Dark Matter Direct Detection with Edelweiss-II

Dark matter direct detection with Edelweiss-II

Eric Armengaud - CEA / IRFU Blois 2008

http://edelweiss2.in2p3.fr/

1 Direct detection of dark matter : principles

. A well-identified science goal: Detect the nuclear recoil of local WIMPs inside some material Target the electroweak interaction scale, m ~ GeV-TeV

 At least 3 strategies:  Search for a global recoil spectrum  V(Earth/Sun): Search for a - small - annual modulation  V(Sun/Wimp gaz): Search for a - large - forward/backward asymetry  Remove backgrounds!!!  Go deep underground  Use several passive shields  Develop smart detectors to identify the remaining radioactivity interactions

2 Direct detection of dark matter

. A well-identified science goal: Detect the nuclear recoil of local WIMPs inside some material Target the electroweak interaction scale, m ~ GeV-TeV

 At least 3 strategies:  Search for a global recoil spectrum

• Cryogenic bolometers: CDMS, CRESST, EDELWEISS… • Liquid noble elements: XENON, LUX, ZEPLIN, WARP, ArDM… • Superheated liquids: COUPP, PICASSO, SIMPLE… • Solid scintillator (DAMA, KIMS), low- threshold Ge (TEXONO), gaz detector (DRIFT)… A very active field with both R&D and large-scale setups

3 The XENON10 experiment

• Quite easily scalable, « high » temperatures • First scintillation pulse (S1) + Second pulse due to e- extraction to the gaz (S2) ⇒ Nuclear recoil discrimination - especially @ low E • Position measurements ⇒ Fiducial cut (~5kg) • Low-energy threshold + Xe mass ⇒ sensitivity to low-mass Wimps

Electronic recoils

Nuclear recoil region

But : not background-free (60 live days run)

4 The CDMS-II experiment • « Towers » of detectors @ 40mK •Heat and ionization pulses ⇒ remove bulk electronic recoils (γs)

• Measurements of phonon delays and risetimes ⇒ remove surface events (βs : miscollected charges), but loose exposure

• 15 cristals used, 1000 « raw » kg.days, 121 kg.days after cuts

timing cut

Still background free - only 1/2 of currently recorded data

5 The Edelweiss-II collaboration

 Based on the experience from Edelweiss-I :  Conceptually close to CDMS  Cryogenic bolometers with heat+ionization channels ⇒ identify bulk gamma-rays  Was already limited, in the end, by the presence of surface (β) interactions with incomplete charge collection  62 kg.d with 3 detectors (final results 2005)  Edelweiss-II :  Completely new setup : host up to 100 detectors, of different kinds  Development of new detectors

EDW-II: • CEA Saclay • CSNSM Orsay • IPN Lyon • Institut Néel Grenoble • FZ / Universität Karlsruhe • JINR Dubna

6 The Edelweiss-II setup - an example of DM direct detection facility

 Operated at the Underground Laboratory of Modane (4μ/day/m2)  Cryogenic installation (~ 20 mK) :  Reversed geometry cryostat  Use of pulse tubes  Shieldings :  Clean room + deradonized air  Active muon veto (>98% coverage)  PE shield  Lead shield ⇒ γ background reduced by ~2 wrt EDW1  Facilities :  Remotely controlled sources for calibrations + regenerations  Detector storage & repair within the clean room  9 cool-downs already operated

7 The Edelweiss-II setup - an example of DM direct detection facility

8 The Edelweiss-II setup - an example of DM direct detection facility

9 Edelweiss-II : the instrumented detectors

 The « standard » EDW-I NTD bolometers:  2 ionization channels (center+guard)  NTD heat channel (edge of detector)  New holders, covers, cabling  May 2008 : 23 detectors (some with a β source on the cover for « calibration »)  R&D bolometers for active β Run 9 setup (April-May) rejection  « NbSi » detectors (May 2008: 4 detectors)  « Interdigit » detectors (May 2008: 4 detectors)  Some other detectors  A Ge73 detector (SD search)

 A heat/scintillation Al2O3 bolometer

10 Standard NTD detector operation ~19 kg.d bckgd Q (spring 07)

Erecoil • First background runs spring 07, « long » « Beta calibration » background run last winter • Already ~100 kg.d of fiducial exposure accumulated (after quality cuts)

• Specific studies of detector’s response to surface events (« beta calibrations ») to have a quantitative understanding of the background and try to take it into account in the analysis

11 NbSi detectors • Detectors developed at CSNSM : identify surface events = athermal phonon measurement with Nb-Si films « replacing » the NTDs • Surface event rejection ok, some problems in 2007 with film contacts / leak currents

ThermometerN bSii A ThermometerN bSi A

1.5 fiducial kg.d, 2007

ThermometterN bSi B ThermometterN bSiB

Heat signal of thermometNerb SiA : 200 Surface event Bulk event

150

100 Transient

Thermal 50

0 0 20 40 60 80 -3 Txi1m0 e (m s)

12 Interdigit detectors

‘a’ electrodes (+2V) collecting ‘b’ electrodes (-1V) field shapping

• Keep the standard phonon detector • Modify the E field near the surfaces with interleaved electrodes (with guards: 6 ‘c’ (-2V) ‘d’ (+1V) ionization channels) • Use ‘b’ and ‘d’ signals as vetos agains surface events From preliminary see-level measurements: • Surface event rejection > 95% • Fiducial volume (for a 200g detector) ~ 50%

13 Interdigit detectors : background runs at LSM

 Few kg.d of background runs with a 200g detector:  Performance as expected  Now installing 3 more 400g detectors, and doing a β calibration, for further performance tests Q

⇒ A promissing, reasonably cheap and simple detector

~ 1.7 kg.d

14 Outlook

 Dark matter direct detection :  A rapidly evolving field  Several detection strategies available now:  Indispensible to cross-check a possible discovery  Each technology has its advantages and unknowns

 EDELWEISS-II is taking data :  Quite large exposures in stable conditions already achieved, « 4-month cycles » with regular improvments and new detectors  New, promissing « Interdigit » detector is operational !  The sensitivity of the experiment will - of course - depend on the ultimate detector performances (surface event rejection) and background encountered  Aim ~10-8 pb, possibly even lower  (Realistic) prospective example : background rate ~ 0.05 evts/kg.d ⊗ rejection factor 97% ⇒ with 1000 kg.d, expect 1 unrejected event, ie. reach ~ 2x10-8 pb sensitivity