Direct Determination of Neutrino Mass
• Beta Decay
– Tritium
– 187Re
– Other ideas?
•
•
•
•
•
• The mass is needed for
• Particle physics
• Interpretation of supernova n signal
• Cosmology
Neutrino Oscillations
Supernova timing
Double beta decay
Cosmology
Z-bursts
Hamish Robertson -- Carolina Symposium 5/08
Masses linked by oscillations
Average mass
> 20 meV
m122
m232
?
Present
Lab Limit
2.3 eV
Claim of Evidence for 0n in 76Ge
<m> ~ 0.2 to 0.3 eV
Single-site events in detectors 2,
3, 4, 5 (56.6 kg-y).
Looks good to me…
H.V. Klapdor-Kleingrothaus, Int. J. Mod. Phys.
E17, 505 (2008)
Beta decay and neutrino mass
3H
Requirements:
• Strong source
Task:
3
187
• Excellent energy resolution Investigate H or Re endpoint with sub-eV precision
• Small endpoint energy E0
KATRIN Aim:
• Long term stability
Improve mn sensitivity tenfold (2eV  0.2eV )
• Low background rate
What are we measuring?
Writing the transition probability as the matrix element of some operator T,
In the degenerate regime where all the masses are the same, the unitarity
of U gives us back the original expression for a single massive neutrino,
an “electron neutrino with mass”
2


m
2
n
 pe E(E  E 0 ) 1
2 
 (E  E 0 ) 
1/ 2

Final Mainz Result -- Kraus et al. hep-ex/0412056
Improved S/N tenfold over
1994 data
20 weeks of data in 1998,
1999, 2001
Stable background: pulsed
RF clearing field applied at
20-s intervals
The next generation experiment
Assemble everyone who has done a tritium experiment. Then…
aim :
improve mn by one order of magnitude (2 eV  0.2 eV )
requires : improve mn2 by two orders of magnitude (4 eV2  0.04 eV2 )
problem : count rate close to ß-end point drops very fast (~dE3)
• improve statistics :
- stronger tritium source (factor 80) (& large analysing plane, Ø=10m)
- longer measuring period (~100 days  ~1000 days)
• improve energy resolution :
- large electrostatic spectrometer with E=0.93 eV (factor 4 improvement)
- reduce systematic errors :
- better control of systematics, energy losses (reduce to less than 1/10)
KATRIN
at Forschungszentrum Karlsruhe
unique facility for closed T2 cycle:
Tritium Laboratory Karlsruhe
TLK
~ 75 m long with 40 s.c. solenoids
5 countries
13 institutions
100 scientists
Windowless Gaseous T2 Source
Principle of MAC-E Filter
Adiabatic magnetic guiding
of ´s along field lines
in stray B-field of
s.c. solenoids:
Bmax = 6 T
Bmin = 3×10-4 T
Energy analysis by
static retarding E-field
with varying strength:
High pass filter with
integral  transmission
for E>qU
KATRIN Experiment
70 m
Rear
Transp/Pump
Source
3H
β-decay
Main spectrometer
Detector
ne
e-
e-
1010 e- /s
1010 e- /s
3He
Source:
Provide the
required tritium
column density
e-
e-
103 e- /s
1 e- /s
3He
3He
3•10-3 mbar
- 1 ± 1 kV
Rear System:
Monitor source
parameters
Pre-spectrometer
0 kV
10-11 mbar
- 1 - 18.4 kV
Transp. & Pump system:
Transport the electrons,
adiabatically and reduce
the tritium density
significantly
Pre-spectrometer:
Rejection of low-energy
electrons and adiabatic
guiding of electrons
10-11 mbar
-1 - 18.574 kV
Main-spectrometer:
Rejection of electrons
below endpoint and
adiabatic guiding of
electrons
Detector:
Count electrons
and measure their
energy
Pre-spectrometer
Parameters:
•Length: 3.4 m (flange to flange)
•Diameter:1.7 m
•Vacuum: < 10-11 mbar
•Material: Stainless steel
•Magnets: 4.5 T
Status:
•Vacuum 7•10-11 mbar (without getter)
•Outgassing 7•10-14 mbar l/ s cm2
•Measurements in progress
Tandem design: -- Pre-filter, Energy analysis
Pre-spectrometer
Filters low energy -decay electrons E<18.4 keV
Moderate energy resolution E80 eV
Test bed for vacuum, electrode design, detector.
Main spectrometer
23 m long, 10 m diameter
High luminosity: dN/dt~Aspect.
 high energy resolution: E/E~Aspect
Vacuum 3x10-11 mbar (reduce backgrounds)
use non-evaporable getter pumps
Inner wire electrode (shape field, reduce backgrounds)
External air coil - compensate for Earths magnetic field
Detector
145 pixel Si PIN diode
~1keV resolution
Image source
 systematics
 backgrounds
Main Spectrometer Manufacture
2 conical end pieces
assembly hall DWE
1 cylindrical centre piece
Voyage of the main spectrometer
Arrival in Leopoldshafen: Nov 24, 2006
Inside the Spectrometer
Detector Section (Univ. of Washington, MIT)
Silicon PIN diode detector
•
9 cm active diameter
•
500 m thick
•
148 segments
detector magnet
B = 3-6 T
“flux tube”
e-
pinch magnet
B=6T
post-acceleration
electrode
shielding
& veto
Pixelized Detector Corrects Focal Plane Resolution
KATRIN Statistical Sensitivity
• Improved over
original design (7
m diameter main
spectrometer, source
luminosity)
• Reduction in
background
• Only shows statistical
uncertainty
Optimized run time at each energy
Tritium Beta Decay History
m  2
2
v
2
A window to work in
Molecular Excitations
Precision Voltage Divider test at PTB, 2006
Improved sensitivity with larger system
Discovery
90% CL UL
Mass Range Accessible
KATRIN
Average mass
> 20 meV
m122
m232
Present
Lab Limit
2.3 eV
Microcalorimeters for 187Re ß-decay
MIBETA: Kurie plot of 6.2 ×106 187Re ß-decay events (E > 700 eV)
MANU2 (Genoa)
metallic Rhenium
m(n) < 26 eV
Nucl. Phys. B (Proc.Suppl.) 91 (2001) 293
10 crystals:
8751 hours x mg (AgReO4)
MIBETA (Milano)
AgReO4
m(n) < 15 eV
Nucl. Instr. Meth. 125 (2004) 125
E0 = (2465.3 ± 0.5stat ± 1.6syst) eV
mn2 = (-112 ± 207 ± 90) eV2
MARE (Milano, Como,
Genoa, Trento, US, D)
Phase I : m(n) < 2.5 eV
hep-ex/0509038
KATRIN outlook
• KATRIN can measure neutrino mass directly via
kinematics of beta decay -- model independent
• Improvement of order of magnitude over
previous best
• Challenging goal of mn < 0.2 eV
(90%
C.L.) looks achievable
• German funding (33.5 M€) is in place
• US DOE funding ($2.6 M) is in place
• Initial operation 2010.
Thanks, Peter
Fin
Supernova Neutrino Time-of-flight
For a supernova at distance D (in 10 kpc) the time delay for a neutrino of
mass m (eV) and energy E (MeV) is:
Beacom & Vogel
hep-ph/9802424
The delay must be ~ the duration of the neutrino signal to avoid
model dependence at short times and not to be drowned in
background at long times. For a 1 eV result with 30-MeV neutrinos,
need D = 175 Mpc. Scaling Kamiokande for the same rate as
SN1987a, detector mass must be 12 Gt.
IceCube will be “only” 1 Gt, and not very sensitive at these low
energies.
Z-bursts
Gelmini, Varieschi & Weiler, hep-ph/0404272
Hypothesis: the extreme-energy
CR spectrum is produced by
neutrinos from distant sources.
The neutrinos can annihilate at the
Z pole on relic neutrinos to
produce the observable EE CR.
(A GZK-style cutoff for neutrinos).
If cutoff is at 2 x 1020 eV, then
mn > 20 eV, in disagreement
with expt. EE CR thus likely
not neutrino Z-burst debris.
Abbasi et al., PRL 92, 151101
Future tritium measurements?
• Ultimate sensitivity of spectrometers
– require instrumental resolution of ~ E /m
e
n
– Linear size X of instrument scales with resolution:
• Differential spectrometers
X
• Integral spectrometers
 Ee /mn
X in
 theElast
e /m
spectral fraction per decay
mnn of the spectrum is ~ (mn/Eo )3

–
– source thickness is set by the inelastic scattering cross-section (3.4 x
10-18 cm2 ), n ≤ 1. Can’t make it thicker, only wider.
– If one wants ~1
event/day in last mn of the spectrum
magnetic spectrometer mn ~ 1.7 eV
• for a 10 m
• for a 3 m dia. solenoid retarding field spectrometer mn ~ 0.3 eV
KATRIN is probably the end of the road for tritium beta decay
NEXTEX
Detector
Cryopump
Deflecting &
Focusing
Analyzer
U of Texas (1 M$)
Pure electrostatics
Possibility for electron diffr.
no magnetic fields <0.1 mG
Mu-Metal
Cylindrical
Mirror
Analyzer
Sensitivity ~ 0.8 eV
Tritium
Getter
Spherical
Deflecting
Analyzer
Required funding 6.5M$
Cold Finger
Tritium Cell
Liquid
Nitrogen
Baffles
Electron Gun
Diffusion Pumps
Not funded, it’s over
Sensitivity with run time
Systematic Uncertainties
Status of KATRIN Hardware Activities
Pre-spectrometer magnets
Delivered in February 2005
Main spectrometer
Final designs by MAN-DWE
Delivered December 2006
WGTS
Manufacturing Started
Delivery 2007
DPS2-F
Manufacturing started
Delivery 2007
Pre-spectrometer
Delivered in Oct. 2003
Vacuum tests started May 2004
El. mag. test start 2006
MC Using Stopping Power (M. Steidl)
Arrows show 99% intensity windows
Figure-of-merit
2 mHz
“Better”
1 mHz
QuickTime™ and a
decompressor
are needed to see this picture.
Final States
Red: 3HeT+
Blue: 3HeH+
1% uncertainty in rovib spectrum  m2 = 6 x 10-3 eV2
Even small mn influences structure
Smn
2DF Galaxy Survey
PRL 89 061301
2.4 eV
0.5 eV
0
Large
Scale
Small
Scale
Minimum Neutrino
and Flavor
Neutrino
massMasses
spectrum
andContent
flavor content
Mass (eV)
e
mu tau
Atmospheric
n3
0.058
0.050
0.049
m232
n2
n1
Solar
n2
n1
Solar
0.009
m122
0
?
Atmospheric
0
n3
?