SuperCDMS Searches for Low-Mass Particle Dark Matter - Noah Kurinsky Lederman Fellow, Fermi National Accelerator Laboratory ICHEP Virtual ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
SuperCDMS Searches for Low-Mass Particle Dark Matter Noah Kurinsky Lederman Fellow, Fermi National Accelerator Laboratory ICHEP Virtual Conference July 29, 2020
Overview
• Low Mass Dark Matter Technical Challenges
• Recent SuperCDMS Results
- ERDM w/ HVeV (arXiv:2005.14067 submitted to PRD)
- NRDM w/ PD2 (arXiv:2007.14289, submitted to PRL)
- ALP/Dark Photon Constraints from SuperCDMS Soudan (arXiv:1911.11905, Phys. Rev. D 101, 052008)
• Upcoming Results
- HVeV @ 0V
- HVeV @ NEXUS (FNAL)
- PD2 @ CUTE (SNOLAB)
- IMPACT @ TUNL (Nuclear recoil yield calibration)
• APS Talk Video
• Future Plans
- Pre-SNOLAB Science
- First Science from SuperCDMS SNOLAB
2 7/29/2020 Noah Kurinsky
What Drives Low-Mass Dark Matter Detector Reach? 3 7/29/2020 Noah Kurinsky
What Drives Low-Mass Dark Matter Detector Reach? 4 7/29/2020 Noah Kurinsky
Collision Kinematics
• Recoil energy for a typical dark matter
particle velocity depends on target
mass and recoil type
• Electron and nuclear recoils have
different kinematics; nuclear recoils are Low Mass High Mass
simple elastic collisions, electron
recoils are largely inelastic and depend
on electron orbital and kinematics
within the bound electron-atom system
• In addition to momentum transfer for a
fixed velocity, using a velocity and
angular distribution yields an expected
energy spectrum p
2mT E
✓ ◆2 m ,N R
1 2 mN v 2
2m v
EN R qmax = 2 E
2mN 2 m + mN
✓ ◆ m ,ER
EER
1 2
µN v =
mN v 2
m v2
2 2 m + mN
5 7/29/2020 Noah Kurinsky
SuperCDMS Athermal Phonon Sensors
• Crystal targets made of Si/Ge
• In any recoil event, all energy eventually returns to the
phonon system
• Prompt phonons produced by interaction with nuclei
• Indirect-gap phonons produced by charge carriers reaching band
minima
• Recombination phonons produced when charge carriers drop
back below the band-gap
• Phonons are also produced when charges are drifted in an
electric field; makes sense by energy conservation alone
• Total phonon energy is initial recoil energy plus Luke (NTL)
phonon energy, as shown at right
Ephonon = Erecoil + V ⇤ neh
✓ ◆
y(Erecoil )
= Erecoil 1 + V ⇤
"eh
• Athermal phonons collected in superconducting aluminum
fins and channeled into Tungsten TES, effectively
decoupling crystal heat capacity from calorimeter (transition
edge sensor) heat capacity
Romani et. al. 2017 (https://arxiv.org/abs/1710.09335 )
6 7/29/2020 Noah Kurinsky
Science Leading to SuperCDMS SNOLAB
Electron Recoil (Charge Readout) Nuclear Recoil (Heat Readout)
Lower Backgrounds, Larger Exposure
Surface R&D (Lower Threshold, Improve Reconstruction)
(Stanford, Northwestern, Berkeley, Texas A&M)
NEXUS @ FNAL
CUTE @ SNOLAB
SuperCDMS SNOBOX
@ SNOLAB
7 7/29/2020 Noah Kurinsky
Science Leading to SuperCDMS SNOLAB First Results Published
Electron Recoil (Charge Readout) Nuclear Recoil (Heat Readout)
Lower Backgrounds, Larger Exposure
Surface R&D (Lower Threshold, Improve Reconstruction)
(Stanford, Northwestern, Berkeley, Texas A&M)
Currently Running
NEXUS @ FNAL
CUTE @ SNOLAB
Installation
In Progress SuperCDMS SNOBOX
@ SNOLAB
8 7/29/2020 Noah Kurinsky
Science Leading to SuperCDMS SNOLAB
Electron Recoil (Charge Readout) Nuclear Recoil (Heat Readout)
Lower Backgrounds, Larger Exposure
Surface R&D (Lower Threshold, Improve Reconstruction)
(Stanford, Northwestern, Berkeley, Texas A&M)
NEXUS @ FNAL
CUTE @ SNOLAB
SuperCDMS SNOBOX
@ SNOLAB
9 7/29/2020 Noah Kurinsky
Science Leading to SuperCDMS SNOLAB
Electron Recoil (Charge Readout) Nuclear Recoil (Heat Readout)
Lower Backgrounds, Larger Exposure
Surface R&D (Lower Threshold, Improve Reconstruction)
(Stanford, Northwestern, Berkeley, Texas A&M)
NEXUS @ FNAL
CUTE @ SNOLAB
SuperCDMS SNOBOX
@ SNOLAB
10 7/29/2020 Noah KurinskyElectron-Recoil Searches
• SuperCDMS HVeV
- gram-scale, single-charge resolving detectors
- Run 1 - 2018
- Run 2 - 2020
- Run 3 - 2021 (underground, lower backgrounds)
• Long-term: HV detectors @ SNOLAB
• Challenges
- Charge leakage
- Understanding quantized backgrounds
- Maximizing charge collection efficiency
- Removing surface backgrounds
11 7/29/2020 Noah KurinskyExperimental Setup (Run 1 @ Stanford) 12 7/29/2020 Noah Kurinsky
HVeV v1 0.5 Gram-Day Science Spectrum • Ran the detector for ~12 hours with SuperCDMS Collaboration 2018 (https://arxiv.org/abs/1804.00088 ) a 1 Hz laser calibration (black line) • Background consistent with infrared photon background at low energy, high-energy tail was not anticipated • Device had 10-14 eV resolution, and the ability to distinguish between ‘true’ events and impurity- mediated events (unlike CCDs which only measure electrons or holes) • Very simple analysis; it’s easy to see how one rules out a quantized signal in light of this background 13 7/29/2020 Noah Kurinsky
Experimental Setup (Run 2 @ Northwestern)
V. Novati, APS April 2020
14 7/29/2020 Noah KurinskySuperCDMS Collaboration 2020 (arXiv:2005.14067)
HVeV Run 2 Supplemental Plots
• HVeV second run taken with 3 eV resolution detector over the course of 3
weeks:
- 60V and 100V spectra show identical backgrounds; signal seen not voltage dependent
- Different prototype, run in a different lab, in a different state
- 0V data acquired with ~12 eV threshold, results still being analyzed
- Rates in every charge bin consistent with Run 1…that is completely unexpected
15 7/29/2020 Noah KurinskySuperCDMS Collaboration 2020 (arXiv:2005.14067)
HVeV Run 2 Results Supplemental Plots
V. Novati, APS April 2020
16 7/29/2020 Noah KurinskyLow-Mass Nuclear Recoil Searches
• Phonon readout
- SuperCDMS CPD (10g, operated at SLAC) - 2020
- HVeV @ 0V (1g, operated at NU) - 2020 (in prep)
- SNOLAB detectors and CPD at CUTE
• Challenges
- Unexplained low-energy excesses persist
- Detectors w/ thresholds below 10 eV highly sensitive to environmental noise
- Science reach driven by threshold and trigger accuracy
- Few calibration points below 1 keV
17 7/29/2020 Noah KurinskyDesign of a Large Area Photon Detector
Design of a Large Area Photon Detector
• The detector is a CDMS-style athermal phonon
sensor
• The detector is a CDMS-style athermal phonon
- 1 mm thick silicon wafer, 45.6 cm^2 surface area
sensor
- Mass ofmm
– 1 10.6 grams
thick silicon wafer, 45.6 cm# surface area
– Mass of 10.6 grams
• The
• The device
device hashasbeen
been optimized
optimized for forphoton
photondetection
– Distributed athermal sensor array read out by TESs
detection
– Single distributed channel gives a fast collection time
- Distributed athermal sensor array read out by TESs
of athermal phonons
- Single distributed
– This channel penalties
reduces efficiency gives a fast collection
due to athermal time
of athermal
phononphonons
down conversion
$% = 41.5 mK,
- This–reduces lowering
efficiency the expected
penalties dueenergy resolution
to athermal
• Designed
phonon downoriginally
conversionfor degraded alpha rejection in
neutrinoless
- T_c=41.5 double beta
mK, lowering decay andenergy
the expected for an active
photon veto for dark matter experiments
resolution
• Designed originally for degraded alpha rejection
in neutrinoless double beta decay and for an
22 July 2019 LTD-18, Milano 4
active photon veto for dark matter experiments
https://agenda.infn.it/event/15448/contributions/95785/
18 7/29/2020 Noah KurinskyarXiv:2004.10257
Reducing Readout Noise
• Readout noise limited by thermal
fluctuation shot noise between sensor
and crystal
- Can reduce this only by weakening the link
between the sensor and absorber, or
dropping the sensor temperature
- Power noise at the level of 1e-17 aW/sqrt(Hz)
- Hitting readout limitations
• Power to current gain impacted by total
heat capacity of the sensor
- Make smaller sensors
• Massive on-sensor multiplexing
- Readout smaller sections of the detector
- Use MKIDs or other naturally multiplexed
sensors rather than DC TES
arXiv:2007.14289
19 7/29/2020 Noah KurinskyCPD Dark Matter Search Results arXiv:2007.14289
• Detector threshold (online FPGA-based trigger) of 20 eV
• World-leading limits below ~150 MeV
- First NR result withFuture Prospects • Following up HVeV and CPD at underground facilities • Preparing for SuperCDMS SNOLAB 21 7/29/2020 Noah Kurinsky
Pre-SNOLAB Science
• HVeV @ NEXUS
- Larger exposures
- Understanding leakage currents and low-
energy backgrounds
- NR/ER calibrations NEXUS Sensitivity Study
- New readout techniques (e.g. mKIDs)
Cryogenic Underground TEst Facility (CUTE) @ SNOLAB
• NR Searches @ CUTE
- 10g R&D Detectors withPlans ER/NR for Follow-up
• 1 gram-day of 0V data taken
- Determine whether background is NR,
ER, or detector-related
- Continuous readout will allow NR limit
with 9eV energy threshold
- Coming late 2020
• More data taking at NEXUS
- Currently operating underground,
constructing lead shield
- 3 eV resolution maintained in new
facility
• Plans for larger payload,
additional studies w/ high-energy
gamma sources, different detector Preliminary
packaging
23 7/29/2020 Noah KurinskyNEXUS: Underground Experimental Site for R&D
• Northwestern EXperimental Underground Site at
Fermilab (NEXUS@FNAL)
- Underground cryogenic detector testing facility in
class 10,000 clean room
- 106 m (300 mwe) depth + lead shielding (in progress)
- expectedNEXUS: Underground Experimental Site for R&D
• Northwestern EXperimental Underground Site at
Fermilab (NEXUS@FNAL)
- Underground cryogenic detector testing facility in
class 10,000 clean room
- 106 m (300 mwe) depth + lead shielding (in progress)
- expectedFrom APS April Meeting SuperCDMS SNOLAB Detectors 26 7/29/2020 Noah Kurinsky
From APS April Meeting SuperCDMS SNOLAB Layout 27 7/29/2020 Noah Kurinsky
SuperCDMS SNOLAB Science Targets 28 7/29/2020 Noah Kurinsky
SuperCDMS SNOLAB Science Targets 29 7/29/2020 Noah Kurinsky
SuperCDMS SNOLAB Science Targets
0V arrays of small to
medium-sized detectors
High Voltage
NTL-based
Phonon + Ionization
30 7/29/2020 Noah KurinskySummary
• SuperCDMS Detectors are seeing significant backgrounds at low energy
- Comparison between HV and 0V will elucidate source
- Runs at NEXUS/CUTE underway will gauge depth/background dependence
- See Yoni Kahn’s talk for a phenomenological take on this question
- See Lauren Hsu’s talk next week for a broader overview of light DM detection
• Upcoming results with ~9 eV threshold will probe new DM masses below 100
MeV, and provide a direct comparison with Run 2 results
• SuperCDMS SNOLAB detectors will substantially improve exposure and
benefit from these early background studies
• Many calibration programs underway to understand nuclear recoils in theThanks! 32 7/29/2020 Noah Kurinsky
Questions?
• Zoom info:
- Topic: Noah Kurinsky ICHEP Questions
- Time: Jul 29, 2020 01:00 PM Central Time (US and Canada)
- Link
- Meeting ID: 991 4345 0731
- Passcode: 688639
33 7/29/2020 Noah KurinskyBackup 34 7/29/2020 Noah Kurinsky
HVeV v1.5: Edge-Dominated Leakage
ArXiv:1903.06517
• Prototype demonstrated position
dependence in the non-quantized data
hinted at during HVeV Run 1
• Nearly contact-free biasing scheme
isolates contact along the crystal edge,
preventing charge tunneling through
most of the high-voltage face
• Surface events have a distinct pulse
shape and can be removed using a cut
in the pulse-shape plane.
• Non-quantized leakage is dominant at
high radius; 95% of non-quantized
events removed by 50% radial cut
efficiency. 80% of quantized events
removed by the same cut
35 7/29/2020 Noah KurinskyHVeV v1.5: Edge-Dominated Leakage
ArXiv:1903.06517
• Prototype demonstrated position
dependence in the non-quantized data
hinted at during HVeV Run 1
ArXiv:1903.06517
• Nearly contact-free biasing scheme
isolates contact along the crystal edge,
preventing charge tunneling through
most of the high-voltage face
• Surface events have a distinct pulse
shape and can be removed using a cut
in the pulse-shape plane.
• Non-quantized leakage is dominant at
high radius; 95% of non-quantized
events removed by 50% radial cut
efficiency. 80% of quantized events
removed by the same cut
36 7/29/2020 Noah KurinskyHVeV v2: Combining Lessons
https://agenda.infn.it/event/15448/contributions/95710/
• Maintain 3 eV resolution, scale back to 1g Device paper in prep
- High coverage for higher dynamic range
• Still seeing some issues with sidewall trapping
and incomplete neutralization
- We observe ~15% trapping
• Physics results from Nuclear Recoil calibration at
TUNL coming soon!
37 7/29/2020 Noah KurinskyR&D ‘HVeV’ Prototype Progress HVeV v2 (NF-C)
HVeV v1 HVeV v1.5
Device: In Prep
ERDM:
Device: https://arxiv.org/abs/1710.09335 Device: https://arxiv.org/abs/1903.06517 NRDM: In Prep
DM: https://arxiv.org/abs/1804.10697
• 3 eV Resolution • 3 eV Resolution
• 10 eV Resolution • 0.06 electron-hole pairs • 30% energy efficiency
• 1 gram mass • 0.25 gram mass, contact-free • 1 gram mass, backside
design contact
• No position resolution
• High position resolution • High position resolution
• ~1.2 Ohm Resistance • ~400-900 mOhm Resistance • ~300 mOhm Resistance
• ~55 mK Tc • 65 mK Tc • 65 mK Tc
• Amorphous Layer • No Amorphous Layer
• No Amorphous Layer
38 7/29/2020 Noah KurinskyScaling Up in Mass 39 7/29/2020 Noah Kurinsky
Scaling Up in Mass
Faster Signal
Lower Sensor
Noise
40 7/29/2020 Noah KurinskyScaling Up in Mass
Faster Signal
Large-Scale Multiplexing
Lower Sensor
Noise
Sets Operating Voltage for NTL Single-Charge Readout
41 7/29/2020 Noah KurinskyDetector R&D Groups
• HVeV (1g Detector Chips)
- Caltech: Sunil Golwala, Yen-Yung Chang, Taylor Aralis,
Osmond Wen
- Fermilab: Noah Kurinsky, Dan Bauer
- Northwestern: Enectali Figueroa-Feliciano, Ziqing
Hong, Tom Ren, Ran Chen
- Stanford/SLAC: Blas Cabrera, Betty Young, Francisco
Ponce, Chris Stanford, To Chin Yu
- University of Minnesota: Matt Fritts, Nick Mast
- University of Florida: Tarek Saab, Corey Bathurst,
Tyler Reynolds
• Large Area Photon Detectors
- UC Berkeley: J. Camilleri, C. Fink, Y. Kolomensky, M.
Pyle, B. Sadoulet, B. Serfass, S. Watkins
- Texas A&M: Nader Mirabolfathi, Rupak Mahapatra,
Fedja Kadribasic
• QET Fabrication R&D
- Mark Platt (TAMU), Paul Brink (SLAC)
42 7/29/2020 Noah KurinskyHVeV v2 1cm Designs
• Best for NR
- UCB A/C - low coverage, single
channel
NF-H
- 600 mOhm normal state resistance
• Best for ERDM UCB-C
- Optimized for baseline resolution with
varying levels of Al coverage
- 900 mOhm Rn - NF F, G, H
- 300 mOhm Rn - NF A, D, E NF-C
• NF-A is roughly the same as QP.4, i.e.
AR64
• Best for Calibrations
- Sacrifice baseline resolution for higher
dynamic range
- 300 mOhm normal state - NF-B/C
https://confluence.slac.stanford.edu/display/CDMS/HVeV+v2+Design
43 7/29/2020 Noah KurinskyNEXUS Si/Ge Dark Matter Search Timeline
• Spring-Fall 2019 (ADR Demonstrator): 1 gram
- 1 gram, 4 eV resolution (20 eV threshold)
- 0.01 electron-hole pair resolution (Future for HVeV Program
• Mounting issues damaged
some prototypes (X’s)
• 2 1g v2 detectors, optimized
X
for dynamic range, now
successfully operated X X
• 2 4g detectors currently being
mounted (both optimized for
low energy resolution)
• Many more prototypes to help
test the resolution and
efficiency models that go into
this detector design X X
- Multiple designs with sub-eV
projected resolutions
45 7/29/2020 Noah KurinskyBroad Picture: Detection Media for Lower Mass
US Cosmic Visions, arXiv:1707.04591
• Looking at current/future technologies, we have a mix of materials with different work functions, to limit dark
counts and maximize energy to carrier conversion
• Extent of these arrows driven by fundamental limitations from kinematics and material properties, and
assumes large current hurdles can be overcome in energy and charge noise across all experiments
• June 2019 workshop at FNAL explored much more than cosmic visions: https://astro.fnal.gov/ldm/
46 7/29/2020 Noah KurinskyYou can also read
