Backgroundandsensitivity predictionsforXENON1T › dm16 › talks › selvi.pdf · 2016-02-19 ·...

Marco Selvi INFN -‐ Sezione di Bologna

(on behalf of the XENON collaboration)

Background and sensitivity predictions for XENON1T

1

Feb 19th 2016, UCLA Dark Matter 2016

M. Selvi – UCLA DM 2016 – XENON1T backgrounds & sensitivity

Outline

2

•  Description of the detector;

•  Prediction of the ER and NR backgrounds through a GEANT4 simulation of the experiment;�

•  Conversion of the energy released in the TPC by signal and background events into the S1 and S2 signals seen in the detector;�

•  Study the sensitivity to WIMP-nucleus spin-independent interactions.�

E. Aprile et al. (the XENON collaboration)“Physics reach of the XENON1T dark matter experiment”,arXiv:1512.07501, submitted to JCAP


The XENON1T experiment

3


The XENON1T experiment

4

96 cm

97 cm χχ


Main Backgrounds

Nuclear Recoils (NR): •  Radiogenic neutrons: spontaneous fission and (α, n) reaction from the U and Th

chains in the detector components.•  Muon-induced neutrons.•  Coherent scattering of neutrinos (mostly solar) off the Xe nuclei.

Electron recoils (ER): •  low energy Compton

scatters from the radioactive contaminants in the detector components: U and Th chains, 40K, 60Co, 137Cs.

•  Intrinsic contaminants: β decays of 222Rn daughters, 85Kr, 136Xe.

•  Elastic scattering of solar neutrinos off electrons.

NR

ER

5


MC simulation in GEANT4

6

Diving bell

PTFE pillars

Supports

Field shaping

electrodes

PMTs

PMT bases

PTFE reflector

Copper plate

PTFE bottom plateCathode

Top PMT array

Bottom PMT array

A single PMT

The whole TPC


ER from the materials

Electronic Recoil Energy [keV]0 500 1000 1500 2000 2500

] -1

keV

)⋅

day

⋅

ER R

ate

[ ( k

g

8−10

7−10

6−10

5−10

4−10

3−10

2−10

1−10

Pb214

Bi214 Bi214

Bi214

Bi214

Bi214Bi214

Bi214

Bi214 Bi214

Bi214

Bi214

Tl208

Tl208 Tl208Bi212

Ac228 Ac228

Ac228 Tl208Cs137

Co60 Co60

K40

Total Cryostat Shells

Cryostat Flanges SS in the TPC

PMT and Bases PTFE and Cu in the TPC

]2 [mm2R0 20 40 60 80 100 120 140 160 180 200 220

310×

Z [m

m]

-400

-300

-200

-100

0

100

200

300

400 ] -1

keV

)⋅

day

⋅

Rate

[ ( k

g

-610

-510

-410

-310

-210

-110

ER background �from the materials�

in 1 t Fiducial Volume, �in [1, 12] keV:�

7.3 10-6 (kg day keV)-1

7

Extensive screening campaign using gamma (Ge ) and mass (ICP-MS) spectrometry.�Eur.Phys.J. C75 (2015) 546


Total ER background

8

Assumptions on the intrinsic backgrounds:�0.2 ppt of natKr (already achieved in XENON1T distillation column tests),10 µBq/kg of 222Rn (conservative estimation based on Rn emanation meas.).

Fiducial Mass [kg]200 400 600 800 1000 1200 1400 1600 1800

] -1

keV

)⋅

day

⋅

ER R

ate

[ (kg

7−10

6−10

5−10

4−10

3−10Total Xe136 Total except materials

Rn222 Kr85 materials

νsolar

Electronic Recoil Energy [keV]0 500 1000 1500 2000 2500

] -1

keV

)⋅

day

⋅

ER R

ate

[ ( k

g

8−10

7−10

6−10

5−10

4−10

3−10

2−10

1−10

Total Rn222

materials Kr85

Xe136 νsolar

222Rn (mainly from 214Pb β-decay) is the most relevant source of ER background in most of the TPC.


Total ER background

9

•  The numbers above are before any ER/NR discrimination.•  In 1 t FV, the background from the material is at the same level as the

one from solar neutrinos and from Kr


NR from radiogenic neutrons

10]2 [mm2R

0 20 40 60 80 100 120 140 160 180 200 220310×

Z [m

m]

-400

-300

-200

-100

0

100

200

300

400 ]-1

keV

)⋅

day

⋅

NR

Rate

[ ( k

g

-910

-810

-710

-610

-510

Fiducial Mass [kg]200 400 600 800 1000 1200 1400 1600 1800

]-1

NR

Rate

[ y

-210

-110

1

[3, 50] keV[4, 50] keV[5, 50] keV

Neutron Energy [MeV]0 2 4 6 8 10

]-1

Neu

tron

Yiel

d pe

r Par

ent D

ecay

[MeV

-1610

-1510

-1410

-1310

-1210

-1110

-1010

-910

-810

-710

-610

-510Th230U - 238

Pb207U - 235

Pb206Ra - 226

Ac228Th - 232

Pb208Th - 228

In 1t FV, [4, 50] keV: �0.6 events / y �

(with no discrimination)

Neutron yield from �SOURCES-4C


Nuclear Recoil Energy [keV]0 10 20 30 40 50 60 70 80 90 100

]-1

keV

)⋅

day

⋅

NR

Rate

[(kg

-1110

-1010

-910

-810

-710

-610

-510

-410

-310

-210

CNNSRadiogenic neutronsMuon-induced neutrons

Total NR background

11

Source Background (ev/y)

Radiogenic neutrons 0.6 ± 20%

Muon-‐induced neutrons

< 0.01 (muon veto ON)

CNNS 0.02 ± 20%

Total NR 0.62 ± 0.12

Single ScaBer, 1 t Fiducial Volume, [4, 50] keVr, 100% NR acceptance

Given the very steep spectrum of NR from CNNS, its contribuQon will become more relevant aSer the conversion into the S1, S2 signals,

considering the detector response and energy resoluQon.


Light & charge emission model

12

ER: NEST model� M. Szydagis et al., JINST 6 (2011) P10002, [arXiv:1106.1613]

NR: direct measurements of Leff,�Qy measured by XENON100�E. Aprile et al., Phys.Rev.D88 (2013) 012006, [arXiv:1304.1427]

NEST is based at low energies on the direct �measurement at Columbia and Zurich:�Phys. Rev. D 86, 112004 (2012) �Phys. Rev. D 87, 115015 (2013) ��


@122 keV

Light Collection EfGiciency

13

MeshMeshMeshWiresMesh

Ly at 500 V/cm | 4.6 PE/keV

]2 [mm2R0 20 40 60 80 100 120 140 160 180 200 220

310×

Z [m

m]

400−

300−

200−

100−

0

100

200

300

400

LCE

[%]

26

28

30

32

34

36

38

40

42

44

Absorption length [m]10 20 30 40 50 60 70 80 90 100

LCE

[%]

20

25

30

35

40

PTFE with 99% reflectivityPTFE with 95% reflectivityPTFE with 90% reflectivityOur assumption

Ly [P

E/ke

V]

4.0

4.5

5.0

5.5

6.0

6.5

7.0

7.5

8.0

We generate optical photons uniformly inside the TPC, and follow their propagation inside it with all the reflections, refractions, absorptions, etc.


Signal and backgrounds in S1

14

S1 [PE]0 10 20 30 40 50 60 70 80

]-1

PE)

⋅ d

ay

⋅Ev

ent R

ate

[ (k

g

-1110

-1010

-910

-810

-710

-610

-510

-410Total (ER+NR) backgroundER backgroundNR background from neutrons

νNR background from 2 cm-46 = 2 10σ, 2WIMP, m = 10 GeV/c

2 cm-47 = 2 10σ, 2WIMP, m = 100 GeV/c2 cm-46 = 2 10σ, 2WIMP, m = 1000 GeV/c

-  Electric field in the TPC: 500 V/cm-  Lower threshold of the region of interest: �

3 PE (lowest one used so far by XENON100)

-  Higher end: 70 PE, corresponding on average to 50 keVr �(include >90% of NR spectrum for 100 GeV WIMP)

-  S2 > 8 electrons

ER background is the most relevant one, �apart from the low S1 region where CNNS becomes dominant.


Statistical model

15

•  2D statistical model: �S1 and discrimination.�

•  Profile Likelihood method, �exclusion test statistic, with CLs�(Eur. Phys. J. C71 (2011) 1554).�

•  Systematic uncertainties are included in the model �as nuisance parameters, and profiled out.��

•  The most relevant uncertainty is related to Leff, which rules the light output of nuclear recoils in LXe. Thus it affects, in a correlated way, both the WIMP signal and the NR backgrounds.�We consider also other systematics: Qy, and the overall uncertainty on ER and NR background (assumed 10% and 20%, respectively).�

•  We study the sensitivity bands via generation of MC toy experiments.

§  ER Background events! NR Background events! (NR) Signal events!


XENON1T sensitivity

16

] 2WIMP mass [GeV/c6 10 20 30 100 200 1000 2000 10000

]2W

IMP-

nucl

eon

Cros

s Sec

tion

[cm

49−10

48−10

47−10

46−10

45−10

44−10

43−10

42−10

41−10

40−10

39−10 DAMA/Na

DAMA/ICDMS-Si (2013)

XENON10 (2013)

SuperCDMS (2014) PandaX (2014)DarkSide-50 (2015)

XENON100 (2012) LUX (2013)

y)⋅XENON1T (2 t

y)⋅XENONnT (20 t

Billard et al., Neutrino Discovery limit (2013)

y⋅XENON1T Sensitivity in 2 tExpected limit (90% CL)

expectedσ 2 ± expectedσ 1 ±

With a 2 t y exposure, with XENON1T we’ll reach a sensitivity to spin-independent WIMP-nucleon interactions of 1.6 10-47 cm2 for a 50 GeV/c2 WIMP.

Dotted blue line shows the XENON1T sensitivity assuming a cutoff below 3 keV


Sensitivity VS time

17

Livetime [days]1 10 210 310

]2 [

cmσ

spin

-inde

pend

ent W

IMP-

nucl

eon

-4810

-4710

-4610

-4510

-4410

XENON100 (2012)

LUX (2013)

XENON1T (2 year exposure)

LUX (2013)

LUX (2015)

In less than 10 days we can reach the sensitivity �of the currently running experiments

mχ = 50 GeV/c2


LUX2015 emission model

18

Nuclear Recoil Energy [keV]1 10 210

Phot

on Y

ield

[ph/

keV

]

0

2

4

6

8

10

12

14

modelXENON100

modelLUX2015

Nuclear Recoil Energy [keV]1 10 210

/keV

]-

Elec

tron

Yie

ld [e

0

2

4

6

8

10

12 modelXENON100

modelLUX2015

With the recent LUX measurement (arXiv:1512.03506):•  Larger photon and electron yield, in particular at low energy (below 3

keV),•  Much smaller systematic uncertainties.


XENON1T sensitivity

19

Assuming the LUX2015 emission model

Potential to detect CNNS �from solar neutrinos.

Significant improvement in sensitivity to WIMPs�

at low masses, below 10 GeV/c2.


Summary & Conclusions

20

•  ER and NR backgrounds from the detector materials are reduced to a subdominant level in the inner 1 t FV.�

•  The most relevant background comes from 222Rn in LXe, (assuming a uniform contamination of 10 µBq/kg). Total ER background: 1.8 10-4 (kg day keV)-1.

•  With a 2 t y exposure, with XENON1T we’ll reach a sensitivity to spin-independent WIMP-nucleon interaction of 1.6 10-47 cm2 for a 50 GeV/c2 WIMP.�

•  Good potential to detect for the first time coherent scattering of neutrinos from the Sun, in particular assuming the LUX2015 emission model. Improved sensitivity at low mass WIMPs too.


Backups

21


MC simulation in GEANT4

Diving bell

PTFE pillars

Supports

Field shaping

electrodes

PMTs

PMT bases

PTFE reflector

Copper plate

PTFE bottom plateCathode

22


ER from the materials Extensive screening campaign using Ge gamma spectrometry

and mass spectrometry (ICP-MS)

Screening of PMTs: E. Aprile et al., Eur.Phys.J. C75 (2015) 546, [arXiv:1503.07698]Dedicated paper on screening of XENON1T materials in preparation

23


NR from µ-‐induced n

24

Negligible, thanks to the water shield and Cherenkov muon veto surrounding the detector


NR from CNNS

25

Nuclear Recoil Energy [keV]1 10

]-1

keV

)⋅

y

⋅Ra

te [(

t

5−10

4−10

3−10

2−10

1−10

1

10

210 TotalB8

hepDSNAtm

Total rate above 4 keV: 1.8 10-2 (t y)-1

Total rate above 1 keV: 90 (t y)-1

(Coherent neutrino-�nucleus scattering)


WIMP signal

26

Halo model: v0 = 220 km/s, vSun = 250 km/s, �vesc = 544 km/s, ρ=0.3 GeV/cm3


2d statistical model

27

ER recoils NR recoils

Total

§  2 t∙y exposure !§  mχ =100 GeV/c2!§  μs=50 (enhanced)!§  μbER=1300"§  μbNR=2.3"

§  ER Background events! NR Background events! (NR) Signal events!

S1 distributionsRecoil energy spectra converted into S1 distributions(n° PE detected) through•  Light Yield LY (n° γ produced/keV)•  Charge Yield QY (n° e- produced/keV)•  Detector performance (LCE, PMTs QE and CE)

Y distributionsY is an idealized version of the discrimination parameter log10(S2/S1)§  ER events Gauss(0,1) §  NR events Gauss(-2.58,0.92)

These distributions reproduce the XENON100 ER/NR discrimination and NR acceptance performance

NO discrimination cut on ER events (full data set used)

S1 [PE]0 10 20 30 40 50 60 70 80

]-1

PE)

⋅ d

ay

⋅Ev

ent R

ate

[ (k

g

-1110

-1010

-910

-810

-710

-610

-510

-410Total (ER+NR) backgroundER backgroundNR background from neutrons

νNR background from 2 cm-46 = 2 10σ, 2WIMP, m = 10 GeV/c

2 cm-47 = 2 10σ, 2WIMP, m = 100 GeV/c2 cm-46 = 2 10σ, 2WIMP, m = 1000 GeV/c


Systematic uncertainty on Leff

28

The number of expected events from CNNS and low mass WIMPs is highly affected by Leff variations

μs= μs(t) and μbNR= μbNR(t)""

!

§  The RELATIVE SCINTILLATION EFFICIENCY Leff rules the light output of nuclear recoils in Xenon

§  Leff is the major systematic uncertainty for XENON1T; no direct experimental measurements below 3 keV.

Parameterization of the uncertainty on Leff Gaussian nuisance parameter t

"

S1 SPECTRAL SHAPES of signal and backgrounds change very slightly under different Leff ! We keep fixed the

shape of S1 spectra"!

mχ=100 GeV/c2"

IMPACT OF Leff VARIATIONS

NUMBER of EXPECTED EVENTS in [3,70] PE

CNNS neutrinos behave like 6 GeV/c2 WIMPNeutrons behave like 30 GeV/c2 WIMP

Conservatively extrapolated down

to zero at 1 keV


ProGile likelihood method

29

μs =10"

Modified p-value CLs method!

μs =5"

Likelihood function(unbinned, extended)

Parameter of interest μs !

Exclusion test statisticProfile Likelihood ratio

Compute test statistic P.D.F.under signal hypothesis H!under bkg-only hypothesis H 0

Run hypotheses tests"§  Generated 104 MC toy experiments under H0 §  Rejection test of each signal hypothesis H! (different !s) using med[qμ|Hμ] as observed test statistic qμ

obs"

§  The significance of each test is given by the p-value

The uncertainty on Leff , affecting both WIMP signal and NR background in a correlated way, is included in the likelihood as a nuisance parameter and profiled out.

Backgroundandsensitivity predictionsforXENON1T › dm16 › talks › selvi.pdf · 2016-02-19 ·...

Documents

Transcript of Backgroundandsensitivity predictionsforXENON1T › dm16 › talks › selvi.pdf · 2016-02-19 ·...