1. M. Di Marco, P. Peiffer, S. Schönert
LArGe A Liquid Argon Germanium hybrid detector system in the framework of the GERDA experiment
M. Di Marco, P. Peiffer, S. Schönert
Thanks to Marik Barnabe Heider
Cryogenic Liquit Detectors for Future Particle Physics workshop,
LNGS 13th-14th March 2006

2. Outline Introduction: GERDA Energy resolution of bare Ge-diodes in LAr
Experimental Setup of
DAQ
Operational parameters
Light yield
Background spectrum
Characterization with various -sources
137Cs, 60Co, 226Ra, 232Th
bkgd suppression in RoI
Outlook on
Conclusions

3. GERDA – GERmanium Detector Array
Physics goal: search for 0ββ-decay   majorana or dirac particle?
Method: operate bare, 76Ge enriched, HP-Ge-diodes in LN (or LAr)
Signal: single-site events in HP-Ge-diode (Qßß=2039 keV)
Background:  - compton or summation, µ-induced, ...
Physics reach:
Phase I: 15 kg*y, existing diodes (HdM, IGEX)
sensitivity goal: T1/2 > 3*1025 y
mee < 0.24 – 0.77 eV
Phase II: 100 kg*y, increased mass, new diodes, additional active background suppression.
sensitivity goal: T1/2 > 2*1026 y
mee < 0.09 – 0.29 ev
H2O
LN/LAr Ge
LNGS
Challenge: reduce background at 2039 keV by ~102  10-3 cts/(kg*keV*y)

4. Background suppression in GERDA
LN as passive shielding (baseline design)
Cerenkov-muon-veto (Phase I)
Anti-coincidence with adjacent crystals (Phase I)
Pulse shape discrimination (Phase I)
Time correlation between events (Phase I)
Detector-segmentation (Phase II)
LAr scintillation anti-coincidence (option for Phase II)
R&D experiment operating HP-Ge-diode in LAr.
With simultaneous LAr-scintillation-light readout.

5. Energy resolution of a bare 2kg HP-Ge-diode in LAr
1.33 MeV
1.17
MeV
1.33
MeV
FWHM
2.3 keV
40K
summation
208Tl
Resolution in 1.33 MeV
2.3 keV FWHM
Resolution in 1.33 MeV
 No deterioration of energy-resolution for bare p-type detectors in LAr !

6. Outline Introduction: GERDA Resolution of bare Ge-diodes in LAr
Experimental Setup of
DAQ
Operational parameters
Light yield
Background spectrum
Characterization with various -sources
137Cs, 60Co, 226Ra, 232Th
bkgd suppression in RoI
Outlook on
Conclusions

7. LArGe@MPI-K: Schematic system description
Bare p-type HP-Ge-diode
Dewar ∅29 cm, h=65 cm
Light detection: WLS (VM2000)
+ PMT(8“, ETL 9357-KFLB )
Active volume ∅20 cm, h=43 cm
≈ 19 kg LAr
Shielding: 5 cm lead
+ 15 mwe underground -
Measurements:
Internal source
- Background from crystal holders
External source
- Background from walls

8. Electronics Trigger on Ge-signal
Shaping 3 µs
Shaping 3 µs
Trigger on Ge-signal
Record Ge-signal and LAr-signal simultaneously
Coincidence time 6 µs
Software cut on recorded data
LAr

9. Operational parameters
Canberra p-type crystal (390 g)
source
Ge-rate
PMT-rate *
Random coinc.**
Back-ground
7 Hz
2,1 kHz
1,2 %
60Co int. 600 Bq
17 Hz
2,8 kHz
1,68 %
226Ra int. 1kBq
23 Hz
3,2 kHz
1,92 %
Data taking: Sept. 05 – Dec. 05
Stability monitored by:
peak position
energy resolution
leakage current
Energy resolution:
~4.5 keV FWHM w/o PMT
~5 keV with PMT
At 1.33 MeV 60Co-line
* Threshold at single pe (~ 2.5 keV)
** Coincidence time: 6 µs
Background suppression is not compromised by signal loss due to random coincidences !
Energy resolution limited
in this setup.

11. Photo-electron yield in LArGe@MPI-K
spe – peak
(LED generated)
57Co peak in LAr
122 keV - 86%
136 keV - 11%
Source position:
57Co-peak at ch 2153, peak energy keV
spe-peak at ch (122.4 ± 3), pedestal at ch 81
photo-electron yield L = (407 ± 10) pe/MeV
- Possible to improve light yield with TPB (WARP)

12. Background spectrum (LArGe@MPI-K)
Ge signal
(no veto)
40K
40 counts/h
208Tl
10 counts/h
Ge signal after veto:
fraction of the signal which „survives“ the cut
energy in Ge (MeV)
Time of data taking: 2 days

13. Outline Introduction: GERDA Resolution of bare Ge-diodes in LAr
Experimental Setup of
DAQ
Operational parameters
Light yield
Background spectrum
Characterization with various -sources
137Cs, 60Co, 226Ra, 232Th
bkgd suppression in RoI
Outlook on
Conclusions

14. Characterization with different sources
137Cs : single  line at 662 keV
full energy peak :
no suppression with LAr veto
Compton continuum:
suppressed
by LAr veto

15. 137Cs real data simulations 662 keV 662 keV ~ 100% survival
Compton continuum:
20% survival
simulations
very well reproduced by MC(MaGe):
shape of energy spectrum
peak efficiency
peak/Compton ratio
survival probability
662 keV
100% survival
Compton continuum:
20% survival

16. Characterization with different sources
full energy peaks :
no suppression with LAr veto
60Co : two  lines (1.1 and 1.3 MeV) in a cascade
external : high probability that only 1 
reaches the crystal  acts as 2 single  lines
internal : if one  reaches the crystal,
2nd  will deposit its energy in LAr
full energy peak :
suppressed
by LAr veto
Compton continuum:
suppressed
by LAr veto

17. 60Co peak suppression
internal source
external source
1.5 m
100%
40%

18. 226Ra real vs. MC RoI (Qββ=2039 keV) 20% survival No suppression
LAr suppressed

19. 232Th real vs. MC (208Tl+228Ac) 228Ac – contribution
No suppression
228Ac – contribution
 228Ac not in secular equilibrium with 228Th
LAr suppressed
RoI: 6% survival

20. 232Th
No suppression
LAr suppressed
RoI: 6% survival

21. Outline Introduction: GERDA Resolution of bare Ge-diodes in LAr
Experimental Setup of
DAQ
Operational parameters
Light yield
Background spectrum
Characterization with various -sources
137Cs, 60Co, 226Ra, 232Th
bkgd suppression in RoI
Outlook on
Conclusions

22. Outlook: LArGe @ Gran Sasso
Active volume ∅20 cm  supression limited by escapes
Active volume ∅90 cm  No significant escapes. Suppression limited by non-active materials.
Exapmles (MC):
Background suppression for contaminations located
in detector support
Bi-214
Tl-208
factor: 10
LArGe suppression and segmentation
are orthogonal !
 Suppression factors multiplicative
3·10²

23. Conclusions LAr does not deteriorate resolution of p-type crystals
Experimental data shows that
LAr veto is a powerful method for background suppression
No relevant loss of 0ßß signal
Results will be improved in larger
MaGe simulations reproduce well the data

24. 137Cs – effective veto threshold
No suppression
LAr suppressed
LAr-veto threshold ~ 1pe = 2.5 keV

25. 60Co MC vs. real

26. Survival probabilities for LArGe-MPIK setup
Source
137Cs
60Co (ext)
1.3 MeV
232Th (ext.)
583 keV
2.6 MeV
RoI
60Co (int)
232Th (int)
226Ra (int)
609 keV
2,4 MeV
Compton
continuum
15%
~ 30%
~ 25 – 33%
12% 6%
19-27%
full-E
peak
100%
~ 100%
40%
30%
full energy peak :
no suppression
by LAr veto
Compton continuum:
suppressed
by LAr veto
full energy peak :
suppressed
by LAr veto
No efficiency loss expected for 0ßß-events
Random coincidence even for 1 kBq source next to the crystal: < 2%
Background suppression limited by radius of the active volume.
R = 10 cm  significant amount of ‘s escape without depositing energy in LAr

27. (nuclear fuel reprocessing plants)
39Ar, 42Ar and 85Kr
Decay mode
Source
Concentration (STP)
222Rn
T1/2 = 3.8 d
, , 
Primordial 238U
1 - ?00 Bq/m3 air
85Kr
T1/2 = 10.8 y
 (687 keV) , 
235U fission
(nuclear fuel reprocessing plants)
1.4 Bq/m3 air
1.2 MBq/m3 Kr
39Ar
T1/2 = 269 y
 (565 keV)
Cosmogenic
17 mBq/m3 air
1.8 Bq/m3 Ar
42Ar
T1/2 = 32.9 y
 (600 keV)
0.5 µBq/m3 air
50 µBq/m3 Ar
Q-value of 39Ar and 85Kr below 700 keV – relevant in case of dark matter detection
Dead-time could be a problem when Ar scintillation is used (slow decay time: ~ 1µs)
42Ar is naturally low

28. 39Ar and 85Kr in argon Dead time: Assume 10 m3 active volume
39Ar rate: 15 kHz  1.5 % Fine!
85Kr rate not higher  ≤ 0.3 ppm Kr required
Results from a 2.3 kg WARP test stand : ~ 0.6 ppm