Experimental Prospects for Direct
Measurements of Neutrino Mass
(Tritium Experiments)
Outline
Directness?
The Various Techniques: for completeness
Astrophysics/Cosmology (very short)
Nuclear and Particle Physics: heart of the talk
beta decay
Steve Elliott
Nuclear Physics
Direct vs. Indirect
Absolute vs. Relative
Kinematic vs. Interference
 No mn experimental technique is direct in that it
measures one of the mi. All experiments measure
parameters that depend on the mass eigenvalues.
 Many laboratory experiments might indicate the
“absolute scale of n mass”.
 These are experiments that search for a kinematic
effect due to mass.
 These are not experiments that search for an
interference effects. Oscillation experiments provide
data on the “relative mass scale”.
Steve Elliott, Tritium Beta Decay, NPSS 2005
2
Mass
Neutrino Mass: What do
we want to know?
Relative
Mass
Scale
Absolute
Mass
Scale
Dirac or Majorana
n 
 
n 
n 
n 

n  or 

 
 
 
n 
Steve Elliott, Tritium Beta Decay, NPSS 2005
ne
n1
2
Ue1
n2
2
Ue2
n3
2
Ue3
Mixing
3
A List of Upcoming Techniques
Supernovas - relatively poor sensitivity
Cosmology - hope of reaching below 100 meV
Nuclear/Particle Physics
t decay - relatively poor sensitivity
m decay - relatively poor sensitivity
b decay - hope to reach below 500 meV
bb decay - hope to reach below 50 meV
oscillations - great sensitivity to mass differences
Steve Elliott, Tritium Beta Decay, NPSS 2005
4
Matter Content of Universe
Steve Elliott, Tritium Beta Decay, NPSS 2005
5
Supernova Tests
50
40
Energy (MeV)
Spread of neutrino
arrival times can give
indication of mass.
SN1987a: about 20 eV
limit but conclusions
varied.
SN dynamics makes for
model dependencies.
Future sensitivity might
be a few eV.
Kamiokande II (PR D38 (1988) 448
IMB (PR D37 (1988) 3361
Baksan (PL B205 (1988) 209)
30
20
10
0
0
2
4
6
8
Time of Event (sec)
10
12
14
But Earth effects might be exploited
for q13 and sgn(dm2) measurments
Steve Elliott, Tritium Beta Decay, NPSS 2005
6
Supernova n Experiments
Detector
Type
Mass (kton)
Location
# events at 10 kpc
status
Super-K
H2O Cerenkov
32
Japan
7000
Running
SNO
Heavy Water
(salt)
1.4 H2O / 1 D2O
Canada
350, 450
Running
LVD
Scintillator
1
Italy
200
Running
KamLAND
Scintillator
1
Japan
300
Running
Borexino
Scintillator
0.3
Italy
100
Soon?
Baksan
Scintillator
0.33
Russia
50
Running
MINIBooNE
Scintillator
0.7
USA
200
Running
AMANDA
Ice
Meff ~ 0.4/PMT
South Pole
N/A
Running
Icarus
Liquid Ar
2.4
Italy
250
Soon
OMNIS
Pb, Fe
4, 1
USA
2000
Proposed
LANNDD
Liquid Ar
70
USA
6000
Proposed
UNO
H2O Cerenkov
600
USA
>100,000
Proposed
Hyper-K
H2O Cerenkov
1000
Japan
>100,000
Proposed
LENA
Scintillator
30
Europe
15,000
Proposed
Steve Elliott, Tritium Beta Decay, NPSS 2005
7
Cosmology
Measure Wnh2
•WMAP measured cosmological
parameters very precisely. This
allowed precise estimates of Wnh2
from LSS measurements.
•WMAP results indicate Smi <
about 1 eV. A very competitive
result. (one interpretation claims
Smi = 0.64 eV!)
•But, correlations between
parameters result in assumption
dependent conclusions.
•Want laboratory experiments.
Steve Elliott, Tritium Beta Decay, NPSS 2005
8
Cosmology - Future
Measurements
MAP/PLANCK CMB measurements with high
precision galaxy surveys (Sloan Digital Sky
Survey): Smi < ~300 meV
If weak lensing by LSS is also considered:
Smi < ~40 meV
Even with the correlations, cosmology will play an
important role in the interpretation of neutrino
mass.
Steve Elliott, Tritium Beta Decay, NPSS 2005
9
Z-burst and high-energy n
 Requires, as yet unknown (and unneeded?) flux of UHE n with
energy > Greisen-Zatsepin-Kuzmin (GZK) energy (5x1019 eV).
 Although could explain existence of cosmic rays with E>EGZK, it
doesn’t explain source of proposed UHE n.
 UHE n + cosmic relic n --> Z --> hadrons, gs
 mi near 500 meV could produce p or g just above GZK cutoff.
Model can be “tweaked” to get lower masses.
 Detect p or g : their multiplicity and energies relate to ER and
hence mi.


eV
R
21
En i  4.2  10 eV
mi 
Steve Elliott, Tritium Beta Decay, NPSS 2005
10
Nuclear and Particle Physics
Techniques
t decay: decays into 5 or 6 p most sensitive
because of restricted phase space for n.
t  np  n t
E* 
mh is invariant mass of n p.
E* is total p energy in t rest frame.
2
2
2
mt  m h  mi Obtain data on degenerate scale m1.
2mt
Steve Elliott, Tritium Beta Decay, NPSS 2005
mn 1  18.2 MeV
Eur. Phys. J. C2 (1998) 3
11
Nuclear and Particle Physics
Techniques
m decay: p --> mn
mn m  mp  mm  2mp mm  pm
2
2
2
2
2
2
2
mn m   Umi mi
2
mn  190 keV
m
PR D53 (1996) 6065
Steve Elliott, Tritium Beta Decay, NPSS 2005
12
The b Spectrum (bare
nucleus)
Energy-independent
Matrix Element
Fund. Constants
2
GF
3 7
dN

dE 2p
Fermi Function
cos q C M F(E, Z  1)
2
2
p(E  me )   mn (  mn )
2
Electron momentum,
Energy, mass
Steve Elliott, Tritium Beta Decay, NPSS 2005
 = E0 - E
2
Step Function
13
b Spectrum (atom or molecule)
plus neutrinos mix
Many possible final states.
j for state j
dN
2
 A M F(E, Z  1)p(E  me )
dE
 W j  j  Uei
j
2
2
j
2
 mi ( j
 mi )
i
Wj - probability for
transition to state j
Steve Elliott, Tritium Beta Decay, NPSS 2005
Uei - mixing
matrix elements
mi - n mass
eigenstate
14
b Spectrum
(atom or molecule)
When convolved with resolution function
with width > mi, we can analyze the
spectrum with one mass parameter: <mb>.
dN
2
 A M F(E, Z  1)p(E  me )
dE
 Wj j
2
j
 mb (  mb )
2
j
Steve Elliott, Tritium Beta Decay, NPSS 2005
15
The Neutrino Mass from b
decay
The shape of the b energy spectrum
near the endpoint depends on mn.
KATRIN LOI
mb 
3
2
2
 Uei mi  2.2 eV
i1
NP B (Proc. Suppl.) 91 (2001), 273
Steve Elliott, Tritium Beta Decay, NPSS 2005
16
187Re
b Decay Experiments
Low Q-value: 2.6 eV
•Long half-life = low specific activity
•Bolometric techniques = measure whole
spectrum
–But also entire energy deposit!!!
•Measure 10-10 of decays near endpoint, but
bolometer response time is 100s ms.
•Future sensitivity should be about 10 eV.
Steve Elliott, Tritium Beta Decay, NPSS 2005
17
Milano Re Experiment
AgReO4
Hope for a sensitivity of 4 eV
Steve Elliott, Tritium Beta Decay, NPSS 2005
18
Genoa Re Experiment
PR C63 014302
Genoa: Metallic Re, mn< 26 eV
Steve Elliott, Tritium Beta Decay, NPSS 2005
NP B (proc. Suppl.) 91 (2001) 293
19
Why Tritium?
•Re bolometers (source =
detector) collect whole
spectrum at once: therefore
need low rate to prevent
pile-up.
•With source ≠ detector,
one can just analyze end of
spectrum: much higher
activities.
•Very low Q-value (2.5 keV)
•Low Q-value (18 keV)
Steve Elliott, Tritium Beta Decay, NPSS 2005
20
Mainz Experiment Setup
Steve Elliott, Tritium Beta Decay, NPSS 2005
21
Mainz Results
mn  2.8 eV (95% CL)
DE ~2-6 eV at 20 keV
Large acceptance
10-40% of 4p
Improvements 94-98
Lower temp source
no dewetting
T2 evaporating into
spectrometer stopped
by tilted solenoids.
Steve Elliott, Tritium Beta Decay, NPSS 2005
22
Fit Function
Fit Parameters: Free Amplitude, Endpoint
energy, neutrino mass squared, and
background
Response function depends on: Potential
distribution within source, backscattering
from source substrate, spectrometer
transmission, and energy dependence of
detection efficiency.
Steve Elliott, Tritium Beta Decay, NPSS 2005
23
Systematic Uncertainties
Inelastic scattering in tritium film
Neighbor molecule excitation
Final states in THe+ molecule
Charging of source film
Statistical Systematic
mn  3.7  5.3  2.1 eV
2
Steve Elliott, Tritium Beta Decay, NPSS 2005
2
24
The Troisk Experiment
Steve Elliott, Tritium Beta Decay, NPSS 2005
25
Troisk Results
mn  2.5 eV (95% CL)
Solid line is
spectrum with
step funcion.
Steve Elliott, Tritium Beta Decay, NPSS 2005
26
Troisk Anomaly
A “peak” shifting
periodically in time
would match the
data.
In a integrating
spectrum, a peak
would appear as a
step.
Steve Elliott, Tritium Beta Decay, NPSS 2005
27
Identifying the Systematics
Mainz: tritium as thin film on flat surface. Temperature
activated roughening led to microcrystals in turn
leading to large inelastic scattering.
Troisk: Gaseous tritium. Large angle scattering of
electrons trapped in source.
Addressing these issues removed the “negative mass
squared” in those experiments.
Steve Elliott, Tritium Beta Decay, NPSS 2005
28
Primary Systematic
Uncertainties
Resolution error
leads to mn error
d mb
2
KATRIN LOI
mb
dN
2
 
dE
2
dN
 2   2
dE
2
 2 2
Steve Elliott, Tritium Beta Decay, NPSS 2005
29
Tritium b decay Experiments
KATRIN
Very big spectrometer using
gaseous and thin sources. A
big step forward.
Univ. of Texas-Austin
t2 source in magnetic
free environment.
Steve Elliott, Tritium Beta Decay, NPSS 2005
30
The MAC-E Filter
dE
Bmin

E Bmax
Magnetic Adiabatic
Collimation followed
by an Electrostatic
Filter
High luminosity
Low background
Good energy resolution

Integrating
high-pass filter
Steve Elliott, Tritium Beta Decay, NPSS 2005
KATRIN LOI
31
KATRIN
1-eV resolution (x4 better)
Electron transport system guides b to pre-spectrometer
while preventing tritium flow to spectrometers.
Pre-spectrometer filters out all b except those near
endpoint. Keeps residual ionization minimized hence
reducing background.
Large analyzing plane increases signal rate
Si drift detectors. 600 eV resolution for 18.6 keV b. Good
electron sensitivity but low efficiency for g.
Steve Elliott, Tritium Beta Decay, NPSS 2005
32
KATRIN (LOI version)
KATRIN will be sensitive to about 200 meV. Thus if
the mi follow a degenerate pattern and m1 is within the
sensitivity, the experiment may see <mb> = m1.
Steve Elliott, Tritium Beta Decay, NPSS 2005
33
Summary of Mass Measurements
Future Results
1400
Range for S for a given
normal heirarchy, qCHOOz=0
7000
b b mass
Current Status
6000
1200
S (meV)
5000
1000
S (meV)
tritium b limit
bb limit
tritium b limit
800
4000
bb limit
3000
2000
1000
600
WMAP limit
0
0
400
Plot a la PL B532, 15 (2002)
(with a guess at the future)
200
400
600
m bb (meV)
800
1000
SDSS limit
200
0
0
100
200
300
400
m bb (meV)
Steve Elliott, Tritium Beta Decay, NPSS 2005
500
600
Absolute scale measures
b: 350 meV
bb: 50 meV
Cosmology: <100 meV
34
A summary of the questions
Are neutrinos Majorana or Dirac?
What is the absolute mass scale?
How small is q13?
How maximal is q23?
Is there CP violation in the neutrino sector?
Is the mass hierarchy inverted or normal?
Is the LSND evidence for oscillation true? Are
there sterile neutrinos?
Steve Elliott, Tritium Beta Decay, NPSS 2005
35
References
Katrin LOI
Lobashev et al. PL B460 (1999) 227
Weinheimer et al. PL B460 (1999) 219
Bilenky Review
hep-ph/0211462 v3
Steve Elliott, Tritium Beta Decay, NPSS 2005
36