- Email: [email protected]
Proposal for a new Ge-semiconductor dark matter detector
Nuclear
Instruments
and Methods
in Physics Research
A 385
( 1997)
265-267
NUCLEAR INSTRUMENTS & METHODS IN PHYSICS RESURCH Section A
EISEVIER
Proposal for a new Ge-semiconductor dark matter detector L. Baudis, J. Hellmig, H.V. Klapdor-Kleingrothaus H. S trecker Max-Planck-Insriturfiir
Kemphyik,
I, A. Miiller, F. Petry, Y. Ramachers*,
PO. Box 10 39 80. D-69029
Received
30 July
Heidelbrrg.
Germon.~
1996
Abstract We present a new proposal for a Cie-semiconductor dark matter detector. As will be shown the use of anti-coincidence between two Ge-detectors will drastically reduce the photon background which mainly limits the direct detection of hypothetical dark matter particles (WIMPS). We expect to improve WIMP cross section limits to a level comparable to planned cryogenic experiments.
The exciting possibility to detect cold dark matter in our galaxy in the form of WIMPS (weakly interacting massive particles) through direct detection of elastic WIMP-nucleus scattering events [ 1,2] is limited by low signal-to-noise ratios due to the very low expected rates [3-51 in the relevant part of the energy spectrum (typically below 50 keV) So the aim of every experimental attempt to directly detect
’ Corresponding
author. E-mail
[email protected].
* Corresponding
author.
[email protected].
E-mail
Isolation
iU
Fig. 1. Schematic
geometry
of our new anti-coincidencedetector
well-type
Ge crystal
is planned
surement
Ge crystal
as n-type.
will
be 90 mm,
height
about 30 mm diameter
0168-9002/97/$17.00 PI/SOl68-9002(96)01019-4
90
P-type
as a p-type
semiconductor,
The approximate
mm;
the small
diameter
measurement
design. The the small mea-
of the well-type crystal
will
have
All
rights reserved
and height 40 mm.
Copyright
@
1997 Elsevier
Science
B.V.
WIMPS is to optimize the interplay of three important parameters: threshold energy (as low as possible), target mass (as high as possible) and background, mainly induced by photon events (as low as possible). In this letter we introduce a new idea for a dark matter detector using Ge-semiconductor technology. We aim at a further reduction of background compared to our first steps in the field of dark matter research [6] with which we obtained the most stringent limits for heavy WIMPS until recently [ 71. We will use our experience with Ge detectors from the O&? Heidelberg-Moscow experiment [8,9] to build the cryostat system and the surrounding of the special crystal configuration introduced below. A small Ge-crystal (natural Ge for the prototype, isotopitally enriched Ge for the second stage) is surrounded by a well-type Ge-crystal and the configuration is run in anticoincidence mode (see Fig. 1) . Both crystals are mounted into a common cryostat system. A thin insulator (about 1 mm) is placed between the two crystals in order to shield leakage currents on the surfaces. The outer well-type Ge detector will be a very effective anti-coincidence counter. This efficiency for different photon energies has been simulated in a Monte Carlo simulation using the GEANT [lo] code and the schematic geometry from Fig. 1. The results are expressed as suppression factors as functions of the initial photon energies. The suppression factor is defined by the ratio of the number of unsuppressed counts in the interesting energy interval 0- 120 keV over the number of suppressed counts in the same energy interval measured by the small inner detector crystal. Gamma-ray radioactivity was simulated in four different volumes: i) a cylinder volume surrounding the well-type Ge detector, ii) a volume above the top and iii) under the bottom side of the well-type and iv) inside the insulator material
L. Baudis et nl./Nucl.
266
h60
s
Instr. and Meth. in Phys. Rex A 385 (1997)
265-267
“‘I”“,““J”.‘I”‘.l’
_ : cvlinder
500
1000
1500
2uBnergy iI%%]
Fig. 2. ‘Ike results of the Monte Carlo simulation for the efficiency of the anti-coincidence (suppression factor in the energy interval O-120 keV) between the two Ge crystals from Fig. I. Shown is the dependence on different photon energies for different photon source locations. The simulated geometry is shown in Fig. I, photons have been started from the cylinder shell surrounding the well-type, from the open top, the bottom and from inside the isolation material between the crystals.
between the small and the well-type detector. In each volume six discrete photon energies were simulated: 238, 511, 911, 1461, 2204, and 2614 keV. For the results, see Fig. 2. In addition we simulated a typical cosmogenitally induced background contribution from @‘Coand the natural decay chain of “sAc. The results are comparable or even better (60Co decays with two coincident photons which increases the anti-coincidence efficiency) than the suppression factors of the 2614 keV photons. The efficiency of the anti-coincidence is lowest for photons starting in the insulator material between the crystals.
200
250
MwlMPLGeVY Fig, 3. Speculative view for obtainable cross section limits on WIMP-neutron (spin-independent interaction) elastic scattering for three future experiments, CRESST and CDMS (Label 2) [ 10-I 21 and our proposal (Label 1). For comparison the present lowest-limit results, the Heidelberg-Moscow result [6] and the UK-COIL result for NaI [7], are also shown together with expected cross sections for neutralinos, first for a SUSY-GUT scenario (filled dots) [ 31 and second for a SUSY-GUT scenario without universal scalar mass unification [ 31 (open boxes).
MWIMP
rGeV]
Fig. 4. The same as Fig. 3 but for WIMP-proton (spin-dependent interaction) elastic scattering. The difference between the SUSY GUT and the relaxed SUSY-GUT scenario is nearly invisible.
Therefore it is essential to use selected radiopure material for this structural part and we are currently testing selected materials in the Gran Sasso underground laboratory. However, even a specific activity comparable to that of the selected copper surrounding the crystals would result in a negligible contribution to the background due to the low mass needed for the insulation material. Suppression factors for photons started from the top of the well detector (open side, see Fig. 1) are lower than for photons from the bottom or the cylinder shell. They are more likely to escape the anticoincidence by Compton backscattering thereby just leaving a signal inside the small detector cystal. The spread of the suppression factors at some energies is due to the low statistics of 10’ simulated photons from each volume and the strong passive shielding by the well detector crystal. This is especially important for low energy y-rays which can hardly reach the inner crystal. From this simulation we roughly estimate a background reduction of a factor of 20 due to the anti-coincidence operation mode. Our experience with low-level Ge-detectors gives us the opportunity to start from a background level at least as low as already achieved in the Heidelberg-Moscow experiment [9]. The main background contribution always originates from materials nearest to the measurement crystals. In our case this would be one of the radiopurest materials we know, a second Ge-crystal. Therefore we think to give a conservative estimation when we prospect the WIMP cross section limit for our new detector design in Figs. 3 and 4 in the following way: We take our first low energy spectrum from which we derived our current WIMP limits [ 61 and reduce the counting rate by scaling with a factor of l/20 and shifting the threshold down to 5 keV which is achievable with the small measurement crystal. We conclude to reach WIMP cross section limits comparable to planned cryogenic experiments [ 1 l-131 with our new Ge-semiconductor dark matter experiment which will start in 1997.
L. Boudis et ol./Nucl.
Instr. und Meth. in Phw Res. A 385 (1997) 265-267
References
171 P.F. Smith
et al., Phys. Lat.
[ 81 Heidelberg-Moscow [ I I P.F. Smith
and J.D.
Lewin.
Phys. Rep.
187 ( 1990)
[ 21 R. Bernabei, Riv. Nuovo Cimento I8 ( 1995) I [ 31 v. Bednyakov, H.V. Klapdor-Kleingrothnus, S.G. Ramachers,
[4 ]
Preprint
V. Bednyakov, Rev. D 50
1S1 V.
Preprint
J. Ellis,
and S.G. Kovalenko,
N. Fornengo.
Phys.
B 336
G. Mignola
and S.
( 1994)
141.
1996);
[I I I
P.D. Barnes et al., Nucl.
[ 131 M.
(1994) Biihler
Double
B 356 ( 1995)
Beta and
Decay S.
Related
Stoica
(World
( 1987).
ReportDDIEE/84-I
Instr. and Meth.
A 370
( 1996)
Temp.
( 1993)
797:
Phys. 93
233.
Phys. Lett.
95. et al.. Nucl.
450.
and
Phys. Rev. D. to be published.
User’s Guide,CERN
et al., J. Low
299.
and
Collaboration,
R. Brun.GEANT3
[ 121 W. Seidel
in:
( 1996)
Phys. Lett.
Klapdor-Kleingrothaus
[IO]
323
HEP-PH-9508249.
Beck et al.. Phys. Len.
H.V.
Singapore
Heidelberg-Moscow
HEP-PH-9606261. Klapdor-Kleingrothaus
A. Bottino.
eds.
Scientific.
and Y.
B 379
Collaboration.
Klapdor-Kleingrothaus,
Topics, Kovalenko
(1994) 7128.
Berezinsky,
ScopeI, [ 61 M.
H.V.
191 H.V.
203.
267
Instr. and Metb.
A 370
( 1996)
237
B
Instruments
and Methods
in Physics Research
A 385
( 1997)
265-267
NUCLEAR INSTRUMENTS & METHODS IN PHYSICS RESURCH Section A
EISEVIER
Proposal for a new Ge-semiconductor dark matter detector L. Baudis, J. Hellmig, H.V. Klapdor-Kleingrothaus H. S trecker Max-Planck-Insriturfiir
Kemphyik,
I, A. Miiller, F. Petry, Y. Ramachers*,
PO. Box 10 39 80. D-69029
Received
30 July
Heidelbrrg.
Germon.~
1996
Abstract We present a new proposal for a Cie-semiconductor dark matter detector. As will be shown the use of anti-coincidence between two Ge-detectors will drastically reduce the photon background which mainly limits the direct detection of hypothetical dark matter particles (WIMPS). We expect to improve WIMP cross section limits to a level comparable to planned cryogenic experiments.
The exciting possibility to detect cold dark matter in our galaxy in the form of WIMPS (weakly interacting massive particles) through direct detection of elastic WIMP-nucleus scattering events [ 1,2] is limited by low signal-to-noise ratios due to the very low expected rates [3-51 in the relevant part of the energy spectrum (typically below 50 keV) So the aim of every experimental attempt to directly detect
’ Corresponding
author. E-mail
[email protected].
* Corresponding
author.
[email protected].
Isolation
iU
Fig. 1. Schematic
geometry
of our new anti-coincidencedetector
well-type
Ge crystal
is planned
surement
Ge crystal
as n-type.
will
be 90 mm,
height
about 30 mm diameter
0168-9002/97/$17.00 PI/SOl68-9002(96)01019-4
90
P-type
as a p-type
semiconductor,
The approximate
mm;
the small
diameter
measurement
design. The the small mea-
of the well-type crystal
will
have
All
rights reserved
and height 40 mm.
Copyright
@
1997 Elsevier
Science
B.V.
WIMPS is to optimize the interplay of three important parameters: threshold energy (as low as possible), target mass (as high as possible) and background, mainly induced by photon events (as low as possible). In this letter we introduce a new idea for a dark matter detector using Ge-semiconductor technology. We aim at a further reduction of background compared to our first steps in the field of dark matter research [6] with which we obtained the most stringent limits for heavy WIMPS until recently [ 71. We will use our experience with Ge detectors from the O&? Heidelberg-Moscow experiment [8,9] to build the cryostat system and the surrounding of the special crystal configuration introduced below. A small Ge-crystal (natural Ge for the prototype, isotopitally enriched Ge for the second stage) is surrounded by a well-type Ge-crystal and the configuration is run in anticoincidence mode (see Fig. 1) . Both crystals are mounted into a common cryostat system. A thin insulator (about 1 mm) is placed between the two crystals in order to shield leakage currents on the surfaces. The outer well-type Ge detector will be a very effective anti-coincidence counter. This efficiency for different photon energies has been simulated in a Monte Carlo simulation using the GEANT [lo] code and the schematic geometry from Fig. 1. The results are expressed as suppression factors as functions of the initial photon energies. The suppression factor is defined by the ratio of the number of unsuppressed counts in the interesting energy interval 0- 120 keV over the number of suppressed counts in the same energy interval measured by the small inner detector crystal. Gamma-ray radioactivity was simulated in four different volumes: i) a cylinder volume surrounding the well-type Ge detector, ii) a volume above the top and iii) under the bottom side of the well-type and iv) inside the insulator material
L. Baudis et nl./Nucl.
266
h60
s
Instr. and Meth. in Phys. Rex A 385 (1997)
265-267
“‘I”“,““J”.‘I”‘.l’
_ : cvlinder
500
1000
1500
2uBnergy iI%%]
Fig. 2. ‘Ike results of the Monte Carlo simulation for the efficiency of the anti-coincidence (suppression factor in the energy interval O-120 keV) between the two Ge crystals from Fig. I. Shown is the dependence on different photon energies for different photon source locations. The simulated geometry is shown in Fig. I, photons have been started from the cylinder shell surrounding the well-type, from the open top, the bottom and from inside the isolation material between the crystals.
between the small and the well-type detector. In each volume six discrete photon energies were simulated: 238, 511, 911, 1461, 2204, and 2614 keV. For the results, see Fig. 2. In addition we simulated a typical cosmogenitally induced background contribution from @‘Coand the natural decay chain of “sAc. The results are comparable or even better (60Co decays with two coincident photons which increases the anti-coincidence efficiency) than the suppression factors of the 2614 keV photons. The efficiency of the anti-coincidence is lowest for photons starting in the insulator material between the crystals.
200
250
MwlMPLGeVY Fig, 3. Speculative view for obtainable cross section limits on WIMP-neutron (spin-independent interaction) elastic scattering for three future experiments, CRESST and CDMS (Label 2) [ 10-I 21 and our proposal (Label 1). For comparison the present lowest-limit results, the Heidelberg-Moscow result [6] and the UK-COIL result for NaI [7], are also shown together with expected cross sections for neutralinos, first for a SUSY-GUT scenario (filled dots) [ 31 and second for a SUSY-GUT scenario without universal scalar mass unification [ 31 (open boxes).
MWIMP
rGeV]
Fig. 4. The same as Fig. 3 but for WIMP-proton (spin-dependent interaction) elastic scattering. The difference between the SUSY GUT and the relaxed SUSY-GUT scenario is nearly invisible.
Therefore it is essential to use selected radiopure material for this structural part and we are currently testing selected materials in the Gran Sasso underground laboratory. However, even a specific activity comparable to that of the selected copper surrounding the crystals would result in a negligible contribution to the background due to the low mass needed for the insulation material. Suppression factors for photons started from the top of the well detector (open side, see Fig. 1) are lower than for photons from the bottom or the cylinder shell. They are more likely to escape the anticoincidence by Compton backscattering thereby just leaving a signal inside the small detector cystal. The spread of the suppression factors at some energies is due to the low statistics of 10’ simulated photons from each volume and the strong passive shielding by the well detector crystal. This is especially important for low energy y-rays which can hardly reach the inner crystal. From this simulation we roughly estimate a background reduction of a factor of 20 due to the anti-coincidence operation mode. Our experience with low-level Ge-detectors gives us the opportunity to start from a background level at least as low as already achieved in the Heidelberg-Moscow experiment [9]. The main background contribution always originates from materials nearest to the measurement crystals. In our case this would be one of the radiopurest materials we know, a second Ge-crystal. Therefore we think to give a conservative estimation when we prospect the WIMP cross section limit for our new detector design in Figs. 3 and 4 in the following way: We take our first low energy spectrum from which we derived our current WIMP limits [ 61 and reduce the counting rate by scaling with a factor of l/20 and shifting the threshold down to 5 keV which is achievable with the small measurement crystal. We conclude to reach WIMP cross section limits comparable to planned cryogenic experiments [ 1 l-131 with our new Ge-semiconductor dark matter experiment which will start in 1997.
L. Boudis et ol./Nucl.
Instr. und Meth. in Phw Res. A 385 (1997) 265-267
References
171 P.F. Smith
et al., Phys. Lat.
[ 81 Heidelberg-Moscow [ I I P.F. Smith
and J.D.
Lewin.
Phys. Rep.
187 ( 1990)
[ 21 R. Bernabei, Riv. Nuovo Cimento I8 ( 1995) I [ 31 v. Bednyakov, H.V. Klapdor-Kleingrothnus, S.G. Ramachers,
[4 ]
Preprint
V. Bednyakov, Rev. D 50
1S1 V.
Preprint
J. Ellis,
and S.G. Kovalenko,
N. Fornengo.
Phys.
B 336
G. Mignola
and S.
( 1994)
141.
1996);
[I I I
P.D. Barnes et al., Nucl.
[ 131 M.
(1994) Biihler
Double
B 356 ( 1995)
Beta and
Decay S.
Related
Stoica
(World
( 1987).
ReportDDIEE/84-I
Instr. and Meth.
A 370
( 1996)
Temp.
( 1993)
797:
Phys. 93
233.
Phys. Lett.
95. et al.. Nucl.
450.
and
Phys. Rev. D. to be published.
User’s Guide,CERN
et al., J. Low
299.
and
Collaboration,
R. Brun.GEANT3
[ 121 W. Seidel
in:
( 1996)
Phys. Lett.
Klapdor-Kleingrothaus
[IO]
323
HEP-PH-9508249.
Beck et al.. Phys. Len.
H.V.
Singapore
Heidelberg-Moscow
HEP-PH-9606261. Klapdor-Kleingrothaus
A. Bottino.
eds.
Scientific.
and Y.
B 379
Collaboration.
Klapdor-Kleingrothaus,
Topics, Kovalenko
(1994) 7128.
Berezinsky,
ScopeI, [ 61 M.
H.V.
191 H.V.
203.
267
Instr. and Metb.
A 370
( 1996)
237
B