The Majorana Project: A Next-Generation Neutrino Mass Probe : The Majorana Project: A Next-Generation Neutrino Mass Probe Craig Aalseth for The Majorana Collaboration
http://majorana.pnl.gov
APS Neutrino WG Meeting, Pasadena, California
February 28, 2004
The Majorana Collaboration : The Majorana Collaboration Brown University, Providence, RI Rick Gaitskell
Duke University, Durham, NC Werner Tornow
Institute for Theoretical and Experimental Physics, Moscow, Russia A. Barabash, S. Konovalov, V. Stekhanov, V. Umatov
Joint Institute for Nuclear Research, Dubna, Russia V. Brudanin, S. Egorov, O. Kochetov, V. Sandukovsky
Lawrence Berkley National Laboratory, Berkley, CA Yuen-Dat Chan, Martina Descovich, Paul Fallon, Reyco Henning, Kevin Lesko, Augusto Macchiavelli, Akbar Mokhtarani, Alan Poon, Craig Tull
Lawrence Livermore National Laboratory, Livermore, CA Kai Vetter
Los Alamos National Laboratory, Los Alamos, NM Ted Ball, Steve Elliott , Victor Gehman, Andrew Hime
North Carolina State University, Raleigh, NC
Jeremy Kephart, Ryan Rohm, Albert Young
Oak Ridge National Laboratory, Oak Ridge, TN
Cyrus Baktash, Thomas Cianciolo, Robert Grzywacz,
David Radford, Krzysztof Rykaczewshi
Osaka University, Osaka, Japan
Hiro Ejiri, Ryuta Hazama, Masaharu Nomachi
Pacific Northwest National Laboratory, Richland, WA Harry Miley (Project Director), Craig Aalseth, Ronald Brodzinski, Shelece Easterday, Todd Hossbach, David Jordan, Richard Kouzes, William Pitts, Ray Warner
Queens University, Kingston, Canada Art McDonald, Aksel Hallin
University of Chicago, Chicago, IL Juan Collar, Andrew Sonnenschein
University of Tennessee, Knoxville, TN
Yuri Efremenko
University of South Carolina, Columbia, SC Frank Avignone, George King
University of Washington, Seattle, WA
Peter Doe, Kareem Kazkaz, R.G. Hamish Robertson, John Wilkerson
Outline : Outline Introduction and Overview
Reference Concept
Configuration
Materials
Backgrounds and Mitigation
Pulse-Shape Discrimination
Detector Segmentation
Experiment Sensitivity
Progress and Status
Conclusions
Germanium Basics : Germanium Basics “Internal Source Method” from Fiorini
76Ge: Endpoint = 2039 keV
Energy above many contaminants
Except: 208Tl, 60Co, 68Ge…
FWHM = 3-4 keV around 2 MeV (~0.2%)
Long experience with Ge bb decay
Previous efforts found 2n at T1/2 ~1021 y
Expect 0n at T1/2 ~ 4 x 1027 y
Ready to go!
Essentially no R&D needed
Majorana Overview : Majorana Overview GOAL: Sensitive to effective Majorana n mass near 50 meV
0n bb decay of 76Ge potentially measured at 2039 keV
Based on well known 76Ge detector technology plus:
Pulse-shape analysis
Detector segmentation
Requires:
Deep underground location
500 kg enriched 86% 76Ge
many crystals, segmentation
Pulse shape discrimination
Time/Spatial Correlation
Special low-background materials
Ge Acquisition : Ge Acquisition Purchase enriched, intrinsic 76Ge
ECP in Russia supplied previous experiments
86% enrichment
Produced at 200 kg/y, delivered quarterly
About 4 years of deliveries
Need to do full market study
Detector Fabrication : Detector Fabrication Need to build about 50 kg of segmented detectors quarterly
About 4-5 years of detector construction
Would prefer to grow crystals and fabricate detectors underground
Detector Configuration : Detector Configuration Granularity, low passive mass, are goals
Optimization underway of performance and risk
Several low-risk designs possible for modular cryostats
Many segmentation schemes possible and equally effective
Nature of Ge crystals allows repackaging
Low-Background Electroformed Copper : Low-Background Electroformed Copper Can be easily formed into thin, low-mass parts
Recent designs reduce mCu/mGe x5
UG Electroforming can reduce cosmogenics
Pre-processing can reduce U-Th
Recent results suggest cleaner than thought Electroformed cups shown have wall thickness of only 250 mm!
Passive Shielding : Passive Shielding Inner
Need about 5 tons of clean, ancient lead
PNNL has identified most of this required need
Outer
About 60 tons of cleaned lead bricks
Electroformed Cu will allow roof span
External Active Shield : External Active Shield Plastic scintillator and phototube will provide ~4p coverage
2-4” plastic scintillator
May want alternating Pb/plastic layers to improve fast neutron tagging
Majorana Reference Concept : Majorana Reference Concept Optimization underway of performance and risk
Several low-risk designs possible
Many segmentation schemes possible
Alternative packaging, cooling, shielding under consideration
Nature of Ge crystals allows repackaging 3-D model of reference configuration
Starting Background Estimate : Starting Background Estimate International Germanium Experiment (IGEX) achieved between 0.1-0.3 counts/keV/kg/y
Documented experiences with cosmic secondary neutron production of isotopes
Calculated
Experimental [Bro95] R. L. Brodzinski, et al, Journal of Radioanalytical and Nuclear Chemistry, Articles, Vol. 193, No. 1 (1995) 61-70.
Physics Motivation for Background Rejection : Physics Motivation for Background Rejection Interaction multiplicity varies with energy, type of radiation
Multiple interactions change shape of induced detector current G. F. Knoll. Radiation detection and measurement. John Wiley & Sons, Inc., second edition, 1989. Multi-site events
Full-energy gamma signals >500 keV
Sum-energy peak signals
Single-site events
Gamma signals <150 keV
72Ge(n, n’ e-) fast neutrons
Double-escape peaks
Internal beta-decay (no g) events
Backgrounds: 68Ge : Backgrounds: 68Ge
Backgrounds: 60Co : Backgrounds: 60Co
Segmentation & Pulse-Shape Discrimination : Segmentation & Pulse-Shape Discrimination Allow rejection of multiple-site interactions
Effective against projected backgrounds
Granularity Costs Money/Should be optimized
Different Schemes Being Evaluated
Segmented Large N-type Crystals
Multiple Small P-type Crystals
Segmented Large P-type Crystals
Why Now is a good time for PSD…. : Why Now is a good time for PSD…. Commercial digital spectroscopy hardware is available with fast (40 MHz), high-resolution (14-bit) digitization
Significant developments in pulse-shape discrimination techniques for HPGe have been made in the past 10 years and are ready to apply to new hardware Full-energy 1621-keV g (top) and 1592-keV DEP (bottom) reconstructed current pulses from 120% P-type Ortec HPGe detector (experimental data) Signal Background
Single-site interaction example : Single-site interaction example Monte-Carlo 2038-keV deposition from 0n bb-decay of 76Ge
Multiple-site interaction example : Multiple-site interaction example Monte-Carlo 2038-keV deposition from multi-Compton of 2615-keV 208Tl g
Multi-Parametric Pulse-Shape Discriminator : Multi-Parametric Pulse-Shape Discriminator Extracts key parameters from each preamplifier output pulse
Sensitive to radial location of interactions and interaction multiplicity
Self-calibrating – allows optimal discrimination for each detector
Discriminator can be recalibrated for changing bias voltage or other variables
Method is computationally cheap, requiring no computed libraries-of-pulses
PSD can reject multiple-site backgrounds (like 68Ge and 60Co) : PSD can reject multiple-site backgrounds (like 68Ge and 60Co) Keeps 80% of the
single-site DEP (double escape peak) Rejects 74% of the multi-site backgrounds (use 212Bi peak as conservative indicator) Improves T1/2 limit by 56% Experimental Data Original spectrum Scaled PSD result
Detector Segmentation : Detector Segmentation Sensitive to axial and azimuthal separation of depositions
Example design with six azimuthal and two axial contacts in a 2-kg detector
This segmentation gives ~2500 segments of 200 g (or 40 cm3) each
Many segmentation schemes give equivalent good background rejection
Monte-Carlo Example (single crystal) : 0nbb efficiency = 91% internal 60Co efficiency = 14%
Improves T1/2 limit by 140% Monte-Carlo Example (single crystal) Sensitive to z and phi separation of depositions Segment multiplicity at 2039 keV Next Steps:
T ½ improvement increases to 260% - 620% when including array self-shielding, depending on position of crystal – not included in earlier background estimate
Time-series analysis of background very promising
Electronics : Electronics Default plan is to use XIA digitizers.
Other commercial options exist or are emerging. CAMAC/CPCI/etc.
Underground Laboratory : Underground Laboratory Would prefer NUSEL
No “foreign” travel
Less customs issues
Very deep (fewer fast neutrons)
Have had very positive interactions with SNOLab
WIPP is available but not as deep as we prefer
Sensitivity vs. Time : Sensitivity vs. Time Slow Production: Gradual ramp to 100 kg/y - total 500 kg 85% 76Ge
Fast Production: 200 kg/y (No ramp)
Present 0nbb 76Ge T1/2 limit rapidly surpassed (T1/2 > 1.9 1025 y) 0nbb Half-Life
Examples of signal intensity : Examples of signal intensity T1/2 = 1 1026y
28 counts
1000 kg-y T1/2 = 1.5 1025y
95 counts
50 kg-y
Collaboration Progress: Optimizations for Full Experiment : Collaboration Progress: Optimizations for Full Experiment Segmented Enriched Germanium Array (SEGA):
Segmented Ge Multi Element Germanium Assay (MEGA):
16+2 natural Ge High density
Materials qualification
Cryogenic design test
Geometry & signal routing test
Powerful screening tool g Full Experiment MAJORANA:
500 kg Ge detectors
All enriched/segmented
Multi-crystal modules 1 to 5 Crystals
First enriched, segmented detector in testing!
Additional tests being planned for other segmented systems
Progress and Status SEGA: Segmentation Optimization : Progress and Status SEGA: Segmentation Optimization First (enriched) 6x2 SEGA operating
Current: Testing (TUNL)
Shallow UG testing at U Chicago LASR facility
Operation in WIPP
Second and third SEGA planning
Funds in hand (LANL, USC)
Alternate segmentation testing underway (USC/PNNL)
40-fold-segmented LLNL detector now available
Figure-Of-Merit vs. Axial & Azimuthal Segmentation for internal 60Co background SEGA crystal initial test cryostat
Progress and Status Ultra-Low Level Screening : Progress and Status Ultra-Low Level Screening Screening facility
Operating in Soudan (Brown U)
Two HPGE detectors (1.05 kg, 0.7 kg)
Planned use for screening
Minor materials used in manufacturing
Improved Cu testing
Small parts qualification
FET, cable, interconnects, etc Dual counter shield: 1.05 and 0.7 kg detectors
Progress and Status MEGA: Cryogenic testing : Progress and Status MEGA: Cryogenic testing Materials in hand
Detectors (20 - 70% HPGe), electronics
Assembly and cryogenic testing of two-crystal modules underway (PNNL, UW, NC State)
Underground facility (WIPP) in prep (LANL, NMSU)
FY 2004 installation anticipated
Sensitive to ~104 short-lived atoms
MEGA Infrastructure at WIPP : MEGA Infrastructure at WIPP Steel sub-floor to support many-ton lead shields
Cleanroom enclosure with antechamber entry
Power, network connectivity
MEGA Underground Site : MEGA Underground Site MEGA will be installed in WIPP
2150 feet of overburden
Q Room Alcove
LANL providing underground clean room (Steve Elliott and Co.)
Module Assembly & Testing : Module Assembly & Testing (MEGA)
Recent Crystal Packaging Test (MEGA) : Recent Crystal Packaging Test (MEGA)
Conclusions : Conclusions Reference Plan meets sensitivity goals
Opportunities for enhancements exist
Potential for discovery
Unprecedented confluence:
Enrichment availability/Neutrino mass interest/ Underground facility development
High Density:
Modest apparatus footprint, no special cavity required
Low Risk:
Proven technology/ Modular instrument / Relocatable
Early results / Incremental deployment
Experienced and Growing Collaboration
Long bb track record, many technical resources
End : End
Majorana and GENIUS : Majorana and GENIUS If no signal seen:
If background (i.e. within Ge mass), then:
Leads to emphasis on
enrichment
balanced mass and time
background rejection
If we assume background is in structural materials only, B=bT then:
Leads to emphasis on
minimized structure
increased natural mass
decreased time Suppressing constants: “Majorana Approach” “GENIUS Approach”
Evidence of 68Ge : Evidence of 68Ge From: NIM A292 (1990) 337-342. Experimental data from two 1.05 kg natural detectors Integral of this spectrum equals integral of this peak. This peak decays with the right half life.
Cosmogenics in Ge crystal account for ~all of early IGEX signals at 2039 keV : Cosmogenics in Ge crystal account for ~all of early IGEX signals at 2039 keV ~70% = 68Ge
~10% = 60Co 77 d 71 d 271 d 5.2 y Early IGEX Data
(Computed) From: Journal of Radioanalytical and Nuclear Chemistry, Articles, 193 1 (1995) 61-70
Background in Structural Materials : Background in Structural Materials We (Jim Reeves) have improved the chemistry for electroformed Cu production
U, Th progeny reduced substantially
Plan to totally eliminate 60Co NIM A292 (1990) 337-342. Early work showing cosmogenics ~1995 to present
Electroformed copper
radiochemistry gains:
H2SO4 Purity
Recrystalized CuSO4
Barium scavenge
Results:
226Ra <25 mBq/kg
(< 1 part in 71019)
(< 2.0 ppt 238U eq.)
228Th 9 mBq/kg
(1 part in 31021)
(2.2 ppt 232Th eq.)
(From Brodzinski et al, Journal of Radioanalytical and Nuclear Chemistry, 193 (1) 1995 pp. 61-70) Worst case estimate today: non-cosmogenics not a show stopper Ge Cu Both
What can we do to improve on this? : What can we do to improve on this? Exploit the single-site (dbd) vs. multi-site (bkg) nature of energy deposition
Pulse shape discrimination: works on “R” separation of depositions
Segmentation: works on “Z” and/or “f” separation
Minimize materials bkg
Conventional but hi capacity cryostats
<1/5 Cu per Ge as IGEX
UG Electroforming
Repeated purification
Alternative cooling approaches
Possibly UG crystal manuf.
~1980 ~1990
Experimental Examples : Experimental Examples Commercial digital spectroscopy hardware is available with fast (40 MHz), high-resolution (14-bit) digitization
Significant developments in pulse-shape discrimination techniques for HPGe have been made in the past 10 years and are ready to apply to new hardware Full-energy 1621-keV g (top) and 1592-keV DEP (bottom) reconstructed current pulses from 120% P-type Ortec HPGe detector (experimental data)
Majorana and GENIUS Conclusion : Majorana and GENIUS Conclusion Both (Maj & Gen) extreme views in background location (crystal and structural materials) are probably somewhat wrong
Reducing structural mass is a good idea for Majorana
PSD and segmentation help with structural background
Structural background concentration can be reduced by
UG electroforming (60Co and short lived cosmogenics)
Multiple reagent purification steps
Re-crystallizing and sub-boiling distillation
Many GENIUS risks can be eliminated or minimized by clever design and operation
Alternative Packaging : Alternative Packaging Would boost crystal to crystal background suppression
Could take advantage of new shielding opportunities
Might use alternative cooling methods
Would require methods for progressive commissioning and periodic maintenance <1m <1m
Majorana DM Sensitivity Majorana dark matter sensitivity similar to and complementary with CDMS-II : Majorana DM Sensitivity Majorana dark matter sensitivity similar to and complementary with CDMS-II Projected 95% C.L. Majorana for an assumed low-energy background of 0.005 counts/keV/kg/day, one order of magnitude lower than in present detectors
Assumes ionization threshold of 1 keV, and SEGA and MEGA limits are < 1 kg-y
Majorana limits are calculated for the total exposure of 5000 kg-y
Dots represent plausible supersymmetric neutralino WIMP candidates Signal to noise analysis (SEGA MEGA)
Annular Modulation (Majorana)