceed the observed spectrum).

• The largest BG at the low energy end is the Cherekov emission from 40K in the PMT photo cathodes.

Light WIMPs & ALP:Low thresholdBG less important

Selection criteria for data: Triggered by the inner detector only (no water tank trigger)RMS of hit timing <100ns (rejection of after pulses of PMTs)Cherenkov rejectionTime difference to the previous/next event >10ms

Detail of the Cherenkov rejectionDetail of the Cherenkov rejection• Basically, separation between

scintillation lights and Cherenkov lights can done using timing profile.

• (# of hits in 20ns window) / (total # of hits) = “head total ratio” is a good parameter for the separation. Total # of hits Total # of hits

20ns 20ns

Cherenkov like Scintillation like

20

• Cherenkov events peaks around 1 scintillation ~ 0.5• Low energy events observed in Fe55 calibration source as well as DM simul

ation (t=25ns) show similar distributions.• Efficiency ranges from 40% to 70% depending on the p.e. range.

0 40 80 120 160 2000 40 80 120 160 200

(PE)(PE)

Head t

ota

l ra

tio

Head t

ota

l ra

tio

Black: real data Black: real data

Blue: DM MC

p.e. distribution after each cutp.e. distribution after each cut• 6.8 days x 835kg data• The Cherenkov events are efficiently reduced by the cut.

• We analyze the events above > 0.3 keVee for entire volume

21

ID only event (trigger ID == 1)dT > 10msRMS of hit timing < 100nsCherenkov cut

Uncertainties• Major uncertainty is the scintillation efficiency of nuclear r

ecoil in liquid xenon.• Uncertainties of the trigger thre. (hard trig. 4hits), cut eff.,

and energy scale are also properly taken into account.

Scintillation efficiency as a function of energyE. Aprile et al., PRL 105, 131302 (2010)

22

Spectrum and Sensitivity Spectrum and Sensitivity • Spectrum shows that observed data and MC WIMPs signal with best fit per WIMPs mass. • Sensitive to the allowed region of DAMA/CoGeNT. • Some part of the allowed regions can be excluded.

DAMA

CoGeNTXMASS

23

WIM

P c

ross

sect

ion

on

nu

cleon

(cm

2)

XMASS observed energy [keVee]

Cou

nt/d

ay/k

g/ke

Vee

GeV

Preliminary

Preliminary

Progress of annual modulation analysis

1-4 keVee(Xe) 2-6 keVee (Na)

• Check the DAMA modulation• QF(Na)0.25, Leff(Xe)0.15• 26keVee(Na) 824 keVNR 14keVe

e(Xe)– but 1/30 sensitivity recoil shape, A2

2: 10.8 for flat, 23.8 for a modulation

Preliminary

Preliminary

2013 data taking (after refurbishment) will cover the maximum and minimum flux region.

Dark matter axion search in XMASSDark matter axion search in XMASS• The DAMA signal may be due to electromagnetic interac

tion of WIMPs to the NaI detectors by such as a non-relativistic axion dark matter. See J. Collar, arXiv: 0903.5068

• XMASS can search for dark matter axion by detection through axio-electric effect.

• We show a preliminary result based on 6.8 days x 835kg data.

gaee

Limit on the axio-electric couplingLimit on the axio-electric coupling

DAMA allowedCoGeNTCDMSXMASS

gaee

Sensitivity withspectrum fittingwith background MC

Preliminary

Preliminary

(keVee)

Observed spectrumExpectation for ma=3keV

Even

ts/k

eV/k

g/da

y

We obtained a limit on the axio-electric coupling (gaee) by comparing the observed spectrum and the expectation.

The result can be further improved above 5keV by fitting spectrum.

Solar axion search in XMASSSolar axion search in XMASS

Observed spectrumMC axion signalsAbs upper limit gaee=4.5e-11

XMASS can also search for solar axion emitted owing to Bremsstrahlung and Compton processes in the Sun, by detection through axio-electric effect. We obtained a limit on axio-electric coupling (gaee)

Solar axion flux (Derbin et al., arXiv:1206.4142)

gaee=10-10

Allowed mass< 200eV for KSVZ<2eV for DFSZ

Originalfigure fromDerbin et al., 1206.4142

XMASSLimit by DM

XMASSLimit by solar axion4.5e-11

Solar limit2.8e-11

Limits on gLimits on gaeeaee from XMASS (DM+solar) from XMASS (DM+solar)

Preliminary

Preliminary

Future plan

29

Plan: Refurbishment work

• Tuning of reconstruction/reduction is on going but for better sensitivity, removing the origins of BG must be done.

• To reduce the BG caused by Aluminum, we are planning to cover the part and surfaces by copper rings and plates:

• BG > 5keV must be reduced significantly (1/100 of current BG).• Schedule: start physics run from 2013

Expected sensitivity with fiducialization

XENON100CDMSII

XMASS 2keVee thre. 100d

Black:signal+BGRed:BG

Expected energy spec.

1 year exposurep=10-44 cm2

50GeV WIMP

Spin Independent

XMASS 5keVee thre. 100d

Initial target of the energythreshold was ~5keVee.Because we have factor ~3better photoelectron yield,lower threshold = smaller massdark matter may be looked for.

31

XMASS1.5 (2015-)• Total mass: 5 tons• Fiducial mass: 1 ton 10

0kg• Backgrounds

– No dirty aluminum– No GORETEX– Less surface 210Pb

• New PMTs.• Expect 10-5 dru• Sensitivity sSI < 10-46 cm2 (> 5 ke

V)• Low threshold analysis could rea

ch a few x 10-42 cm2 around a few GeV region.

XENON100

CDMSIIEDELWEISSII

DAMA

CoGeNT

SummarySummary• The XMASS 800kg detector was constructed and started co

mmissioning late 2010.• We completed commissioning data-taking and physics anal

yses are on-going.• BG level is not as low as originally expected, but now the c

omposition is well understood above 5keV.• Some preliminary results on dark matter and axion searche

s are shown. More results will come later.• By improving software (reconstruction/BG reduction) and ha

rdware refurbishment, we aim at the dark matter search with the original sensitivity.

• And XMASS1.5 will progress. Physics run is planed from 2015.

XMASS collaborationICRR, University of Tokyo K. Abe, K. Hieda, K. Hiraide, Y. Kishimoto, K. Kobayashi, Y. Koshio,

S. Moriyama, M. Nakahata, H. Ogawa, H. Sekiya, A. Shinozaki, Y. Suzuki, O. Takachio, A. Takeda, D. Umemoto, M. Yamashita, B. Yang

IPMU, University of Tokyo J. Liu, K. Martens

Kobe University K. Hosokawa, K. Miuchi, A. Murata, Y. Ohnishi, Y. Takeuchi

Tokai University F. Kusaba, K. Nishijima

Gifu University S. Tasaka

Yokohama National University

K. Fujii, I. Murayama, S. Nakamura

Miyagi University of Education

Y. Fukuda

STEL, Nagoya University Y. Itow, K. Masuda, H. Takiya, H. Uchida

Kobe University K. Ohtsuka, Y. Takeuchi

Seoul National University S. B. Kim

Sejong University N.Y. Kim, Y. D. Kim

KRISS Y. H. Kim, M. K. Lee, K. B. Lee, J. S. Lee

Cryogenic

700LLiq.

Storage

liquid pump

Condenser360W

gas pump

filtersCalibration line

Cable line

filters

Outer vacuum

857kg

10 m3 x 2

gas storage

emergency gas pump

100L/min

evaporator

Water tank

Liquid circulation ~5L/min

Gas circulation <30L/minXMASS Xenon

operating system

Xenon keeping : Pulse tube refrigerator.Xenon circulation : Gas phase: < 30 L/minLiquid phase: ~ 5 L/minXenon collection :700L liquid storage tank.10m3 x 2 gas storage tank.

We successed the xenon filling/collection for 5 times with keeping the amount and quality of xenon.

Internal RI in xenon

100 500 1000Time difference (s)

Fitting withan expecteddecay curve

1st event (214Bi )2nd event (214Po )

222Rn 220RnRadon concentration: 222Rn : 8.2+/-0.5mBq220Rn : <280microBq (90%C.L.)

220Rn candidate216Po candidate ~accidental

164us 145ms※214Pb, 212Pb(daugther of radon) is source of background in low energy. These background is remained even after fiducial volume cut. 8.2mBq ～ 2x10-4dru contribution.