GRETINA - CNEAscoccola/iguazu/Pt/09_PT09... · 2005-10-28 · Summary γ1 Large arrays of Ge...
Transcript of GRETINA - CNEAscoccola/iguazu/Pt/09_PT09... · 2005-10-28 · Summary γ1 Large arrays of Ge...
SLAFNAP6IGUAZU - ARGENTINA
October 3-7, 2005
A. O. MacchiavelliLawrence Berkeley National Laboratory
GRETINA
Many Thanks toI-Yang Lee, P.Fallon, M.Descovich and S.Ettenauer
Outline
Introduction
Concept of γ-ray tracking
Impact on spectroscopy
Proof of principleSegmented detectorsElectronicsPosition reconstructionIn-beam results
GRETINADesign and expected performanceStatus of the project
Summary and conclusions
Gamma-ray spectroscopy has played a major role in the study of the atomic nucleus.
Development of new detectors and techniques have always led to discoveries of new and unexpected phenomena.
Gamma-ray Spectroscopy and Nuclear Physics
“Spectroscopic history”of 156Dy
Resolving power - Figure of Merit
“Effective” Energy resolution (δE), Efficiency (ε), Peak-to-total (P/T)
Plus auxiliary devices
Weakest branch that can be resolved
fold
α(lo
g)
1/rf
(N/No)/εf
Resolving power - Figure of Merit
f*
Resolving powerα*
Evolution over the years
- HV signal
p n
Intrinsic energy resolution determined by statistics of charge carriers ~
valence band
conduction band
0.7 eV ε=3 eV
Germanium Semi-conductor Detectors
Best energy resolution
Moving nucleus
θ
γ-ray detector
V±ΔVΔθΝ
ΔθD
Broadening of detected gamma ray energy due to:Spread in speed ΔVDistribution in the direction of velocity ΔθΝDetector opening angle ΔθD
Need accurate determination of V and θ.Position sensitive γ-ray detector and particle detector
Doppler shift
Doppler Broadening
Doppler Broadening
Peak/Total = 20%ε=20%
Compton suppressor
20%
Veto
P/T=60%ε =20%
Compton Suppression
Improve peak-to-total ratio
Number of modules 110Ge Size 7cm (D) × 7.5cm (L)Distance to Ge 25 cm
Peak efficiency 9% (1.33 MeV)Peak/Total 55% (1.33 MeV)Resolving power 10,000
Almost 10 !Almost 10 !
Ν = 100ΝΩ ε = 0.1Efficiency limited
Veto
Compton Suppressed Ge Ge Sphere Gamma Ray Tracking
Ν = 1000 (summing)ΝΩ ε = 0.6Too many detectors
Ν = 100ΝΩ ε = 0.6Segmentation
sum
Towards the “Ultimate” Ge Array
Pulse shape analysis in segments3D position of interaction points
Tracking of photon interaction points energy and position of γ-ray
Gamma-ray Tracking
• High position resolution
• High efficiency
• High peak to background• High counting rate• Background rejection
Large recoil velocity– Fragmentation and– Inverse reactions
Low beam intensity
High background rate– Beam decay– Beam impurity
Advantages of γ-ray Tracking(In particular for Radioactive Beams)
Resolving power: 107 vs. 104
– Cross sections down to ~1 nb• Most exotic nuclei• Heavy elements (e.g. 253,254No)• Drip-line physics• High level densities (e.g. chaos)
Efficiency (high energy) (23% vs. 0.5% at Eγ=15 MeV)
– Shape of GDR– Studies of hypernuclei
Efficiency (slow beams) (50% vs. 8% at Eγ =1.3 MeV)
– Fusion evaporation reactionsEfficiency (fast beams) (50% vs. 0.5% at Eγ =1.3 MeV)
– Fast-beam spectroscopy with low rates -> RIA
Angular resolution (0.2º vs. 8º)– N-rich exotic beams
• Coulomb excitation– Fragmentation-beam spectroscopy
• Halos• Evolution of shell structure• Transfer reactions
Count rate per crystal (100 kHz vs. 10 kHz)
– More efficient use of available beam intensity
Linear polarization
Background rejection by direction
Physics opportunities with a 4π array GRETA
Signal Generation
Dx
Q (%)
x=0
x=D/2
V=1V=0
V=0
Ex weighting field
weighting potential
AB
Q (%) time
A
B
A+B
Advances in detector segmentation
Electronics developmentLow noise high bandwidth preamplifier
Rapid sampling, high resolution pulse digitizer
Signal analysis algorithm
Tracking algorithm
Computing power for on-line processing
Technical Challenges
SeGA MSUANL - GARBO EXOGAM
Segmented Germanium Detectors
MINIBALL AGATATIGRESS
Segmented Germanium Detectors
GRAPE
Prototype detectors at LBNL
PII
PIII
Segments and segment size
Pre-Amp
Eurisys PSC823
FET IF1320
Gain 200mV/MeV
Rise Time ~40nsec
Decay Time 50μsec
Power 50mW
Performance with detector: Energy resolution 1.15 keV Am
2.5 keV Co
Noise level 4 keV (20MHz)
LBNL VME 8-channel digital signal processing boards (100 MHz- 12 bit) Individual or Common clockProgrammable triggerFPGA Energy (P/Z)FPGA CFD timingUp to 10μsec data samples
Digitizer and Data Acquisition
Acquisition rate > 8 Mbytes/secData stored to a redundant disk array (1Tbytes) over Gigabit networkSuite of offline analysis programs
Detector characterization
At the hospital
UC Berkeley Medical Center
Singles Scan
Am source scans to determine position of segment boundaries.
X
Y114μCi Am source (60keV)
Vertical and horizontal collimators (2mm) to define x,y and z
Automatic scan on pre-determined x,y,z pattern.
1min per point
0
200
400
600
800
1000
1200
1400
0 10 20 30 40 50 60 70
Y (mm)
Cou
nts
60ke
V A
m
Singles Scan
Deviation from nominal values
dX =0.1+/-0.1 mm
dY =0.2+/-0.1mm
dα = 0 +/- 1 deg
ε3
central
Energy central contact
Crystal Orientation
Gate on Am
Coincidence Scans
Pulse shape measurement using a prompt coincidence requirement between GRETINA and Clover(s)
1mCi 137Cs source
Vertical and slit collimators to define 90 deg scattering
500nsec overlap
Coincidence trigger ~ 200 events/day
Coincidence Scans
A
Clover 2 crystals
369-379 keV
Eclover+EA= 662keV
Pulse Shapes
12
34
5
6
Δx =5mm
~ 200 events
Pulse Shapes
Tracking
First step – cluster finding
Any two points with θ < θp are groupedinto the same cluster
Tracking
Second step – Tracking of Compton scattering interaction points
Problem: 3!=6 possible sequences Assume: Eγ = Ee1 + E e2 + Ee3 ;γ-ray from the source
Sequence with the minimum χ2 < χ2 max
correct scattering sequencerejects Compton and wrong direction
Eγ
Ideal Shell
Let’s put it all together
Full analysis of simulated and experimental data
GEANT Signal calculation
Signalanalysis Tracking
Signaldigitizer
XYZE
XYZE
γ-ray
γ-ray
• Detector geometry
• Drift velocity• Electronics response
signals
signals
Spectra →Efficiency, P/T
Imaging
Field calculation
Electric field
Austin Kuhn, PhD Thesis, UC Berkeley, 2002.
• Eγ=0.662 MeV (137Cs), source distance= 12cm
• Signal analysis (least square method)• up to 4 segments and 2 interactions per
segment (98% of all events)
• For single interaction per segment• position determination <1.5 mmsuccess rate ~80%
• For two interactions per segment• position determination < 1.5 mm
success rate ~70%• minimum separation 2 mm
• Compared data and simulation with and without tracking
137Cs source
Simulation
Data
P/T ~ 38%
Relative Eff ~ 0.67
P/T ~ 31%
Relative Eff ~ 0.62
Tracking
P/T no tracking ~ 16%
“Battle” conditions
• 82Se + 12C @ 385 MeV• 90Zr nuclei (β ~ 8.9%)• 2055 keV (10+ 8+) in 90Zr• Target-detector @ 4 cm• Beam-detector @ 900
Experimental measurement of position resolutionDoppler broadening related to ΔrMaximize Doppler effect
In-beam Test of PII and PIII
θβ
γγ cos11 2
0
−−
= EEDoppler shift:
θ
γ-rayinteraction
position
Beam
Target
Digitized pulse shapes have been analyzed to extract energy and position of individual γ-ray interactions (signal decomposition ).
The position of the 1st
interaction was used to correct for Doppler shift.
Data Analysis
Comparison between the observed signals and a linearcontribution of pre-calculated basis signals:
Q(t)=Eiqi(t)+Ej qj(t)+…
Basis signalDetector signal
Signal DecompositionNew codeDavid Radford - ORNL
From simulations, 14.5 keV FWHM
σ = 2.4 mm (RMS in 3D)correct for segment contact
correct for 1st interaction
FWHM28.3 keV
FWHM14.8 keV
Results
Single segment
correct for central contact
correct for segment contact
correct for the first interaction
FWHM18 keV
FWHM28 keV
FWHM14 keV
Two segment
Results from PIII
GG R E T I N A = 1/4 of 4π “First stage of GRETA”
But not But not ““Little GRETALittle GRETA””
GGamma Ray Energy Tracking In beam Nuclear Array
Number of hexagons
Number of different hexagonal shapes
80 2 (20, 60)
110 3 (20, 30, 60)
120 2 (60, 60)
150 3 (30, 60, 60)
180 3 (60, 60, 60)
200 4 (20, 60, 60, 60)
12 pentagons and …
Geodesic tiling of the sphere
GS
AGATA
4π Simulations GEANT4
Quad
Triplet
E.Farnea and D.Bazzacco - Padova
- 7 quad modules + 1 triplet (31 crystals)- Cost of $17M- Construction period from 2005 – 2010
GRETINA
Aux. Det. Trigger
Global TriggerModule
SignalDigitizers
Local TriggerModule
30 Crystals
Data Storage
2.2 MB/s
Aux. Det. Data
66 MB/s
75 dual Processors
Workstations, Servers
6.9 MB/s
2.3 MB/s + Aux. Data
NetworkSwitch
Processing Farm
ReadoutComputer
Data Acquisition System
How does nuclear shell structure and collectivity evolve in exotic neutron-rich nuclei?
What is the influence of increasing charge on the dynamics and structure of the heaviest nuclei?
How do the collective degrees of freedom and shell structure evolve with excitation energy and angular momentum?
What are the characteristics of Giant Resonancesbuilt on excited states and loosely bound nuclei?
GRETINA Physics
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1 2 3 4 5 6 7 8
GS (lo)GS (hi)SeGA (lo)SeGA (hi)gretina (lo)gretina (hi)
Performance
Relative to a 4π arraysolid angle ~ 0.8, position resolution ~ 2mm
Low (lo) and high (hi) beam intensity
n-rich nuclei from fragmentation reactions
Gamma-ray energy (2keV/channel)
30Na from 32Al Beam
30Na from 30Mg Beam
340
370410250
175
190
150
140
340
370410
250
175
770430 (3+--2+)
Simulation SeGA Simulation GRETINA
30Mg (pn) → 30Na (100 MeV/u)v/c=0.43
charge exchange reactionGamma-gamma coincidence
NSCL data SeGA(E. Rodriguez-Vieitez et al.)
Performance 1π
High spin states from fusion reactions
Simulation GRETA, ε =0.25
4-fold, I=10-5
Simulation GS
Simulation GS, ε =0.09
3-fold, I=10-4
v/c=0.04
3-fold, I=10-3
64Ni ( 48Ca, 4n) 108Cd, Gammasphere
Performance 4π
DOE - Critical DecisionsCD0 : Mission need Aug. 2003
CD1 : Preliminary Baseline Range Feb. 2004
CD2A/CD3A : Performance Baseline range for long lead time items June 2005
• CD2B/CD3B : Start Construction Sept. 2007
• CD4 : Start of Operation May 2010
Schedule
• Argonne National Laboratory– Trigger system– Slow control software
• Michigan State University– Detector testing
• Oak Ridge National Laboratory– Liquid nitrogen supply system– Data acquisition
• Washington University– Target chamber
Collaborating Institutions
Physics M. A. RileyDetector A. O. MacchiavelliElectronics D. C. RadfordSoftware M. CromazAuxiliary Detectors D. G. Sarantites
Working Groups
http://grfs1.lbl.gov/
http://radware.phy.ornl.gov/greta/join.html
Summary
γ1 Large arrays of Ge detectors such as Gammasphere and Euroball had a large impact in nuclear structure research in the last decade.
γ2 A 4π Gamma-Ray Tracking Array appears as the next frontier and was identified as a new initiative by the nuclear structure communities in Europe and USA
γ3 GRETINA has just received CD2/3A approval from DOE.
γ4 With the AGATA demonstrator, first realization of a tracking device and offers a very compelling physics case
γ5 We look forward to an exciting research program.
Important applications in many fields
Kai Vetter -LLNL
High energy astrophysicsCorrelate the detected photon to the source object as known from more precise observations in other wavelengths
Biomedical researchPrecise localization of radioactive tracers in the bodyCancer diagnosisMolecular targeted radiation therapyMonitor changes in the tracer distribution -> dynamical studies
National securityNuclear non-proliferation/ nuclear counter terrorismContraband detectionStockpile stewardshipNuclear waste monitoring and management
Industrial non-destructive assessmentsDetermination of the material density distribution between the source and detector
Why imaging gamma-rays?
( )1
2011cos
EEEcmE−
−=γγ
θThe Compton scattering formula gives θ:
4321 EEEEE +++=γ
r1
r2r3
r4
θ
1E
γE
12rr
sourcesourceGamma rays interact several times with detector via Compton interaction (e.g. until it is stopped by the photo-electrical effect)
Measuring positions and energies of individual interactions enables to determine pathway of gamma ray in detector (tracking)
Energies and positions of first two interactions define cone of incident angles (electron path is not measured)
Cones are projected on plane or sphere (one circle per event) for 2D or into cube (one cone per event) for 3D imaging
source
Compton gamma-ray imaging
4π photocamera4π photocamera
K.Vetter et al.
4π gamma-ray imaging
Finally, to conclude, a word about the future …
“The future isn’t what it used to be.”
Arthur C. Clarke
“Prediction is very difficult, especially about the future.”
Niels Bohr
Shrinking budgets
Limited resourcesNeed to find new opportunities
However, with
Electric field
ionconcentratimpurityVVE
:2 ρρ=∇
∇=rr
Boundary condition : applied bias voltage
Weighting potential for segment k
02 =∇ kVBoundary condition : 1 V on the segment k
0 V on all other segments
Signal Generation
Trajectory : for electrons and holes
∫+=
=t
dtvxtx
Evv
00)(
)(rrr
vrv
Induced charge (S. Ramo, Proc. IRE 27(1939)584)
If a charge q moves from position x1 to position x2, then the induced charge on electrode k is
( ))()( 12 xVxVqQ kkkrr
−=Δ
anisotropic
Signal Generation
Measurements using the digitizer board
Energy Resolution
Crystal A
0
0.5
1
1.5
2
2.5
3
3.5
0 6 12 18 24 30 36segment #
FWH
M (k
eV)
Am 60keV60Co 1173keV60Co 1332keV
2.55 keV
1.28 keV
Segment number
σ (keV)
3.5 keV
Noise level at high-frequency
Calculate signal in each segment for interactions on a grid
base signals
Decompose the composite signal into alinear combination of base signals
Interpolate to improve position resolution
Signal decomposition
Results
152Eu full analysisGain in peak/total vs. efficiency
Neutron damage
Fast neutrons damage the Ge structure and create hole-traps.Reduction of charge collection efficiency.Signal loss depends on the interaction position.Hole-trapping has been introduced in the calculation through
attenuation length λ [T. Raudorf, R. Pehl, NIM A255 (1987)].λ ~ 1/φ and depends on bias, detector temperature, crystal type
⎟⎟⎠
⎞⎜⎜⎝
⎛ −−⋅=
h
rrnnλ
'exp0
For GRETINA: φ = 1010 n/cm2 λ = 10 cmφ = 109 n/cm2 λ = 100 cm
Initial number of holes
Attenuation length
Distance traveled
Neutron damage
Degradation in E resolution occurs for λ<100 cm, beforecorrection and for λ<30 cm, after correction, but only forλ<17 cm position resolutionbecomes worse than 1 mm (achievable limit).
A measurable effect of neutron damage on position resolution is never reached before annealing is required!
Results