SubGeV Searches with EDELWEISS

The EDELWEISS SubGeV program

SIMPs search results [PRD 99 082013 (2019)]

New: search results for light DM interacting with electrons

J. Gascon UCB Lyon 1, CNRS/IN2P3/IP2I

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches EDELWEISS-SubGeV: Scientific context A widening search domain eV keV MeV GeV TeV

Absorption DM-electron scattering DM-Nucleus scattering Electronic recoil Electronic recoil Nuclear recoil

Hidden sector Dark Matter and others Standard WIMP

New theories, New 8B neutrinos (~ 6 GeV) experiments... Reactor neutrinos (~ 2.7 GeV) EDELWEISS-SubGeV program EDELWEISS-III

Not competitive with High Voltage Low Voltage High Voltage Low Voltage noble gases experiments single e/h Part. ID + Fid single e/h Part. ID + Fid

J. Billard (IPNL) J. Billard, New Directions in the Search for Particles, Fermilab 2019 2

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 2 This talk

EDELWEISS EDELWEISS-Surf Electron-DM search SIMP results

Evolution of the detectors from EDELWEISS- III to EDELWEISS-SubGeV

EDELWEISS-III

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 3 EDELWEISS-III detectors

210 210 [JINST 12 (2017) P08010]Pb β Pb β 210 210 206 Po α 206 Po α Heat: DT = E/Ccal + : Npairs = E/eg orPb en Pb <2x10-5 rejection of electron/nuclear recoils

Readout of all electrodes provides a <4x10-5 surface event rejection. yield

18000 kg.day equivalent 18000 kg.day equivalent 210Pb β Ionization 210Po 206Pb α

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 4

18000 kg.day equivalent EDELWEISS SubGeV program

n Current+future projects: background limited n Event-by-event identification of nuclear recoils down to 1 GeV/c2 ( and reach 10-43 cm2 with kg-size array) requires

sphonon = 10 eV and sion = 20 eVee n Ionization resolution is key to particle identification + surface rejection : Cold HEMT preamp + low capacitance (joint development with Ricochet)

n Reducing the detector mass from 860 to 33 g is key to meet these resolution goals 1st milestone: EDELWEISS-Surf

n Keeping the ability to apply HV to amplify heat signal: low thresholds in ER searches & statistical NR/ER separation 1st milestone: Electron-DM results

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 5 7

1 2 Toward 20 eVee ionization resolution 3 4 n Transition from JFET to HEMT Goal #3: EM background[ rejectionnew arXiv:1909.02879 of O(103) ] FET to 5 n Lower intrinsic noise + reduce cabling HEMT 20 eV ionization resolution: HEMT preamplifiers + new electrode design 6 capacitance by working at 1K or 4K n Data driven HEMT models show that the 7 FET to HEMT goal of 20 eVee is reachable with ~20 pF 8 Goal Goal 9 total input impedance n Ongoing HEMT characterizations • As initiated by the CDMS-Berkeley group10 (arXiv:1611.09712) we are transitioning to HEMT based preamplifiers. • HEMT have lower intrinsic noise than11 JFET • Work @ 4/1 K allowing to reduce the stray capacitance • Based on our data driven HEMT model12, O (10) eV rms reachable with ~20 pF total input impedance • HEMT characterizations are ongoing 13 • First HEMT-based preamp to be tested in winter 2019 ! • Cryogenic and cold cabling challenges14 ahead ! Left: Contribution to the ionization resolution of the voltage noise, current noise, bias • Synergie with the Ricochet-CryoCube collaboration Fig. 5 J. Billard (IPNL) 8 15 n HEMTresistor-based (10 preamp GW @ tests 20mK). end The of 2019 total detector + cabling capacitance is 20 pF. 5 pF and 100 pF geometries have been studied. FFT of 1 keV event are shown. Right: Evolution of the resolution 16 n Cryogenics + cabling challenges ahead with the detector + cabling capacitance for the 5 HEMT geometries and the IF1320 Si-JFET 17 n Workfrom done InterFET. in synergy Noise ofwith the biasthe resistor is included.Optimization (Color figureof 33g online.) FID design: large 18 Ricochet-CryoCube collaboration fiducial volume & low capacitance 19 Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 6 20 Ge-NTD Thermal Sensor Heat 21 Veto Top Interdigitized Al Electrodes 22 Fiducial Top 23 24 25

26 (V) Field 27 10mm 28 29 15 mm 30 31 Fig. 6 Electrostatic simulation of a Full Inter-Digitized electrodes scheme on a 38 g 32 crystal (F = 30 g, h = 10mm). The crystal is surrounded at 2 mm distance by a chassis connected 33 to the ground (not shown). The capacitance of the 4 electrodes with respect to the ground is about 20 pF (Color figure online.) 34 35 36 37 38 tector. Electrostatic simulations (Fig. 6) show that the EDELWEISS Fully Inter- 39 Digitized (FID) 800g detector [1] ionization electrode pattern can be adapted to 40 match the 20 pF capacitance requirement with a 30 g detector. This FID detec- ⇠ 41 tor geometry has been shown to efficiently reject surface events, which may be 42 advantageous for CEnNS and low-mass WIMP searches where the low energy 43 backgrounds are unknown. Future work will address the issues of very low ca- 44 pacitance cabling and optimization the balance between individual detector mass, 45 fidicual volume, and detector capacitance when considering the 10 eV heat and 46 20 eVee ionization baseline RMS resolutions required to the upcoming Ricochet 47 and EDELWEISS-SubGeV experiments. The development of HEMT-based am- 48 plifiers represents a clear path to achieving the energy resolution requirements of 49 cryogenic semiconductor detectors for pursuing new physics in the low mass Dark 50 Matter and CEnNS sectors. 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Searching for low-mass dark matter particles with a massive Ge bolometer operated above-ground, Phys. Rev. D 99 (2019) 082003

ABOVE-GROUND SEARCHES WITH A 33G DETECTOR

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 7 Theory & experiment motivations for DM surface searches n Relevance of strong interactions of ~GeV DM particles

• Main focus of direct DM searches so far: DM-nucleon cross sections below 10-31 cm2 Shielding from Earth + atmosphere can be neglected, i.e. experiments are located in deep underground sites, to reduce cosmic-ray induced backgrounds • O(10-24) cm2 DM-DM cross-section of ~GeV DM particles could actually help CDM problems at small-scale (DM halo, satellites...) [Spergel+Steinhardt PRL 84 3760 (2000)] • Natural extension: test for O(10-24) cm2 DM-nucleon interactions [e.g. Chen et al, PRD 65 123515 (2002)] n Technological:

• First physics with prototype low-threshold 33g Ge detectors

• Detector development program based in surface laboratory

• Proof that relatively massive EDELWEISS-like detectors can be used in surface experiment, i.e. relevant for the study of the coherent elastic scattering of reactor neutrinos on nucleons (like Ricochet)

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 8 EDELWEISS-Surf Above-ground DM search n Context: EDELWEISS and Ricochet common R&D for low-threshold detectors performed in easy- access surface lab @ IPN-Lyon n <1 m overburden: ideal for SIMP search (strongly interacting DM) n Dry cryostat (CryoConcept) with RED20 (33g) <30h cool-down (fast turnover RED11 (200g) ideal for detector R&D) [NIM A858 (2017) 73] n < µg/√Hz vibration levels (spring-suspended tower). [JINST 13 (2018) No.8 T08009] n RED20: 33g Ge with NTD sensor, with no electrodes • No ER/NR discrimination, but no uncertainty due to ionization yield or charge trapping) n 55Fe source for calibration

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 9 EDELWEISS-Surf data n Streamed data processed PSD from 137 h displayed 26 0.65 ]

Hz 24 0.6 with optimum filter 10-7 1 n Stability of noise & 22 0.55

20 0.5 Trigger rate [Hz] resolution over 137h 10-8 10-1 (6 days) of data taking 18 0.45 16 0.4

n -9 -2 Baseline energy resolution [eV]

1 day set aside a priori Linear power spectrum [V/ 10 10 Measured Noise 14 0.35

Signal template Optimal filter transfer function [a.u.] Expected from OF theory for blind search Optimal filter 12 0.3 1 10 102 0 20 40 60 80 100 120 n Baseline: s = 17.8 eV Frequency [Hz] Time [hour] Calibration and Resolution 12 n @5.9 keV: s = 36 eV 7 10 Data Noise induced triggers 6 6 1018 eV RMS10 energy resolutionResidual

105

4 0.4 105 10

Efficiency 0.05 0.1 0.15 0.2

0.3Event rate [evts/kg/keV/day)] 104

0.2

0 1 2 3 Trigger4 and Livetime5 6 7 8 Calibrated[PRD with 99 low 082013 energy X-rays (2019)]from 55Fe + ∆χ2 cuts 0.1 Energy [keV] + χ2 cut normal Jan. 7th,No ionisation 2020 read-out (only )TMEX2020 - - Edelweiss SubGeV DM searchesAnalysis threshold 10 total deposited energy is collected, 0 allowing for a low threshold 10−1 1 60 eV energy threshold Energy [keV]

Bradley J Kavanagh (GRAPPA) Tiny, tough WIMPs with EDELWEISS-surf LHC Results Forum - 3rd June 2019 Efficiency, signal prediction: pulse simulation

n Efficiency (including deadtime, pileups and 0.4 60 eV analysis c2 cuts) obtained by inserting pulses at threshold Efficiency

random times in actual data stream 0.3 n Same technique used to evaluate response to WIMPs of given masses 0.2 Trigger and Livetime • Case 1: NR from standard WIMPs + Dc2 cuts 0.1 + c2 cut • normal Case 2: ER+NR including Migdal effect Analysis threshold (ejection of n=3-shell e- in WIMP-atom 0 10-1 1 collision [cf Ibe et al, JHEP 03 (2018) 194] ) Energy [keV] 2 GeV/c2 Unsmeared Background Model Background Model 5 Analysis Threshold (60 eV) 105 Analysis Threshold (60 eV) 10 Data Data 2 35 2 2 29 2 Excluded WIMP model: 0.7 GeV/c ,9.8 10 cm Excluded WIMP model: 50 MeV/c ,9.0 10 cm ⇥ ⇥ 2 37 2 2 30 2 Excluded WIMP model: 2.0 GeV/c ,4.5 10 cm Excluded WIMP model: 100 MeV/c ,7.0 10 cm ⇥ ⇥ 2 37 2 2 32 2 Excluded WIMP model: 10.0 GeV/c ,1.1 10 cm Excluded WIMP model: 1.0 GeV/c ,1.6 10 cm 104 ⇥ 104 ⇥ Standard Spectra Migdal Spectra

103 103

2 Number of counts [evts/keV] Number of102 counts [evts/keV] 2GeV/c 102 After pulse simulation 0.03 0.1 1 2 0.03 0.1 1 2 Energy [keV] Energy [keV]

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 11 Filling the gap between ground & space searches

n Shaded regions: with full Earth-Shielding (ES) calculation

n Lines: underground limits (w/o ES calculation, ~ok for <10-31 cm2)

10-24 Stronger upper -25 CMB cutoff for Migdal 10 10-26 XQC rocket -27 (subleading ]

2 10 component) 10-28 EDELWEISS-surf Migdal 10-29 10-30 -31 2 10 EDELWEISS-surf Sharp 45 MeV/c -32 10 CRESST – n-cleus cutoff due to ES -33 EDELWEISS-Surf (Standard) 10 EDELWEISS-Surf (Migdal) effect on velocity 10-34 EDELWEISS-III LT 10-35 CRESST Surface -36 CRESST-II + CRESST-III 10 SuperCDMS LT -37 10 CDMSLite 10-38 LUX (Standard) -39 LUX (Migdal) 10 XENON1T (Standard) 10-40 XENON100 LT -41 NEWS-G 10 DarkSide (Standard) WIMP-nucleon cross section [cm -42 10 XQC -43 CMB * Also: 10 10-44 Neutrino discovery limit spin-dependent CDEX Migdal Underground 10-45 limits 10-2 2´10-2 10-1 2´10-1 1 2 3 4 5 6 7 10 2 [PRD 99 082013 (2019)] WIMP Mass [GeV/c ]

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 12 Next step: going deep underground to reduce backgrounds New results:

UNDERGROUND SEARCH FOR ELECTRON-DM INTERACTIONS WITH A 33G GE DETECTOR

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 13 Theory + experiment motivation: searches for electron-DM interactions n New class of light DM models where the interaction with normal matter comes from the coupling of a Dark Sector photon with the normal photon n Preferred detection mode: DM-electron scattering n Benchmark model predictions: complete coverage (10-40 cm2) possible with ~1 kg.year exposure of a detector sensitive to the single- electron in the absence of any background Community Dark Sectors Report (arXiv:1608.08632): Report 2016 Workshop: Workshop: 2016

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 14 Searching with semiconductors: Ge

n Semiconductors: small gap energy + small leakage current n Challenges: leakage current (1 e-/kg/y ~ 10-27 A/kg) and electrons from radioactivity (1 e-/kg/y ~ 10-8 Bq/kg) n Interesting improvements in bkg levels already obtained in Si with ~g pixel CCD arrays (DAMIC, SENSEI) and <1g cryogenic detector (CDMS) n Ge interesting alternative for a massive detectors : need to assess performance in terms of currents & backgrounds

n HV mode of EDELWEISS-SubGeV program:

• A 33g EDELWEISS-surf detector (equipped with electrodes) operated at ~100 V should reach sub-electron resolution (<<3 eV)

• EDELWEISS has been able to obtain at LSM the lowest radioactive background levels below 10 keV in massive Ge detectors (~0.1 evt/kg/day/keV)

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 15 RED30 Detector n Essentially: EDELWEISS-surf detector + with electrodes added n Ge f 20 x H 20 mm2 (33.6 g) n 1 Ge-NTD sensor (1.6 mm3) glued directly on bottom Ge surface

NTD sensor Al grid electrode n Electrodes on flat surface: lithographed Al grid (500 µm pitch, 4% coverage to reduce trapping) n Outer rings of the grid acts as separate guard electrodes (outer ~2 mm) n No side electrodes on this prototype (mitigation of risk of leakage at HV)

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 16 EDELWEISS detectors High voltage (~100 V) operation mode

Fully Inter-Digitized (FID) 870 g High-Purity Ge detectors operated at ~ 18mK Simultaneous measurement of heat and ionization signals

Heat measurement 2 NTDs (thermal sensors)

m particle-ID dependent m

0 4

Ionization measurement Amplification of the Heat signal due to Neganov-Luke effect =

h 4 sets of electrodes concentric Al rings (2mm spacing) particle-ID dependent

 = 70 mm Huge Amplification of the signal only (not noise) Surface event rejection Luke-NeganovParticle-typeamplification identificaiton lower thresholds + improved resolution (planar mode, no veto electrodes) (thermal sensor measures ionization!) EDELWEISS detectors High voltage (~100 V) operationEDELWEISS mode detectors High voltage (~100 V) operation mode EDELWEISS detectors High voltage (~100 V) operationn modeAddition to heat signal due to drift of charges in applied electric field Fully Inter-Digitized (FID) 870 g High-Purity Ge detectors operated at ~ 18mK Simultaneous measurement of heat and ionization signals Fully Inter-Digitized (FID) 870 g High-Purity Ge detectors operated at ~ 18mK Simultaneous measurement of heat and ionization signals Fully Inter-Digitized (FID) 870 g High-Purity Ge detectors operated at ~ 18mK Simultaneous measurement of heat and ionization signals Heat measurement Heat measurement Heat measurement 2 NTDs (thermal sensors) 2 NTDs (thermal sensors) 2 NTDs (thermal sensors)

m particle-ID dependent m m particle-ID dependent m

0 particle-ID dependent m

4 m

Ionization measurement Amplification of the Heat signal due to Neganov-Luke effect

0 = 4 0

Ionization measurement Amplification of the Heat signal due to Neganov-Luke effect 4 h

= 4 sets of electrodes Ionization measurement Amplification of the Heat signal due to Neganov-Luke effect

= h 4 sets of electrodes concentric Al rings (2mm spacing) particle-ID dependent concentric Al ringsh (2mm spacing) 4 sets of electrodes particle-ID dependent concentricn Amplification Al rings (2mm spacing)proportional to ionization signal and to biasparticle-ID : dependent

 = 70 mm  = 70 mm • Loss of discrimination as heat signal becomes a copy of the ionization signal  = 70 mm Huge Amplification of the signal only (not noise) Surface event rejection Particle-type identificaiton Huge Amplification of the signal only (not noise) Surface event rejection • GainParticle-type in resolution identificaiton by a factor (1+DV/3) for electron signals (planar mode, no veto electrodes) (thermal sensor measures ionization!) lower thresholds + improved resolution (planar mode, no veto electrodes) (thermal sensor measures ionization!) lower thresholds + improved resolution Huge Amplification of the signal only (not noise) Surface event rejection Particle-type identificaiton lower thresholds + improved resolution (planar mode, no veto electrodes) (thermal sensor measures ionization!) Total heat energy [keV] Amplification gain [1+V/3]

Quentin Arnaud IRN Terascale, Bruxelles 3

Total heat energy [keV] Amplification gain [1+V/3] Total heat energy [keV] Amplification gain [1+V/3]

Quentin Arnaud IRN Terascale, Bruxelles Jan. 7th, 2020 TMEX2020 - Edelweiss 3 SubGeV DM searches 17 Total heat energy [keV] Amplification gain [1+V/3] Quentin Arnaud IRN Terascale, Bruxelles 3

Quentin Arnaud IRN Terascale, Bruxelles 3 EDELWEISS setup at LSM

40 228 228 212 212 208

238 214 214

137

210 210 10 2 RED30 tower 210Pb 208 Th, U x-ray (93 - 114), 228Ac, 129.1 226 Ra, 186.2 40 K, 1460.8 + 1464

10 Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 18 214 Bi, 2204.2 208 40 214 Tl, 2614.5 Pb, 351.9 87 212Pb, 238.6 90 90

Counts / (day kg 10 keV) 1 100 2ν2β Mo 210 210 2νββ 100 100 100 -1 2νββ 10 100 0+ 1130.3

1000 2000 3000 210 210 Energy (keV) 210 α 100 210 210 210 2 4 10.308 × 31.927 × γ 210 46.5

γ 40 40 87

40 γ 90 90

238 232 α 2ν2β

100 2 4 2νββ 1+ 2νββ 100 Current EDELWEISS run at LSM n Continuous running at ≤22 mK Jan-Dec 2019 n Eleven Ge detectors • Rest of cryostat used for joint physics run with CUPID-Mo n Compare detector physics in 33g, 200g and 800g detectors n Compare performance of NTD and NbSi-TES heat sensors n Study of low-energy backgrounds in Ge detectors operated with large Luke-Neganov amplification n Limits on Dark Matter using ER and NR spectra of detectors with large Luke- Neganov amplification n Data taking to continue in 2020

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 19 Detector response measurement with 71Ge n Extensive detector studies (calib., fid. volume, charge collection) using 71Ge 10.37 keV line from exposure to a AmBe source prior to LSM installation n April 1st-7th: DM search at highest stable bias (78V) with reduced (but still visible) 71Ge activation n April 10th : in-situ re-activation to Blinded confirm the stability of the detector response and obtain reference sample of 10.37 keV events to be used in the data analysis of the DM search data n May-December: AmBe calibration + further studies of properties of backgrounds at low energy, behavior at high bias, leakage currents, charge collection, etc..

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 20 Detector response

n Ionization channel only used for detector studies, not for DM search n Heat channel calibrated using 71Ge KLM lines (0.16, 1.30 and 10.37 keV) at different biases (0V to 81V)

n Tail due to incomplete charge collection on side surface (confirmed by ring guard signal analysis) is <25%: conservatively taken as a 25% loss of efficiency in the following analysis.

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 21 Detector resolution n Search data: 2.44 day (at 78V) blinded below 1 keV

• Baseline resolution stabilizes around s= 42 eV (1.60 eVee) 4 days after the ramp to 78V

• Constant resolution (from random trigger data) during these 2.44 days n Reference sample: 1.3 days before/after search data

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 22 Efficiency

n Efficiency determined by injecting pulses at random time in the recorded data streams

n The injected pulses are actual 71Ge K-line events (from the activation that immediately followed the search), rescaled to the desired energy

n Some sensitivity to 1 electron-hole pair signal

Simulated number of electron-hole pairs 0.03 7 9 30.0 eV Line 5 10 27.0 eV Line 0.025 4 24.0 eV Line

0.02 3 21.0 eV Line

18.0 eV Line 0.015 2 EDELWEISS 15.0 eV Line RED30 12.0 eV Line 0.01

9.0 eV Line Blinded data 0.005 1 6.0 eV Line Distributions normalized to the surivival probability 3.0 eV Line Non-blinded 0 0 10 20 30 40 50 Energy [eVee]

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 23 Limit setting strategy n No bkg subtraction: 90%CL Poisson upper limit on rate in fixed energy range n Determine most sensitive range using 1.3 day sample non-blinded data (smoothed with KDE) recorded just before/after the search n Signal calculation: QEdark [R. Essig et al., JHEP05 (2016) 046] + charge quantization [as in SuperCDMS, PRL 121 (2018) 051301] + pulse simulation

EDELWEISS Non-blind data KDE RED30

Pre-unblinding: Search interval determination (shaded colors)

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 24 Limit setting strategy n No bkg subtraction: 90%CL Poisson upper limit on rate in fixed energy range n Determine most sensitive range using 1.3 day sample non-blinded data (smoothed with KDE) recorded just before/after the search n Signal calculation: QEdark [R. Essig et al., JHEP05 (2016) 046] + charge quantization [as in SuperCDMS, PRL 121 (2018) 051301] + pulse simulation

EDELWEISS Search data RED30

After unblinding (same search intervals)

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 25 Results

n Limits slightly better than the expectation from the nonblind sample,

because of a slightly better resolution (s = 1.60 vs 1.63 eVee)

n Sensitivity extends into the domain of sub-MeV DM particles: with s = 0.53 e--h pairs there is some sensitivity to single-e- events

n Competitive with Si expts (that use smaller detectors: few g vs 33 g)

2 Heavy DarkBkg : td04a000-td07a000 Sector DetResponse mediator : td05a000-td06a000 case , F=1(F=1) Light DarkBkg : td04a000-td07a000 Sector DetResponse mediator : td05a000-td06a000 case , F=1/q2(F=1/q ) Bkg : td04a000-td07a000 DetResponse : td05a000-td06a000Bkg , F=1 : td04a000-td07a000 DetResponse : td05a000-td06a000 , F=1 −28 −28 ] ]

−28 2 10 −28 2 10 ] ] CDMS CDMS 2 10 2 10 CDMSSuperCDMS CDMSSuperCDMS DAMIC DAMIC [cm DAMIC [cm DAMIC [cm [cm

e DAMIC e DAMIC e e σ σ σ −29 σ XENON XENON XENON XENON −29 −29 XENON XENON 10 10 10 −29 SENSEISENSEISENSEI 10 SENSEISENSEI SENSEI EDELWEISSEDELWEISSEDELWEISS Limit limit Limit EDELWEISSEDELWEISSEDELWEISS Limit limit Limit −30 −30 EDELWEISS expected sensitivity EDELWEISS expected sensitivity 10 10−30 10 EDELSEISSEDELWEISS expected expected sensitivity EDELSEISSEDELWEISS expected expected sensitivity sensitivity 10−30 sensitivity 10−31 10−31 10−31 −31 32 32 10 10− 10− 10−32

10−33 10−33 10−32 10−33 10−34 10−34 10−33 10−34 10−35 10−35 1 10 1 102 10 102 DM Mass [MeV] DM Mass [MeV] 10−35 10−34 1 10 102 1 10 102 DM Mass [MeV] DM Mass [MeV]

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 26 Next steps

Pursuing the study of the noise (& detectorGoal performance) #4: Operation that at high voltage (~100 V) currently limits the sensitivity High Voltage: Exploring DM-electron/nucleus interactions with near single-electron sensitivity achieved n Gain in baseline resolution: x2 on NTD sensorin massive sensitivity bolometers operated and x2underground on bias (low-background environment ~ 1 - 0.1 DRU). had been obtained on other prototype detectors n Study of leakage currents & release of trapped charges (using also the presently available ionization signals) + electrode design n If events are not associated to charge: stress release in crystal? n Noise related to similarEDELWEISS rise at low energy R&D: in theheat spectra sensors observed by other cryogenic detectors? NbSi transition edge sensor NbSi TES on thermal chip: 10.4 keV NbSi209 Preliminary 10.4 keV high impedanceRED30 Preliminary Alternative heat (210 g) sensor under 66 Volt on sapphire or germanium chip 70 Volt 1.3 keV s = 5 eVee investigation: 1.3 keV s = 1.8 eVee high-impedance should reach <10 eV 160resolution eV (RMS) DM 160 eV with standardDM JFET pre-amplifiers NbSi TES search search zone zone 100 nm thick, 20mm diameter spiral NbSi sensor lithographied on a 200 g Ge First EDELWEISS DM-electron scattering and absorption results expected by fall 2019. Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeVJ. BillardDM searches (IPNL) 27 10

S. Marnieros, HDR (2014) http://hal.in2p3.fr/tel-01088881/document

14 Bernhard Siebenborn - EDELWEISS dark matter search Institut für Kernphysik Conclusions

The EDELWEISS-SubGeV program aims at probing MeV-GeV particles via Electron Recoil and Nuclear Recoil signals, with a ~kg-size array n Low-voltage objectives : s = 10 eV (phonon), 20 eVee (ion.) • Particle identification and surface event rejection down to 50 eV n High-voltage objectives : s = 10 eV (phonon) + 100 V bias

• Single e-/h+ pair sensitivity on massive (~30 g) bolometers First results with 33g prototypes: n EDELWEISS-Surf: above-ground search with 18 eV phonon resolution • First Ge-based limit below 1.2 GeV, best above-ground limits down to 600 MeV, Migdal effect used to probe down to 45 MeV n First results with 33g Ge detector with s = 0.6 electron resolution • First use of Ge: currents + background competitive with Si

• Limits below 1 MeV competitive with best results

Jan. 7th, 2020 TMEX2020 - Edelweiss SubGeV DM searches 28