1
SuperCDMS Soudan Ray Bunker
for the SuperCDMS Collabora4on
Ray Bunker - SuperCDMS
Detec4on Technique
2
Nuclear Recoil (NR) yields ≈20–30% ioniza<on vs. Electron Recoil (ER)
holes+
e-‐
Nuclear recoil (NR)
Electron recoil (ER)
Quasipar4cle trap Aluminum
collector
Ge or Si crystal
Diffusion
Phonons
Cooper pair
tungsten Transi4on
Edge Sensor (TES)
Phonons provide equal measure of NR & ER
recoil energies
Ioniza4on & Phonons
Ray Bunker - SuperCDMS 3
Detec4on Technique
Ioniza4on & Phonons
Ioniza4o
n yield
Recoil energy (keV)
γ-‐calibra4on ERs
Neutron-‐calibra4on NR
s
detector
copp
er hou
sing
133Ba γ
adjacent detector
252Cf neutron
Residual backgrounds are near surfaces: • U/Th/K decays à γ-‐induced electrons, sidewall γ’s • 210Pb decay chain à betas, x-‐rays, 206Pb recoils
Low-‐energy region, all of the above plus: • U/Th/K decays à γ-‐induced bulk ERs • Internal ac<va<on (e.g., 68Ge) à bulk ERs
Surface events
γ
γ
e-‐
γ
e-‐
210Pb
x-‐ray
210Bi β
210Po 206Pb
Low-‐ene
rgy
region
γ
γ γ
e–
γ
Ray Bunker - SuperCDMS 4
The SuperCDMS iZIP Bulk event à Side-‐symmetric ioniza<on signal
Electrons
Holes
Surface event à Asymmetric ioniza<on signal
Electrons Holes
Significantly improved surface-‐event rejec4on
Phonon sensors @ 0V
Ioniza<on electrodes @ ±2V
Interleaved ioniza<on & phonon sensors
5
SuperCDMS Soudan
Tower 3 Sources
15 detector array of germanium iZIPs:
• 0.6 kg per detector • in situ 210Pb sources • Installed in CDMS II shielding • In opera<on since early 2012 • Primary iZIP science run ended summer 2014 • More CDMSlite & calibra<ons through fall 2015
CDMS II Shielding Broad science program:
• Demonstra<on of surface-‐event discrimina<on • WIMP-‐search analyses:
• 3 CDMSlite runs à MWIMP < 5 GeV/c2 • Low-‐threshold iZIP à MWIMP = 5–10 GeV/c2 • High-‐threshold iZIP à MWIMP > 10 GeV/c2
• Y/Be & Sb/Be photoneutron calibra<ons • In general, high degree of experimenta<on to inform SuperCDMS SNOLAB
see talk by Sunil Golwala (next session)
Ray Bunker - SuperCDMS 6
Surface-‐event Discrimina4on 210Pb sources placed above and below 2 detector faces:
• 50 live days à 0 of 132,968 leaked surface events in (symmetric) NR signal region • More than sufficient face-‐event discrimina<on for SNOLAB iZIP payload
Ioniza<on-‐derived side-‐symmetry parameter
Ioniza<o
n yield
Recoil energy (keV)
Ioniza<on yield
1.2
1
0.8
0.6
0.4
0.2
0
Bulk ERs
Bulk ERs Surface Betas & X-‐rays
Bulk NRs
Bulk NRs
206Pb Recoils
APL 103, 164105 (2013)
Ray Bunker - SuperCDMS 7
Low-‐threshold Analysis
Tower 3 Sources
Low-‐energy WIMP search:
• Use 7 iZIPs w/ best low-‐energy performance • Data period = Oct 2012 – July 2013 • 577 kg-‐day exposure (2.4x CDMS II) • Range of thresholds: 1.6–5 keV • Improved discrimina<on of sidewall events via phonon-‐derived fiducial volume
Lowest-‐energy background >10x lower than CDMS II,
dominated by sidewall events
PRL 112, 241302 (2014)
CDMS II Si
DAMA Further LT studies:
• Maximum likelihood analysis using background model from 2014 PRL
• Understand anomalous high-‐energy candidates using Detector Monte Carlo à Stay tuned… PRD long paper in progress
Anomalous high-‐energy candidates from detector w/ shorted charge channel
8
High-‐threshold Analysis
Tower 3 Sources
High-‐energy WIMP search:
• Different strategy: target high-‐mass WIMPs • Higher thresholds to prevent background at low energy due to resolu<on smearing • Use 10 iZIPs w/ fully opera<onal charge channels • Mix of WIMP search and γ & neutron calibra<ons
Mar 2012 Sep 2012 Apr 2013 Nov 2013 May 2014
Accumulated
Live Time (days)
0
100
200
300
400
500
Nov 2014 Jun 2015
Blinded HT WIMP-‐search exposure
252 Cf, ne
utron calib
ra4o
ns
133Ba gamma calibra4on (≈daily)
CDMSlite
Run
1
Low-‐threshold WIMP-‐search exposure
CDMSlite
Run
2a
CDMSlite
Run
2b
Extend
ed γ calibra4o
n
CDMSlite
Run
3
Photon
eutron
calibra4o
n
9
High-‐threshold Analysis Analysis challenges: • Exposure-‐limited à maximize fiducial volume with near-‐zero background • Surface events (γ-‐induced electrons & 210Pb decay chain):
• Faces: expect a few hundred such events à demonstrated ≈10-‐5 rejec<on more than sufficient! • Sidewalls: more challenging to understand low-‐yield event rate vs. radial-‐posi<on es<mator (ongoing)
Ray Bunker - SuperCDMS
Detector-‐response Simula<on
True radius squared (cm2) 0 4 8 12 2 6 10
0.0
0.5
1.0
1.5
2.0
2.5
True
z po
si<on
(cm)
Reconstructed radial param
eter
1.0
0.75
0.45
0.15
0.0
0.9
0.6
0.3
14
Detector-‐response Simula<on
True radius squared (cm2) 0 4 8 12 2 6 10
0.0
0.5
1.0
1.5
2.0
2.5
True
z po
si<on
(cm)
Reconstructed z p
aram
eter
1.0
0.5
0.0
-‐0.5
-‐1.0
14
Ray Bunker - SuperCDMS 10
SuperCDMS HT Projected Sensitivity
High-‐threshold Analysis
Projected sensi<vity for:
• 1700 kg-‐day exposure (≈2.75x CDMS II)
• 7-‐15 keV thresholds
• 50-‐65% fiducial volume
Nice improvement vs. CDMS II considering modest exposure
Important demonstra<on of iZIP technology in advance of SuperCDMS SNOLAB
Ray Bunker - SuperCDMS 11
Background Measurements Global γ background:
• Geant4 simula<on of U/Th/K in materials of construc<on • Simultaneous fit for ac<vi<es using measured ioniza<on spectrum • Mystery: dominated by U/Th gammas coming from below! • Auemp<ng to understand with forensic assay à e.g. dust?
Surface α background:
• Alphas are easily iden<fied! • Primary contribu<on from 210Po at 5.3 MeV à Normalizes 210Pb surface contamina<on
• Decay rate indicates more Po vs. Pb at start • Ge crystal faces show ≈50 nBq/cm2
• >10x higher rate at sidewalls à Copper housings or crystal sidewalls? à Planning to directly α assay both!
Bouom iZIP from 4th Tower
Measured spectrum Best-‐fit simula<on
500 1000 1500 2000 2500 3000 0 Ioniza<on energy (keV)
10-‐5
10-‐4
10-‐3
10-‐2
Normalize
d even
t rate All selected events
Alpha selec<on
2000 4000 6000 0 8000
1000
500
2000
1500
3000
2500
Middle iZIP from 2nd Tower
Recoil energy (keV)
Ioniza<o
n en
ergy (keV
)
Fully contained at sidewall
fully contained at faces
12
Photoneutron Calibra4on Purpose:
• Improve understanding of nuclear-‐recoil energy scale for CDMSlite & iZIP modes
Technique: • Pair 88Y and 124Sb γ sources with Be wafer à ≈mono-‐energe<c neutrons (152 & 23 keV) • Look for maximum elas<c scauering shoulder off Ge at 8.1 & 1.3 keV
Approxim
ate trigger
threshold
elas4c scag
er
endp
oint
103
102
101
100
Even
t rate (arb.)
0 1 2 3 4 5 6 7
Detected recoil energy (keV)
Geant4 simula<on with approximate
shielding geometry
Pb γ shield
Outer poly (40 cm)
Lead shield (22cm)
Inner poly (10cm)
Copper vacuum cans
Detectors Source
Energy (a.u.)
Even
ts per hou
r
PRELIMINARY
124Sb 124Sb9Be
Rates match in tails where no
neutrons expected
Excess due to neutrons
CDMSlite mode
Ray Bunker - SuperCDMS 13
Summary
January 2016 Collabora4on Mee4ng at PNNL
• SuperCDMS Soudan is a rich source of data for improving WIMP-‐search sensi<vity and understanding SuperCDMS detectors
• The iZIP technology significantly improves rejec<on of the CDMS II limi<ng backgrounds • e.g., demonstrated face-‐event rejec<on beuer than needed for SuperCDMS SNOLAB & improved phonon-‐sensor layout enables >10x reduc<on of low-‐energy backgrounds
• High-‐threshold analysis is ongoing (expect result this year) + backgrounds & calibra<ons
Ray Bunker - SuperCDMS
Backup Slides
14
15
CDMS II ! Background-‐Limited Ioniza<o
n yield
0.6
0.4
0.2
0.0
-‐0.2
40
30
20
10
0
Normalize
d ioniza<o
n yield
10 15 5 0 -‐5 Normalized phonon <ming parameter
-‐10
100 10 2 Recoil energy (keV)
PRL 106, 131302 (2011)
Science 327, 1619 (2010) High-‐threshold analysis:
• Final CDMS II dataset • 612 kg-‐day Ge exposure • 10 keV threshold • Limited by near-‐face ERs in high-‐energy region
Bulk ERs
Surface ERs
Signal Region
Low-‐threshold analysis:
• Final CDMS II dataset • 241 kg-‐day Ge exposure • 2 keV threshold • Limited by sidewall events in low-‐energy region
Bulk ERs
Near-‐face ERs
Sidewall Events
Signal Region
Ray Bunker - SuperCDMS 16
LT Analysis – Detector Monte Carlo
True radius (cm)
True
z po
si<on
(cm)
Fully Func<onal Detector
Reconstructed Ch
arge Radius (top)
True radius (cm)
True
z po
si<on
(cm)
Shorted-‐Channel Detector
Reconstructed Ch
arge Radius (top)
Charged-‐based radial preselec<on not effec<ve on top side
PRELIMINARY
PRELIMINARY
Ray Bunker - SuperCDMS 17
LT Analysis – Detector Monte Carlo
True radius (cm)
True
z po
si<on
(cm)
Fully Func<onal Detector
Reconstructed Ph
onon
Radius
Phonon-‐based radial fiducializa<on less effec<ve near top corner
True radius (cm)
True
z po
si<on
(cm)
Shorted-‐Channel Detector
Reconstructed Ph
onon
Radius
Slide courtesy Lauren Hsu (FNAL) 18
Photoneutron Calibra4on
lead shield (22cm) x-‐axis units are not calibrated…but they are the same scale shown on previous slide
arbitrary energy units arbitrary energy units
Even
ts per hou
r
Even
ts per hou
r
88Y9Be subtracted
124Sb9Be subtracted
Plots shown here are result axer subtrac<ng neutron-‐off from neutron-‐on spectra; these plots are not efficiency corrected
approxim
ate trigger thresho
ld
approxim
ate trigger thresho
ld
error bars shown are sta=s=cal; data shown is 5 weeks of Sb and 4 weeks of Y (addi=onal data not yet analyzed)