Measurement of low energy Electronic Recoil Response and Electronic/Nuclear Recoils

Discrimination in XENON100

Jingqiang Ye, UC San DiegoOn behalf on XENON Collaboration

LIDINE 2017, Sep. 22 - 24, 2017, SLAC

1

Introduction

2

ER,NR

𝑁"#

𝑁$

S1

S2

g1

g2

g1 =S1

Nph, g2 =

S2

Ne• Scintillation signal S1• Charge signal S2• Different S2/S1 for ER/NR

• Primary scintillation gain g1• Secondary scintillation gain g2• g1 is proportional to photon detection efficiency(PDE)

ER

NR

Data Extraction

0

20

40

60

80

100

120

140

160

180

s]µD

rift T

ime

[

0 20 40 60 80 100 120 140 160 180 200 220]2 at Liquid Surface [cm2Detected Radius

Rel

ativ

e Li

ght Y

ield

to C

ente

r0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

2.20

20

40

60

80

100

120

140

160

1800 20 40 60 80 100 120 140 160 180 200 220

FV#1FV#2FV#3FV#4FV#5FV#6FV#7

3

Calibration source & detector condition:• CH3T calibration (<18.6 keVee, ER) • AmBe calibration(NR)• 3 different drift and extraction fields

To compare with different PDE:• 7 sub-FVs(‘small detector’)

• 50% quantile in 𝑅&direction, equal in drift time• Enough statistics in each sub-FV• Avoid strong field distortion in top and bottom region

• No position dependent correction of S1 and S2• Small S1 and S2 variation in each sub-FV(6% for S1,5% for S2)• PDE increases from top part to bottom part

Drift field(V/cm)

Extraction field(kV/cm)

Electron lifetime(us)

Max drift time(us)

Events in sub-FV(10))

CH3T 400 10.0 1470 ± 190 182 43.4

CH3T 167 8.2 390 ± 160 202 11.9

CH3T 100 8.2 590 ± 30 220 8.9

AmBe 400 10.0 1490 ± 100 182 3.5

AmBe 167 8.2 490 ± 130 202 3.6

AmBe 100 8.2 550 ± 60 220 6.5

Detector calibration(g1, g2)

4

E = W · (S1

g1+

S2

g2),W = 13.7eV

S2

E= � g2

g1

S1

E+

g2W

Calibration principle: Doke method

Detector calibration(g1, g2)

Calibration result:• g1 z dependence due to geometry effect• g2 z dependence due to electron lifetime• g1 is consistent under different drift fields• g2 increases with larger extraction field

5

XeIncoming particle Xe*

Xe+

e-

Xe2+

e- e-

S1S2

Fano fluctuationGaussian

𝑁+~ N(E/W, 𝐹𝐸/𝑊� )F=0.059

Recombination fluctuationGaussian

r~ N(<r>,∆𝑟)Electron drift & extraction

Binomial~B(𝑁$, 𝜀4 ∗ 𝜀$6)

Photon detectionPoisson

RecombinationBinomial~ B(𝑁7,r)

r: recombination factor<r>: average recombination fraction(<r> is tuned)∆𝑟: recombination fluctuation(Δ𝑟/<r> is tuned) 6

Excimer productionBinomial

𝑁$6~ B(𝑁+, 𝛼/(1 + 𝛼))

Simulation model

MC-data matching

7

0500

100015002000250030003500

Cou

nts

S1 spectrum (data)15.4%-84.6% credible region

1.5

2.0

2.5

3.0

Log 1

0(S2

/S1) Point estimation MC

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

1.5

2.0

2.5

3.0

Log 1

0(S2

/S1) Data ER band medians (data)

ER band medians (mc)

0

1000

20000<S1<10 S2 spectrum (data)15.4%-84.6% credible region

0

500

1000

150010<S1<20

0

200

400

600

800

Cou

nts20<S1<30

0

100

200

30030<S1<40

0 1000 2000 3000 4000S2[PE]

0

20

40

60

8040<S1<50

100

101

102

Cou

nts

102

103

Cou

nts

Use Binned Maximum Likelihood Estimation(MLE) in Log10(S2/S1) vs S1 spaceto extract ER response

Light yield

8

2 4 6 8 10 12 14

10

20

30

40

50

hnphi/

E[p

h/ke

V]

a) 100 V/cm

Best estimation±1� fitting uncer.Credible regionNEST v0.98LUX @ 105 V/cm

2 4 6 8 10 12 14

b) 167 V/cm

LUX @ 180 V/cm

2 4 6 8 10 12 14

c) 400 V/cm

2 4 6 8 10 12 14

�4�2

024

Unc

.[ph

/keV

]

2 4 6 8 10 12 14Energy[keV]

2 4 6 8 10 12 14

• Lower light yield at higher drift field as expected• Consistent with LUX measurement• Light yield deviates from NEST at high energy, especially at high drift fields

Recombination fluctuation

2 4 6 8 10 12 14

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

�r

d) 100 V/cm

2 4 6 8 10 12 14

e) 167 V/cm

Best estimation±1� fitting uncer.Credible regionLUX @ 180 V/cm

2 4 6 8 10 12 14

f) 400 V/cm

2 4 6 8 10 12 14

�0.03�0.02�0.01

0.000.010.020.03

Abs

.unc

.

2 4 6 8 10 12 14Energy[keV]

2 4 6 8 10 12 14

No significant change observed between different drift fields

9

ER/NR discrimination

10

• Normalize S1 to photons generated to compare ER leakage under different g1• S2 is corrected for electron lifetime

ER/NR discriminationER leakage is smaller at larger g1

11

ER/NR discrimination for different g1 and drift fields

12

• S1 range(100-400 photons), energy range(11-34 keVnr)• ER leakage is smaller at larger g1• No significant difference for ER leakage between 100 V/cm and

400 V/cm drift field

1 g

0.04 0.05 0.06 0.07 0.08 0.09

ER L

eaka

ge F

ract

ion

3−10

2−10

400 V/cm

167 V/cm

100 V/cm

Summary

• Light yield and recombination fluctuation for low energy under three drift field are measured

• Light yield under 100 and 167 V/cm are consistent with LUX measurement

• ER leakage is smaller for larger photon detection efficiency

• No significant difference in ER leakage is observed between 100 V/cm and 400 V/cm drift field

• The paper will be available on arXiv next week

13

Back up

14

• Drift field increases:• ER/NR separation increases• ER band width increases

• g1 increases:• ER/NR separation doesn’t change significantly• ER band width decreases

Recombination factor

0 2 4 6 8 10 12 14 16Energy[keV]

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

hri

100 V/cm167 V/cm400 V/cm

Black: 180 V/cmBlue: 105 V/cm

LUX CollaborationarXiv: 1512.03133

15

16

0500

100015002000250030003500

Cou

nts

S1 spectrum (data)15.4%-84.6% credible region

1.5

2.0

2.5

3.0

Log 1

0(S2

/S1) Point estimation MC

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

1.5

2.0

2.5

3.0

Log 1

0(S2

/S1) Data ER band medians (data)

ER band medians (mc)

0

1000

20000<S1<10 S2 spectrum (data)15.4%-84.6% credible region

0

500

1000

150010<S1<20

0

200

400

600

800

Cou

nts20<S1<30

0

100

200

30030<S1<40

0 1000 2000 3000 4000S2[PE]

0

20

40

60

8040<S1<50

100

101

102

Cou

nts

102

103

Cou

nts

P-value = 0.01

P-value = 0.16

P-value = 0.10

P-value = 0.37

P-value = 0.38