Backgrounds in a one tonne-year dark matter search of XENON1T - Fei Gao Columbia University IDM 2018, Providence July 24, 2018 - CERN Indico
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Backgrounds in a one tonne-year
dark matter search of XENON1T
Fei Gao
Columbia University
IDM 2018, Providence
July 24, 2018
www.xenon1t.org 1Development of XENON program
XENON1T
XENON10 XENON100 XENON1T XENONnT
96cm
● 3.5 t liquid xenon in total
● 2.0t active target
● ~1t after fiducialization
● 248+6 PMTs
M. Schumann (AEC Bern) – XENON 8
2005-2007 2008-2016 2012-2018 2019-2023
25 kg - 15cm drift 161 kg - 30 cm drift 3.2 ton - 1 m drift 8 ton - 1.5 m drift
~10-43 cm2 ~10-45 cm2 ~10-47 cm2 ~10-48 cm2
2Make Dark Matter Search Easy
XENON10 XENON100
XENON1T
3Make Dark Matter Search Easy
• Discrimination
XENON10 power between ER and NR remains constant (determined by
XENON100
LXe properties) —> The total number of ER and NR background we can
tolerate (in ROI) is almost the same:
• O(100) ER events before discrimination
• O(1) NR background
• O(1) Other anomalous backgrounds
• Key is to get lower background for larger detectors
XENON1T
4ER Background Budget
Internal beta backgrounds (Goal): JCAP04(2016)027
• 85Kr: 0.2 ppt Kr/Xe
• 8.9% of 136Xe abundance
• 222Rn: 10 uBq/kg
Dominating backgrounds
In 1 ton FV, gamma bkg from materials is of the
same order as the one from 85Kr
5Background Control via Screening Measurements
• Gamma screening
• Ultra-low background HPGe Detectors
• A few facilities for measurement:
• GEMSE @Freiburg
• UZH: GATOR @ LNGS
• MPIK: Both @LNGS and @MPIK
• Careful material selection based on
screening measurement —> Subdominant
BG in DM search GATOR
• Radon screening
• 222Rn is the dominating ER BG source
• Dedicated system @MPIK to measure
radon emanation rate
• ~10 uBq/kg 222Rn level controlled well.
6Online Krypton / Radon Distillation
• Commercial Xe: 1 ppm - 10 ppb of natKr
• XENON1T sensitivity requires: < 1 ppt
• Solution: 5.5 m distillation column, 6.5 kg/h throughput
>6.4×105 separation, output concentration < 48 ppq (RGMS)
EPJ-C 77, 2017
Evolution of Kr level during online distillation
Rate [t-1 y-1 Fraction
Source
keV-1] [%]
222Rn 56 ± 6 75.0
85Kr 7.7±1.3 10.3
Solar ν 2.5±0.1 3.3
Successfully distilled Materials 8±1 10.7
Kr to 0.66 ppt
136Xe 0.8±0.1 1.1
Total 75±8
222Rn became the dominant BG Measured 82±5
after krypton distillation
(Expectations in 1-12 keV search window, 1t FV,
single scatters, before ER/NR discrimination)
7The Global Picture of ER Background
ROI for WIMP search
up to 11 keV
• Energy reconstructed from anti
correlated S1 and S2. Excellent
XENON Preliminary linearity from keV to MeV
• Best energy resolution measured
with this large LXeTPC ~1.6%
resolution (sigma) at 2.5 MeV
XENON Preliminary
• Good agreement between predicted and
measured background spectrum
• natKr: 0.6 ppt; 214Pb: 10 uBq/kg
• Gammas based on screening measurements
8Nuclear Recoil Backgrounds
Fraction
Source Rate [t-1 y-1]
[%]
Cosmogenic µ-induced neutrons Radiogenic n 0.6 ± 0.1 96.5
significantly reduced by rock CEνNS 0.012 2.0
overburden and muon veto Cosmogenic n < 0.01 < 2.0
Coherent elastic ν-nucleus (Expectations in 4-50 keV search window, 1t FV, single scatters)
scattering, constrained by 8B JCAP04 (2016) 027
neutrino flux and measurements, XENON Preliminary
is an irreducible background at
very low energy (1 keV)
Radiogenic neutrons from (α, n)
reactions and fission from 238U MC- radiogenic neutrons
and 232Th: reduced via careful
materials selection, event
multiplicity and fiducialization
9Neutron Multiple Scatters
• Absolute neutron flux from simulation has large uncertainties!
• Constraining Single Scatter (SS) background via Multiple Scatter (MS) samples
• SS/MS ratio can be calculated with calibration data, so MS constrain can be
independent of absolute flux.
• Expected (6.4±3.2) multiple scatters, observed 9, statistically consistent
• Final SS background constrained by both MC prediction and MS sample.
XENON Preliminary
XENON Preliminary
10Multiple Scatter Waveform
XENON Preliminary
11Surface Background
• 210Pb and 210Po plate-out on PTFE surface produce
events with reduced S2 -> can be mis-reconstructed
into NR signal region
• Suppressed by fiducialization of volume
• Data-driven model derived from surface
event control samples
XENON Preliminary
12Accidental Coincidence Background
A “lone” S1 or S2 signal produced XENON Preliminary
in charge and light insensitive
regions of the TPC may be
accidentally combined to produce
fake events in signal region
XENON Preliminary
Empirical model shows an overall small
rate in the ROI for NRs
• Select unpaired S1/S2 from data
• Randomly pair to form events
• Apply selection conditions from
analysis
• Performance verified with 220Rn data
and background sidebands
13Fiducial Volume Optimization
Optimize FV prior to unblinding to reduce materials and surface background
• FV volume increased from 1 tonne (in SR0 First Result) to 1.3 tonne thanks to
improvements in position reconstruction, including PTFE charge-up and field corrections
• new surface background model allowed inclusion of radius, R, in statistical inference to
maximize useful volume. Analysis space became cS1, cS2b, R and Z
14Background Prediction
Mass 1.3t 1.3t • Reference region is defined as
between NR median and NR -
(S2, S1) Full Reference 2sigma
ER 627±18 1.6±0.3 • ER is the most significant
background and uniformly
Neutron 1.4±0.6 0.8±0.3 distributed in the volume
CENNS 0.05±0.01 0.03±0.01 • Surface background
contributes most in reference
AC 0.47±0.15 0.10±0.03 region, but its impact is
subdominant in inner volume
Surface 106±11 4.8±0.4 • Neutron background is less
BG 735±20 7.4±0.6 than one event, and impact is
further suppressed by position
WIMPs best-fit information
3.6 1.7
(200GeV) • Other background components
Data 739 14 are completely sub-dominant
15Dark Matter Search Results
● Results interpreted with unbinned profile likelihood analysis in cs1, cs2, r space
10 43
101
Normalized
WIMP-nucleon sSI [cm2]
10 44 100
10 1
101 102 103
10 45
WIMP mass [GeV/c2]
(2017)
n d aX-II
Pa
17) rk)
LUX
(20
th is wo
r,
10 46
(1 t⇥y
T
ENON1
X
10 47
101 102 103
WIMP mass [GeV/c2]
● Sensitivity of all current DM
searches are limited by BG and
its fluctuations
● Further reduction and well-
established understanding of
backgrounds in future detectors
like XENONnT is the key to DM
16 discoverySummary
• XENON1T had achieved the projected background level in the one
tonne-year DM search
• A few backgrounds are highlighted in the talk:
• Lowest ER background at (82±5) [t-1 y-1 keV-1], agrees well with projection
• Direct constrain of neutron background via search of neutron’s multiplicity
• Complete model of anomalous background from TPC surfaces and
accidental S1-S2 pileup
• The complete knowledge of all background components are used
in the DM search analysis via the unbinned profile likelihood
analysis. Resulting the best DM sensitivity to date
• Further reduction and complete understand of background in future
detectors like XENONnT is the key for DM discoveries
17You can also read
