1. From Edelweiss I to Edelweiss II
Véronique SANGLARD
CNRS/IN2P3/IPNL

2. Outline The Edelweiss-I experiment 2003 results
Ionization trigger data taking
Phonon trigger data taking
The second phase : Edelweiss-II
Perspectives and conclusion

3. The Edelweiss* collaboration
CEA-Saclay DAPNIA/DRECAM
CRTBT Grenoble
CSNSM Orsay
FZK/Univ. Karlsruhe
IAP Paris
IPN Lyon
Laboratoire Souterrain de Modane
1700 m depth under the Fréjus tunnel (4800 we)
4 µ/m²/d (106 less than at the surface)
*Expérience pour DEtecter Les WIMPs
En SIte Souterrain
(Underground experiment to detect WIMP)

4. Heat and ionization detectors
threshold
Simultaneous measurement of
charge and heat signals for each
interaction
Different charge/heat ratio for
nuclear and electron recoils
(γs, βs ionize more than WIMPs
and neutrons)
Neutrons Ge(n,n',γ) Gammas
Event by event discrimination
Discrimination > 99.9 % for Erec>15 keV

5. Charge collection
Miscollected charge events can simulate nuclear recoils
Use of gamma calibration (57Co, 137Cs) to check the detector charge collection quality
Few miscollected charge events with amorphous layer
Since 2002 use detectors with amorphous layer
WITH
WITHOUT

6. The 1 kg stage Shield : 30 cm paraffin 15 cm lead 10 cm copper
Al sputtered
electrodes
NTD heat sensor
Ge or Si amorphous layer
Guard ring
fiducial vol. : 57 %*
*(O.Martineau et al. Nim A in press)

7. 2003 Edelweiss data Additional ~45 kg.d recorded with 3 new detectors
2 phases with 2 different triggers
On the ionization signal
On the phonon signal
data represent ~ 62 kg.d
Results :
Events observed in nuclear recoil band (40 for Erec > 15 keV)

8. 2003 data (ionization trigger)
20 kg.d
Energy threshold : 20, 30 keV
3 events observed in the nuclear recoil band (above these thresholds)

9. 2003 data (phonon trigger) Lower energy threshold : 15 keV
18 events observed in nuclear recoil band, most (12) below 30 keV in 22 kg.d
1 coincidence n-n observed between detectors (10% prob.)
Stable behavior of 3 detectors over total exposition

10. Edelweiss new limit PRELIMINARY
Unknown backgrounds
"Yellin method"* used to derive exclusion limit
*(PRD 66, (2002))
No background subtraction
New (prel.) limit consistent with the previous publication* *(Phys. Lett. B (2002))
PRELIMINARY

11. Experimental spectrum
Low energy spectrum
inconsistent with Wimp
mass > ~ 20 GeV
Possible backgrounds
Neutrons (n-n coinc.)
Miscollected charge
events
(surface events)

12. Lessons from Edelweiss-I
With 3 new detectors and an extended exposure, the preliminary 2003 exclusion limit confirms the previous published one
Surface events :
Improved radiopurity in Edelweiss-II
Identification (or suppression) possible with NbSi thin film sensor
Neutron background :
Improved shielding against neutron
Anti-coincidences more efficient with increased number of detectors

13. Identification of surface events
2 NbSi athermal phonon sensors for surface event rejection
Two components :
Thermal (energy)
Athermal/transitory (near-surface tag)
For this surface event, the athermal component is higher in NbSi 1
First tests of 200g modules in Edelweiss-I promising :
10 x less background while retaining 50 % efficiency

14. Perspectives : Edelweiss-II
Aim : x 100 improvement in sensitivity
1st phase :
21*320g Ge bolometers with NTD heat sensor
7*400g Ge bolometers with NbSi thin film sensor
Installation started in April 2004
Data taking in 2005

15. Edelweiss II : new cryostat
Larger experimental volume
Low radioactivity cryostat
Innovative reversed geometry
10 mK base temperature
First phase : 28 detectors, up to 120

16. Edelweiss II : new setup
Clean room
Efficient shielding against neutron and gamma ray background
20 cm lead
50 cm PE
Muon veto
Sensitivity
Edelweiss I :
0.2 evt/kg/day
Edelweiss II :
0.002 evt/kg/day

17. Conclusion 2003 preliminary data confirm the 2002 exclusion limit
Edelweiss-I is sensitive to -optimistic- SUSY models (10-6 pb)
Edelweiss-II, goals :
To reach more favored SUSY models (10-8 pb)
Competitive with CDMS-II, CRESST-II
Testing the bulk of SUSY parameter space (>10-10 pb) will require one-ton detector array and an extreme background rejection

18. The Modane Underground Laboratory
Edelweiss I
Edelweiss II
1700 m depth under the Fréjus Tunnel (4800 we)
4 µ/m²/day (106 less than at the surface)
1500 neutrons (>1 MeV)/m²/day (rock radioactivity)

19. Direct detection
Principle : Elastic scattering of a WIMP on a target nucleus
Constraints :
Low energy deposited : Erecoil <1OO keV
Small interaction rate : R<1 evt/kg/day
Electronic recoil :
background
Nuclear recoil :
signal

20. Neutron-gamma discrimination
Use of neutron calibration
to simulate WIMP nuclear recoils
Evt/evt Discrimination (between nuclear and electronic recoils)
> 99.9 % for Erec > 15 keV
Use n-n coincidences to identify neutron background
Ionization
Threshold
Neutrons 73Ge(n,n',γ) Gammas

21. Edelweiss-I 2002 data 20

22. Resolutions

23. Phonon trigger data Trigger efficiency very precisely determined
by coincidence
spectrum in neutron data
Threshold at 50 % efficiency (in recoil energy)

24. Expected sensitivity

25. 137Cs calibration with new trigger (1)
High statistics γ 137Cs calibration
(~105 events)
Charge collection quality (check of γ rejection)
Exposure (in gamma rays) corresponding to ~ 2 years of data taking
~ 10 nuclear recoil events observed after one week
~ 31 after two weeks
Coincidences (between nuclear recoils) observed between detectors

26. 137Cs calibration with new trigger (2)
Top detector
5 events in nuclear recoil
band with 1 coincidence
Further away from background pollution
Middle detector
11 events in nuclear recoil
band with 2 coincidences
Closer to background pollution
Bottom detector
15 events in nuclear recoil
band with 1 coincidence
Closest to background pollution

27. 137Cs calibration with new trigger (3)
Summary of all three detectors
~ 22 (Erec > 15 keV) events in nuclear recoil band
> events below 200 keV recoil energy

28. 137Cs calibration with new trigger (4)
What are really these events ?
Compatible with nuclear recoils
Some of them are coincidences between detectors
Probably neutrons
137Cs calibration suggests a contamination by 252Cf source
Confirmed by measurement of the source holder in low-background Ge diode setup
Edelweiss sensitive to low nuclear recoil rates