Experimental%20Investigation%20of%20Geologically%20Produced%20Antineutrinos%20with%20KamLAND

1
Experimental Investigation of Geologically
Produced Antineutrinos with KamLAND

• Stanford University
• Department of Physics
• Kazumi Ishii

2
Outline

• Geologically Produced Antineutrinos
  (Geoneutrinos)
• KamLAND
• Background Events
• Results

3
Structure of the Earth

• Seismic data splits Earth into 5 basic regions
  core, mantle, oceanic crust, continental crust,
  and sediment.
• All these regions are solid except the outer core.

Image by Colin Rose and Dorling Kindersley
4
Convection in the Earth
Image http//www.dstu.univ-montp2.fr/PERSO/bokelm
ann/convection.gif

• The mantle convects even though it is solid.
• It is responsible for the plate tectonics and
  earthquakes.
• Oceanic crust is being renewed at mid-ocean
  ridges and recycled at trenches.

5
Total Heat Flow from the Earth
Bore-hole Measurements

• Conductive heat flow measured from bore-hole
  temperature gradient and conductivity
• Deepest bore-hole (12km) is only 1/500 of the
  Earths radius.
• Total heat flow 44.2?1.0TW (87mW/m2), or 31?1TW
  (61mW/m2) according to more recent evaluation of
  same data despite the small quoted errors.

Image Pollack et. al
6
Radiogenic Heat

• 238U, 232Th and K generate 8TW, 8TW, and 3TW of
  radiogenic heat in the Earth

• Beta decays produce electron antineutrinos

7
Urey Ratio and Mantle Convection Models

• Urey ratio indicates what fraction of heat
  dissipated comes from radiogenic heat. Urey ratio
  can be defined as
• Some mantle convection models predict
• Urey ratio gt 0.7.

8
Discrepancy?

• The measured total heat flow, 44 or 31TW, and the
  estimated radiogenic heat produced in the mantle,
  13TW, gives Urey Ratio 0.3 or 0.5.
• Problem with
• Mantle convection model?
• Total heat flow measured?
• Estimated amount of radiogenic heat production
  rate?
• Geoneutrino can serve as a cross-check of the
  radiogenic heat production.

9
Geoneutrino Signal

• KamLAND is only sensitive to antineutrinos above
  1800keV
• Geoneutrinos from K decay cannot be detected with
  KamLAND.

10
U and Th in the EarthChondritic Meteorites

• U and Th concentrations in the Earth are based on
  measurement of chondritic meteorites.
• Chondritic meteorites consist of elements similar
  to those in the solar photosphere.
• Th/U ratio is 3.9
• Th/U ratio is known better than the absolute
  concentrations.

11
U and Th Distributionsin the Earth

• U and Th are thought to be absent from the core
  and present in the mantle and crust.
• The core is mainly Fe-Ni alloy.
• U and Th are lithophile (rock-loving), and not
  siderophile (metal-loving) elements.
• U and Th concentrations are the highest in the
  continental crust and continental sediment.
• Mantle crystallized outward from the core-mantle
  boundary.
• U and Th prefer to enter a melt phase.

12
Reference Earth ModelConcentrations of U and Th

• Total amounts of U and Th in the Earth are
  estimated from the condritic
• meteorites.
• Concentrations in the sediments and crusts are
  based on the samples
• on the surface, seismic data, and tectonic
  model.
• Concentrations in the mantle are estimated by
  subtracting the amounts in
• the sediments and the crusts.

13
Geological Uncertainty

• We assigned 10 for the observable geological
  uncertainty.
• This does not include uncertainties in the total
  amounts or
• distributions of U and Th.

U concentrations
U and Th concentration variations due to various
crustal types contribute 7 error in the total
flux.
Variations in local U and Th concentrations
contribute 3 error in the total flux.
14
Neutrino Oscillations

• The weak interaction neutrino eigenstates may be
  expressed as superpositions of definite mass
  eigenstates
• The electron neutrino survival probability can be
  estimated as a two flavor oscillations

15
KamLAND Neutrino Oscillation Measurement

• KamLAND saw an antineutrino disappearance and a
  spectral distortion.
• KamLAND result combined with solar experiments
  precisely measured the oscillation parameters.

16
The Expected Geoneutrino Flux

• Given an Earth model and neutrino oscillation
  parameters, the antineutrino flux per unit energy
  at KamLAND is given by

• The decay rate per unit mass

• The number of antineutrinos per decay chain per
  unit energy

• The mass concentration as a function of position
  in the Earth

• The density as a function of position in the Earth

• A survival probability due to neutrino
  oscillations,
• 
  for geoneutrino energy range.

17
Reference Earth Model Flux

• Expected geoneutrino flux at KamLAND
• 238U geoneutrinos 2.34?106 cm-2s-1
• 232Th geoneutrinos 1.98 ?106 cm-2s-1

18
Expected Geoneutrino Detection Rate

• By multiplying the expected geoneutrino flux and
  cross-sections, detection rates for geoneutrinos
  from U and Th at KamLAND are
• 238U geoneutrinos 3.0?10-31 per target proton
  year
• 232Th geoneutrinos 0.85?10-31 per target proton
  year

19
Geoneutrino Map of the Earth
Simulated origins of geoneutrinos detectable with
KamLAND using the reference Earth model
KamLAND
20
Geoneutrino References

• G. Marx, Menyhard N, Mitteilungen der Sternwarte,
  Budapest No. 48 (1960)
• M.A. Markov, Neutrino, Ed. "Nauka", Moscow, 1964
• G. Eders, Nucl. Phys., 78 (1966) 657
• G. Marx, Czech. J. of Physics B, 19 (1969) 1471
• G. Marx and I. Lux, Acta Phys. Acad. Hung., 28
  (1970) 63
• C. Avilez et al., Phys. Rev. D23 (1981) 1116
• L. Krauss et al., Nature 310 (1984) 191
• J.S. Kargel and J.S. Lewis, Icarus 105 (1993) 1
• R.S. Raghavan et al., Phys. Rev. Lett. 80 (1998)
  635
• C.G. Rothschild, M.C. Chen, F.P. Calaprice,
  Geophys. Rev. Lett. 25 (1998) 1083
• F. Montovani et al., Phys. Rev. D69 (2004) 013001

21
Have Geoneutrinos Been Measured before?
Fred Reines neutrino detector (circa 1953)
By Gamow in 1953
22
Were Fred Reines Background Events from
Geoneutrinos?
30TW
23
Outline

• Geoneutrinos
• KamLAND
• Background Events
• Results

24
KamLAND Detector
1km Overburden
Electronics Hut
Steel Sphere, 8.5m radius
Inner detector 1325 17 PMTs 554 20 PMTs 34
coverage
1 kton liquid-scintillator
Transparent balloon, 6.5m radius
Buffer oil
Water Cherenkov outer detector 225 20 PMTs
25
Inside the Detector
26
Determining Event Vertices

• Vertex determined using the photon arrival times
  at PMTs.
• Calibrated using sources deployed down the center
  of the detector.

27
Determining Event Energies

• The visible energy is calculated from the
  amount of photo-electrons correcting for spatial
  detector response.
• The real energy is calculated from the visible
  energy correcting for Cherenkov photons and
  scintillation light quenching.

28
Tracking Muons
Monte Carlo (line) and Data ()
29
Detecting Antineutrinos with KamLAND
Delayed
Prompt

• KamLAND (Kamioka Liquid scintillator AntiNeutrino
  Detector)

2.2 MeV g
0.5 MeV ?
e-
e
0.5 MeV ?
n
p

• Inverse beta decay
• ne p ? e n
• E? Te 1.8MeV

p
d
ne

• The positron loses its energy then annihilates
  with an electron.

• The neutron first thermalizes then captures a
  proton with a mean capture time of 200ms.

30
Selecting Geoneutrino Events
Delayed
Prompt
2.2 MeV g
0.5 MeV ?

• ?r lt 1m
• 0.5µs lt ?T lt 500µs
• 1.7MeV lt E?,plt 3.4MeV
• 1.8MeV lt Edlt 2.6MeV
• Veto after muons
• Rp, Rd lt 5m
• ?dgt1.2m

e
0.5 MeV ?
These cuts are different from the reactor
antineutrino event selection cuts because of the
excess background events for lower geoneutrino
energies.
31
Outline

• Geoneutrinos
• KamLAND
• Background Events
• Results

32
Reactor Background Introduction
KamLAND

• KamLAND was designed to measure reactor
  antineutrinos.
• Reactor antineutrinos are the most significant
  background.

33
Reactor Background Measurement

• Reactor antineutrino signals are identical to
  geoneutrinos except for the prompt energy
  spectrum.
• To calculate the reactor antineutrino interaction
  rate per target proton per year, we need to know
  the neutrino oscillation parameters, the
  detection cross-section (0.2) and each
  reactors
• Location
• Reactor thermal power (2.1)
• Fuel composition (1.0)
• Antineutrino spectrum (2.5)

34
Long-lived Reactor Background
Fractional Increase in energy spectra

• Fission fragments with half-lives greater than a
  few hours (97Zr, 132I, 93Y, 106Ru, 144Ce, 90Sr)
  may not have reached equilibrium.
• The reactor antineutrino spectrum is based on the
  measured ß spectrum after 1day exposure of 235U,
  239Pu, and 241Pu to a thermal n flux.
• Long-lived isotopes occur in the core and spent
  fuel.
• Spent fuel is assumed to be at the reactor
  location.

235U fission products
239Pu fission products
Antineutrino EnergyMeV
Kopeikin et al. Physics of Atomic Nuclei 64
(2001) 849
35
13C(a,n)16O Background

• Alpha source, 210Po?206Pba.
• Natural abundance of 13C is 1.1
• 13C(a,n)16O.
• n loses energy creating a prompt event, and is
  later captured creating a delayed event.

np scattering
13C(a,n)16O
n(12C,12C)n
36
Cosmic Muon Induced Background

• Muons produce unstable isotopes and neutrons as
  they go through the detector.
• 9Li and 8He ?-decay producing n, mimicking
  inverse ?-decay signals.
• Any events after muons are vetoed.
• 2ms after all muons
• 2s within 3m cylinder of the muon track
• 2s whole detector for muons with high light yield

37
Random Coincidence Background

• There is a probability that two uncorrelated
  events pass the coincidence cuts.
• The random coincidence background event rates are
  calculated by different delayed event time window
  (10ms to 20s instead).

38
Background Event Summary

• The following is a summary of the expected
  numbers of background coincidence events.

39
Pulse Shape Discrimination

• Antineutrino prompt event is caused by e whereas
  13C(a,n)16O prompt event is caused by n.
• These different prompt events produce different
  scintillation light time distributions allowing a
  statistical discrimination.

From AmBe source
Neutrons
Gammas
40
Pulse Shape Discrimination Part 2

• This study assumes similarities in time
  distributions of positrons and gammas.
• This method yields consistent 13C(a,n)16O
  background event rate.

From AmBe source
Neutrons
Gammas
41
Outline

• Geoneutrinos
• KamLAND
• Background Events
• Results

42
Data-set

• From March, 2002 to October, 2004.
• 749.10.5 day of total live-time.
• (3.46 0.17) 1031 target protons.
• (7.09 0.35) 1031 target proton years.
• 0.6870.007 of the total efficiency for
  geoneutrino detection.
• 14.8 0.7 238U geoneutrinos and 3.9 0.2 232Th
  geoneutrinos are expected.

43
Geoneutrino Candidate Energy Distribution
Expected total
Candidate Data
Expected total background
Expected reactor
(?,n)
Expected U
Random
Expected Th
44
Rate Analysis

• 152 candidate events
• 12713 expected background events.
• geoneutrinos.
• / (target proton-year)
  detected geoneutrino rate.
  

45
Likelihood Analysis

• Uses un-binned likelihood analysis.
• Uses the expected prompt event energy
  distribution.
• Uses the neutrino oscillation parameters
  determined from results of KamLAND reactor
  antineutrino and solar neutrino experiments.

46
Log Likelihood Equation
For given NU and NTh, log L is maximized by
varying the other parameters.
47
How Many Geoneutrinos Did We See?
Expected ratio from chondritic meteorites
Best fit 3 U geoneutrinos 18 Th geoneutrinos
Expected result from reference Earth model
48
How Many Geoneutrinos Did We See, Part 2?
??2 2(logLmax - logL)
Expected result from reference Earth model
Central Value 28
49
Reality Check

• Could all geoneutrinos come from an
  undiscovered uranium deposit?
• Not likely
• The antineutrino flux from a 100kton uranium
  deposit (the worlds largest) located 1km away
  from KamLAND would be only 3 of expected
  geoneutrino flux.

50
Conclusions

• This is the first experimental investigation of
  geoneutrinos.
• This is the first chemical analysis of the mantle
  of the Earth.
• We observed 4.5 to 54.2 geoneutrinos with 90
  C.L.
• Scaling concentrations in all regions of our
  reference Earth model, the 99 upper limit on
  geoneutrino rate corresponds to radiogenic power
  from U and Th decays of less than 60TW.
• The measurement is consistent with the current
  geological models.

51
Future of Geoneutrino Measurement with KamLAND

• The reactor background is irreducible for
  KamLAND.
• We are working on purifying the liquid
  scintillator, which will reduce the (?,n)
  background events.
• More accurate (?,n) cross section can lower the
  error on the (?,n) background rate.
• S. Harissopulos et al. submitted to Phys. Rev. C
  calculated new (?,n) cross sections with more
  accuracy.
• G. Fiorentini et al. arXivhep-ph/0508048
  recalculated the number of geoneutrinos using the
  above cross sections and our data. They claim
  that we detected geoneutrinos, 2.5?
  above 0.

52
Future Geoneutrino Experiment Considerations

• Location and geoneutrino data purity
• No nearby nuclear reactors
• On oceanic crust to probe mantle
• On continental crust to probe continental crust
• Needs to be shielded from cosmic muons
• Low radioactive background
• People are talking about
• Hawaii (oceanic crust with no reactors)
• Canada, South Dakota, Australia, the Netherlands,
  and South Africa (continental crust with no
  reactors)
• Geoneutrino Meeting in Hawaii, December 2005

53
Acknowledgement

• Prof. E. Ohtani (Tohoku University) and Prof. N.
  Sleep (Stanford University)
• Japanese Ministry of Education, Culture, Sports,
  Science, and Technology
• United States Department of Energy
• Electric associations in Japan Hokkaido, Tohoku,
  Hokuriku, Chubu, Kansai, Chugoku, Shikoku, and
  Kyushu Electric Companies, Japan Atomic Power Co.
  and Japan Nuclear cycle Development Institute
• Kamioka Mining and Smelting Company

54
KamLAND Collaborators
55
Geoneutrino Results in Nature
Nature 436, 499-503 (28 July 2005) doi
10.1038/nature03980
http//www.nature.com/nature/journal/v436/n7050/fu
ll/nature03980.html