Dark Matter Search in the EDELWEISS expt

1
Dark Matter Searchin the EDELWEISS expt

• G. Chardin
• DAPNIA/SPP, CEA-Saclay

2
EDELWEISS Collaboration

• CEA-Saclay DAPNIA G. Chardin, M. Gros,A.
  Juillard, A. de Lesquen, M. Loidl, J. Mallet,
  L. Miramonti, L. Mosca, X-F. Navick
• CEA-Saclay DRECAM M. Chapellier, P. Pari
• CRTBT Grenoble A. Benoit, M. Caussignac
• CSNSM Orsay L. Bergé, A. Broniatowski,
  L. Dumoulin, A. Juilliard, S. Marnieros,N.
  Mirabolfathi
• IAP Paris C. Goldbach, G. Nollez
• IPN Lyon B. Chambon, M. De Jesus, D. Drain,
  J. Gascon, J-P. Hadjout, O. Martineau,C. Pastor,
  E. Simon, M. Stern
• Fréjus Underground Lab Ph. Charvin

3
EDELWEISSDark Matter Search

• Introduction
• Limitations of present experiments
• Cryogenic double detection detectors
• First 70 g germanium detectors discrimination
  performances
• Anomalous NaI events
• Second generation 70 g detectors
• Present stage 3 x 320 g Ge detectors
• First results and expected sensitivity
• The EDELWEISS-II experiment
• Perspectives and conclusions

4
WIMPs direct detection(some) conventional and
cryogenic experiments
CRESST Al2O3(Münich/Oxford) _at_ Gran Sasso
EDELWEISS (Ge _at_ Fréjus) ROSEBUD _at_ Canfranc
Milano/Genova/Napoli/ (Te02) _at_ Gran Sasso CDMS
(Ge and Si, Berkeley/Stanford) DAMA (NaI, Xe _at_
Gran Sasso) UKDMC (NaI _at_ Boulby Mine)
ELEGANT, LiF _at_Japan
UKDMC
EDELWEISS
ELEGANT, LiF
CDMS
ROSEBUD
CRESST, HDMS, DAMAGran Sasso
ANTARES (indirect)
AMANDA (indirect)
EDELWEISS Fréjus underground lab
5
The Fréjus Underground Laboratory
Muon flux 4 muons/m2/day Neutron flux
(mainly from rock) 4 10-6 s-1 cm-2
6
EDELWEISSWIMP search

• Measure charge and heat signal to separate
• electron recoils (g background)
• nuclear recoils (neutrons, WIMPs)

7
EDELWEISSWIMP search

• Physics data takingsNot two, but four
  populations observed(nuclear and electron
  recoils, volume and surface events)

8
Unexpected events

• Anomalous events in UKDMC and Saclay NaI
  experiments (not reported by DAMA) events almost
  like nuclear recoils but faster
• EDELWEISS example not two, but four categories
  of events
• Surface nuclear recoils A. Benoit et al., Phys.
  Lett. B 479 (2000) 8
• Surface events probably represent the main
  limitation of NaI, classical germanium (e.g.
  HDMS, GENINO) and cryogenic germanium experiments

9
Anomalous events in NaI experiments

• Tails on the time distribution of NaI events
  Events faster than nuclear recoils
• Ratefew 10-1 - 10-2 evts/kg/keV/day
• Total number lt Nb of alpha events

Smith and Spooner, Phys. World Jan. 2000
10
Low energysurface nuclear recoils

• Consider for example the a-decay of 210Po
  206Pb a due to a pollution at the detector
  surface
• For a total released energy of 5 MeV, 4.9 MeV
  will be carried away by the a particle and 100
  keV by the 206Pb nucleus.

11
Anomalous NaI events surface nuclear recoils ?

• Surface nuclear recoils will give events in the
  5-10 keV range(since quenching factor lt0.1)
• Heavy nuclei at low energy and on surface
  scintillation decay times expected to be faster
  than nuclear Na or I recoils, or alphas(see
  UKDMC and Saclay publications)
• Note energy observed in detector will depend
  strongly on state of detector surface (very flat
  surface  apparent energy 5-10 keV, rougher
  surface  a energy loss in addition possible)
• Event rate appears compatible with a event rate
  (up to trigger effects, should be the same as
  surface alphas)

12
Anomalous eventsConsequences

• Discrimination analysis statistical
  discrimination (distributions overlapping)
  requires that all populations be identified(fit
  with 2 degrees of freedom ? 4 d.o.f)
• Necessary to include additional degrees of
  freedom of surface electron recoils (cf. Saclay)
  and surface nuclear recoils (for NaI all
  discrimination expts)
• Expected to play a significant role in
  spin-dependent analysis (marginally for coherent
  coupling) (factor 10)
• Note even if population is not seen, it should
  be included in the fit (if surface state is
  excellent, visible energy will be 5-10 keV in NaI
  crystals)

13
Anomalous eventsConsequences (ff)

• Important to study in detail surface events for
  charge detectors,and also for scintillation
  detectors
• Further study to confirm our hypothesis analysis
  of coincident data taken with Ge-4 and Ge-5
  detectors (but saturation and non linearity of
  heat channel)
• Hypothesis can be tested on a NaI crystal by
  using an implanted source (Po or Rn, e.g., cf.
  Milano group
• Events present in DAMA ? (same crystals as
  Saclay), discrimination factors taking into
  account additional population ?

14
The DAMA claim

• Detected not by direct observation, but by
  searching for an annual modulation effect (105
  events if correct)

15
2nd generation 70g germanium bolometers

• Improved radioactive environment
• Close roman lead shield
• Removable paraffin shielding (30 cm thick)
• Acoustic isolation improved
• Nitrogen flushing against radon
• New implantation schemes for the electrodes

16
330 g Ge detectors of the "1 kg" stage of
EDELWEISS-I
17
320 g Ge detectors of the "1 kg" stage of
EDELWEISS-I

• Three 320 g Ge detectors
• Close spacing, guard ring for middle detector
• Improved radioactive environment

18
"1 kg" stage of EDELWEISS-I first results

• Very preliminary data from first 320 g
  detector
• Still a factor 5 from CDMS and DAMA benchmarks,
  but resolution can be significantly improved and
  new detector with guard ring germanium
  protection starting to be operated

19
Reaching SUSY models
20
EDELWEISSRD to get rid of surface events

• Discrimination performances of charge heat
  detectors are excellent, but surface events are
  uncomfortably close to nuclear recoil band
• To remove surface events
• record risetime structure of charge signal (A.
  Broniatowski et al.)
• detect out-of-equilibrium phonons and fast vs.
  slow component ratio
• lower charge energy threshold and improve energy
  resolution(AsGa FETs, Single Electron Transistor)

21
Remove surface events (ff)
Radioactive source

• Large NbSi film sensors efficiently collect
  out-of-equilibrium phonons ( fast  component)
• Fast/slow component ratio depends on location of
  energy deposition

Sensor 1
Sensor 2
22
Remove surface events (ff)

• Record risetime structure of charge signal (A.
  Broniatowski et al. LTD8 conf. proc.)
• Cut on risetimeis able to removeelectron
  eventsat 60 keV
• Pb reach thelt10 keV level

23
Decrease charge threshold using theSingle
Electron Transistor

• Charge manipulation in the quantum regime
• Ultra small capacitances now achievable 1 fF
  (10-15 F)
• Coulomb blockade when 1 electron present, other
  electrons must wait outside
• 3-4 orders of magnitude increase in charge
  sensitivity over best FET-based devices if
  capacitance is adapted
• Dual of the SQUID system

24
An SET-basedSi photodetector

• Cleland et al. (Appl. Phys. Lett. 61 (1992)
  2820) proof of principle, coupling of a Si
  detector to a Single Electron Transistor circuit

25
Extending the bandwidth of the SET for charge
detection

• Idea move to high-frequencies
• Ex couple the SET to a GHz resonant circuit and
  measure damping (Schoelkopf et al. Science 1998)

• Fantastic sensitivity still-unoptimized device
  10-5 e/vHz
• Sensitive enough to feel a single particle
• Problem capacitive coupling is difficult
• EDELWEISS realize a prototype with C 2 pF and
  500 eV charge threshold

26
EDELWEISS-II shielding
Efficient protection against neutron and
gamma-ray backgrounds Designed to improve by gt
2 orders of magnitude the best present
performances
27
Innovativereversed cryostat (100 l)
Large volume (100 l) Efficient geometry
Innovative cryogenic solutions (pulsetubes,
compact cryogenic system) 10 mK
base temperature
28
Conclusions

• EDELWEISS Dark Matter Searchnow entering the
  allowed region ofSUSY models
• The present "1 kg" stage should be able to test
  the whole DAMA region
• Detector developments are pursued to
• get rid of surface events
• lower charge energy threshold
• apply developments to other physics applications
  coherent n scattering, solar neutrinos, mass of
  antiproton
• Testing most of SUSY parameter space will require
  gt 50-100 kg stage
• EDELWEISS is exploring with CRESST the best
  strategy for a cryogenic experiment at the 100 kg
  scale.