The COBRA Double Beta
Decay Search Experiment
Danielle Stewart
July, 2006
To Follow……
 What is COBRA?
 Shielding work at Warwick
 Current status of COBRA
 Future Plans
University of Warwick, University of Liverpool,
University of Birmingham, University of Sussex,
University of York, University of Dortmund
Danielle Stewart
July, 2006
The COBRA Concept
Cadmium-Tellurium 0-neutrino Beta decay Research Apparatus
 A 64,000 array of 1cm3 CdZnTe semiconductor crystals based at
Gran Sasso (K. Zuber, Phys. Lett. B 519,1 (2001)).
Why use CdZnTe detectors for
a 0νββ search?
Danielle Stewart
July, 2006
Advantages of CZT
 CdZnTe provides 5 β-β-, 4 β+β+ isotopes, e.g. 130Te (2529keV, β-β-),
116Cd
(2809keV, β-β-), 106Cd (2771keV, β+β+)
 T1/2 sensitivity, background limited, scales as:
Maximise Abundance, a, and
Efficiency of Detection, ε:
Maximise Target Mass and
Exposure Time:
 Natural 130Te a ~ 34%
 Scalable, modular design
 Can enrich 116Cd to 90%
 Room temperature operation
T10/2  a
Mt
EB
Maximise Energy Resolution:
Minimise Background:
 Semiconductor
 Clean material manufacture
 Source = detector
 High Q values
 ΔE ~ 1% at 2-3 MeV possible
 Multi-crystal events, pixels
Danielle Stewart
July, 2006
Simulation of Shielding
Background Sources:
Gamma radiation from decay chains of
238U
and 232Th
U/Th from LAAPD's (Large Area Avalanche PhotoDiodes)
Low energy neutrons
High energy neutrons
MCNP – Design Strategy (Monte Carlo N Particle
transport code)
GEANT4 – Realism
Danielle Stewart
July, 2006
Standard Neutron Attenuation
Materials:
•Water
•Polyethylene (Pe)
•Pe + Bi
•Premadex
•Pe + B (30%)
•Pe + B (5%)
•Pe + Li
Danielle Stewart
July, 2006
(n,g)-Reactions
Danielle Stewart
July, 2006
Neutron Energy-Flux Dependence
Depletion
Feeding
Danielle Stewart
July, 2006
Building a Multilayer Shield
Tested:
• block structure sequence
• layer materials in block
• layer ratio’s in block
• number of block repetitions
Danielle Stewart
July, 2006
Comparison to Standard Shields
Danielle Stewart
July, 2006
Comparison to Standard Shields
Danielle Stewart
July, 2006
Shielding Results
 Multilayer shields outperform standard shielding
structures.
 Best results in this study for Metal, Moderator, Capture
Material combination.
 Best Ratio 3:8:4
 Best materials: Lead, Pe-Bi and Pe-Li
 Iron for Lead as cheaper metal against neutrons outside
 Best fine-tuned full shield: Single or double block in
clamp of outer Pe moderator and inner Lead layer.
Danielle Stewart
July, 2006
Current Status: Prototype
 R&D Prototype: 4 x 1cm3 CZT
eV PRODUCTS crystals
 August 2003 – January 2006,
Gran Sasso
 Studied Background and
Electronics
 Uranium contamination in
 Identified passivation paint on
Crystals limited to <490μBqKg-1
crystals as main source of
from 214Bi β-α coincidence.
background
Danielle Stewart
July, 2006
The COBRA Concept: 64 Array
 Installing now at Gran Sasso
 R&D on:
 Energy resolution (N2 cooling).
 Backgrounds: measurement of
contamination levels.
 Background reduction via multicrystal events.
 New passivation paint: has at
least x10 lower background.
Danielle Stewart
July, 2006
Optimising Energy Resolution
1) Used CPG detectors initially due
to best known characterisation.
2) Measurement gives energy
resolution of 3% @2.8MeV.
Collecting Anode
Difference
Pulse Out
Cathode~-2kV
Danielle Stewart
3) Know from He et al. (NIM A388
(1997) 180):
 ΔE dependent on event depth.
 ΔE=1.29% @662keV possible.
 Not limited by CdZnTe
material.
4) Investigating CPG improvements
 Better grid design (He and
Sturm, NIM A554 (2005) 291).
 Digital subtraction.
 Anode÷Cathode readout
 Pixelized readout (see later).
July, 2006
Background Reduction
“King Cobra” – preliminary design for sensitivity to mee~50meV.
 418kg mass in 64000 1cm3 CdZnTe crystals with 90% 116Cd.
Sensitive to 50meV if B<10-3keV-1kg-1yr-1, ΔE<2% at 2805keV (116Cd).
Have to study contribution of potential sources to signal window to
find requirements for shielding and acceptable contamination levels.
Danielle Stewart
July, 2006
Building a Background Model: 1
Flexible Geant4 framework, Venom, developed for COBRA simulation.
1)
2bb decay continuum ‘tail’
 Negligible, with E<2%, B2bb<2x10-7kg-1yr-1keV-1
2)
Neutrons and Muons:
 Simulation of shielding shows these to be negligible.
3)
Ultimately left with a,b,g sources:
 Explore simple model initially, using Geant4 Radioactive Decay
Module.
Danielle Stewart
July, 2006
Building a Background Model: 2
Gas
222Rn gas
Delrin Holder
238U,232Th decay chain
40K
137Cs
Danielle Stewart
Crystals
238U, 232Th decay chains
40K
137Cs
210Po
210Pb on surface
Chamber walls
210Pb on surface
July, 2006
The COBRA Concept
 Simulation results
analysed to give
energy spectrum.
 Reject events with
>1 crystal with
Edep>10keV.
 Count events in
signal window,
2805±28keV.
 Convert counts to event rate as function of contamination.
 Resultant levels conservative – no active veto around crystals.
 U/Th major contributors: O(mBqkg-1) acceptable – same as for other
0bb experiments.
Danielle Stewart
July, 2006
Reducing Background: Pixels
 Pixellating CdZnTe readout enables
tracking:
– Range of a ~15mm.
– Range of 2.8MeV b- ~1mm.
- g’s: separated hits.
3.2 mm
 Simulations with 200-500mm pixels
indicate
- as vetoed with 100% efficiency.
- gs vetoed with ~70% efficiency.
 Testing detectors with 16 (2x2mm)
and 256 (1.6x1.6mm) pixels.
2.8MeV electrons 1.4+1.4MeV electron pairs  Further studies of discrimination of
bb from b events underway.
Danielle Stewart
July, 2006
Conclusions from Current Status
 COBRA’s use of CdZnTe semiconductors offers many advantages for 0bb
searches.
 64 crystal array being installed. Reduced major paint background, limited U
contamination in CdZnTe <490mBqkg-1.
 Detector development underway to optimize energy resolution.
 Detailed study of backgrounds underway. U/Th at mBqkg-1 levels acceptable,
use 64-array to begin contamination measurement.
 Development of pixellated readout offers further background reduction via
discrimination of as and gs from bs through tracking.
Danielle Stewart
July, 2006
Outlook:1
 Background Model – Paper in preparation by Ben (Collection of all
background work done)
 3 detectors to “play” with! New Warwick responsibility.
 New passivation methods
 New readout schemes (home-made pre-amplifier already)
Not usual CPG technique
 Surface characterisation – Chris McConville
Pre-amplifier
Danielle Stewart
July, 2006
Outlook:2
Liquid Scintillator project
 Set up for avalanche photo diode (apd) readout
 Try to replace apd’s with home-made light detectors
-Thick GEM’s (Gas Electron Multipliers) with photocathode
 Mechanical shielding design
Light tight box
Danielle Stewart
July, 2006
Introduction
Oscillation experiments => non-zero neutrino mass
0νββ decay can probe absolute mass scales
If 0νββ decay is detected:
ν is majorana particle
non-conservation of Lepton no. by 2 units
New Physics beyond Standard Model
Postgraduate Seminar – November 2005
Double Beta Decay
Isotopes on left/right decay by β-/(β + and EC) decay
Parabola split due to nuclear pairing energy
Single beta decay is forbidden
Neighbouring odd-odd nucleus becomes virtual intermediate state
Only 35 isotopes have necessary ground state configuration
Postgraduate Seminar – November 2005
Double Beta Decay Possibilities
e
e
e
e
n1
n2
2νββ
n1
n2
0νββ
Second order weak decay
2νββ decay: A Z , N
A Z 2, N 2 2e 2 e
Simultaneous single beta decays – T1/2 ~ 1021 - 1024
yearsA Z , N
A Z 2, N 2 2e
0νββ decay:
Emission and re-absorption of a virtual light neutrino
Involves helicity change, observed decay rate => ν
mass
Postgraduate Seminar – November 2005
Energy Spectrum
A peak at the Q-value is the signature of 0νββ decay
The Q-value corresponds to released energy in
nuclear transition
Half-life varies as Q5
Require good energy resolution
Postgraduate Seminar – November 2005