The contribution of  to the understanding of oscillations Leslie Camilleri CERN, PH University of Bologna November 10, 2005

Plan of the talk A very brief theory of neutrino oscillations. The Past: The discovery of oscillations in Solar and Atmospheric neutrino experiments. The Far Future programmes: A Super-conducting Proton Linac (SPL). A  beam. A Neutrino Factory The Present programmes: Double-  decay Reactors Accelerator long baseline experiments. NO A

Theory 2

Theory 3

The PAST Discovery of Oscillations

Solar spectrum

Real Time  Charged Current (CC) reactions on nucleons  e + n  p + e - At quark level e + d  u + e- Sensitive ONLY to  e  the flux of e  Neutral Current (NC) reactions on nucleons x + n  n + x Sensitive to flux from ALL flavours,  e, ,   CC and NC on nucleons: negligible in WATER due to Oxygen being very tightly bound (>15 MeV) Important in HEAVY WATER: deuterium binding energy only 2 MeV.  Elastic Scattering (ES) on electron: ES is large in WATER and HEAVY WATER Sensitive to  (  ) only Sensitive to flux from ALL flavours,  ( e, ,  ) But rate smaller than W exch.

Super-Kamiokande The Detector tons ultra-pure water 1 km overburden = 2700 m.w.e tons fiducial volume Sensitive to Elastic Scattering ONLY Mostly e ’s

SK-I: 8 B Solar Neutrino Flux 8 B flux = 2.35  0.02  0.08 [x10 6 /cm 2 /s] Data / SSM BP2004 =  0.004(stat.) (syst.)  230 solar  events PLB539 (2002) 179 Electron total energy: MeV May 31, 1996 – July 15, 2001 (1496 days ) Data / SSM BP2000 =  0.005(stat.) (syst.)

Suppression relative to Standard Solar Model Suppression relative to Standard Solar Model is observed in all experiments. Is it due to a misunderstanding as to how the sun “works” ? Standard solar model. Or are the neutrinos “disappearing” ?

SNO (Heavy water):Sensitive to CC, NC, ES. Calculate flux from each. Flavor content of solar flux. e only e mostly e     Neutrinos DO NOT disappear. They just Change Flavour ! Driven by enhanced oscillations in the dense matter of the sun. Using ALL neutrinos Fully Consistent with Standard Solar Model NC ES SK ES CC SSM   ee

Matter effects: Mikheyev-Smirnov-Wolfenstein e     e e e    N ee ee All  flavours Only e  flavour Introduces extra potential for e ZoZo W-W-

Confirmed by KAMLAND: Reactor antineutrinos to detector at Kamioka Solar Experiments KAMLAND KamLAND + Solar Completely consistent

Atmospheric Neutrinos: e and   Zenith angle  Baseline Produced by  and K decays in upper atmosphere

 /e identification  sharp ring e fuzzy ring due to many particles in shower

Suppression of  zenith angle and energy dependent Explained by oscillations No oscillations Oscillations

Confirmed by K2K: KEK accelerator to Super-Kamiokande Completely consistent

Near Future (Accelerators) OPERA (CNGS beam) ~732km MINOS (NUMI beam) 732km Confirm   ~  oscillations by searching for  appearance in a  beam using    X  ...  Emulsions to observe the kink in  decays They look for  disappearance to observe oscillatory pattern in energy spectrum. Measure  m 2 and  23

3-family oscillation matrix S = sine c = cosine   CP violation phase.    drives SOLAR oscillations: sin 2  12 = (+- 16%)   23 drives ATMOSPHERIC oscillations: sin 2  23 = (+44% -22%)   13 the MISSING link ! sin 2  13 < 0.03  Set by a reactor experiment: CHOOZ.

CHOOZ: A reactor experiment to measure  13  Reactors emit 6 e ’s per fission 5.6 x e /second for a 3 GW reactor (Few MeV)  Excellent source of antineutrinos.  With a detector at 1 km, L/E = 1km/1MeV ~ same as atmospheric ~ 1000km/1GeV.  Can probe same  m 2  If they oscillate to   or   they would NOT have enough energy to create  ’s or  ’s via CC interactions.  Cannot study oscillations through an “appearance” experiment.  Must study oscllations via e disappearance. P ee = 1 – sin 2 2  13 sin 2 [(  m 23 2 L)/(4E )] Looked for distortions of the expected energy spectrum or in the rate Did not find any. Set a limit on sin 2 2  13 > 0.12 for  m 2 atm or sin 2  13 > 0.03

Mass hierarchy Sign of  m 2 23 m2m1m2m1 m3m3 m2m1m2m1 m3m3  m 23 2 = 2.4 x eV 2  m 23 2 = 2.4 x eV 2  m 12 2 = 7.9 x eV 2 > 0.05 eV 2 Normal Hierarchy Inverted Hierarchy Oscillations only tell us about DIFFERENCES in masses Not the ABSOLUTE mass scale: Direct measurements or Double  decay Upper limit: Tritium  decay: mass ( e ) < 2.2 eV Lower limit: (2.4 x ) 1/2  > 0.05 eV  m 12 2 = 7.9 x eV 2 e  

Why are neutrino masses so low???? Other particles

What’s needed next?  Determine  13.  Determine the mass hierarchy.  Any CP violation in he neutrino sector?

Correlations in Oscillation Probability From M. Lindner: Measuring P (  ~ e ) does NOT yield a UNIQUE value of  13. Because of correlations between  13,  CP and the mass hierarchy (sign of  m 2 31 ) CP violation: Difference between Neutrino and Antineutrino Oscillations

8-fold degeneracies   13 -  ambiguity.  Mass hierarchy two-fold degeneragy A measure of P  e can yield a whole range of values of  13 Measuring with ’s as well reduces the correlations   23 degeneracy: For a value of sin 2 2  23, say 0.92, 2  23 is 67 o or 113 o and  23 is 33.5 o or 56.5  In addition if we just have a lower limit on sin 2 2  23, then all the values between these two are possible.

Correlations In vacuum and without CP violation: P(   e ) vac = sin 2  23 sin 2 2   sin 2  atm with  atm = 1.27  m 2 32 (L/E) For  m 2 32 = 2.5 x eV 2 and for maximum oscillation  We need:  atm =  /2  L(km)/E(GeV) = 495 For L = 800km E must be 1.64 GeV, and for L = 295km E = 0.6 GeV Introducing matter effects, at the first oscillation maximum: P(   e ) mat = [1 +- (2E/E R )] P(   e ) vac with E R = [12 GeV][  m 2 32 /(2.5x10 -3 )][2.8 gm.cm -3 /  ]~ 12 GeV +- depends on the mass hierarchy. Matter effects grow with energy and therefore with distance. 3 times larger (27%) at NO A (1.64 GeV) than at T2K (0.6 GeV)

The FAR Future

Superconducting Proton Linac  Power : 4 MW  Kinetic Energy : 2.2 GeV (3.5 GeV)  Repetition Rate: 50 Hz  Spill Length: 11 msec.  Accumulator needed to shorten pulse length.  Target: Liquid Mercury Jet to cope with stress due to high flux.  Focusing: Horn and Reflector optimized for 600 MeV/c particles  Decay Tunnel: 20m long 1m radius  Distance: 130km  Neutrino energy to be at oscillation maximum for  m 23 2 = 2.5 x eV MeV  Detector mass: 440 kton fiducial.  Type: Water Cerenkov (Super-K)  Location: New lab in Frejus tunnel

Advantage of mixing neutrino and antineutrino running 3.5 and 4.5 GeV proton beam 260 and 350 MeV options 5 years of running. 2 years of running and 8 years of running The limit IMPROVES near  = 90 o Limit: ~0.001

Beta beams Idea introduced by Piero Zucchelli. Accelerate radioactive ions decaying via  + or  -. Because of Lorentz boost, the decay electron neutrinos or antineutrinos will be focused forward into a beam. Look for: Appearance of  or  Advantages: “Clean” beams with no intrinsic  component. Precisely calculable energy spectra. Energy of beam tunable through acceleration of ions.  Accelerate protons in SPL  Impinge on appropriate source  Bunch resulting ions  Accelerate ions in PS and SPS.  Store in decay ring. 8 bunches.  Favourite scheme:  6 He 6 Li + e - + e 18 Ne 18 Ne + e + + e Half lives: 0.8 sec and 0.64 sec. Detector: Same as for SPL (Frejus)

 sensitivity for  = 60,100 Statistics limited CP violation Asymmetry decreases with increasing    2% Syst. Unc. 2.9 x He ions and 1.2 x Ne ions per year decaying in straight sections M. Mezzetto SPSC Villars 3  Down to ~  = 35 o Sin 2 2   

Neutrino Factory   p  s  s =  yr  yr e  yr   yr    e + e   oscillates e     interacts giving    WRONG SIGN MUON interacts giving   Need to measure charge  Magnetic detector    e + e  

Neutrino Factory

The NEAR Future

Neutrinoless Double-  decay e-e- e-e- W-W- W-W- e-e- e-e- i W-W- W-W- N N´N´ N´N´ N Standard 2-neutrino double  decay Neutrinoless double  decay Can only happen if the neutrino is reabsorbed as an Antineutrino Helicity must flip  non-zero mass If the neutrino is its own Antiparticle: Majorana i i (A,Z)  (A,Z+2) + 2 e -

Detection arbitrary units (Q  ~ MeV) Look for a peak at the end point of the2-neutrino spectrum One claim: not generally believed New experiments will use: 130 Te, 132 Xe, 76 Ge, 100 Mo Will observe the 2 electrons through bolometric, calorimetric or tracking techniques

Limits Claim Rate = (T o ½ ) -1 = (Phase space factor) x (Matrix element) 2 x 2 = | U e1 2 m 1 + U e2 2 m 2 +U e3 2 m 3 | New experiments will go down milli eV Small if m 3 is heaviest state, because multiplied by U e3 2 (= sin 2  13 ) which is small (<0.03). Better with inverted hierarchy

 13 with Reactors P ee = 1 – sin 2 2  13 sin 2 [(  m 23 2 L)/(4E )] near oscillation maximum Advantage: NO dependence on  CP or mass hierarchy: No ambiguities. Disadvantage: Cannot determine them! Measured through inverse  decay: e + p = e+ + n Distortion of the e energy spectrum Oscillation effects are SMALL Must know e energy spectrum well to control SYSTEMATICS. CHOOZ: One detector at 1100m Systematic uncertainty: 2.7%

Technique Solution: Use 2 detectors Additional NEAR detector: measure flux and cross sections BEFORE oscillations. Even better: interchange NEAR and FAR detectors part of the time to reduce detector systematics Detectors : Liquid scintillator loaded with gadolinium: Neutron capture  photons p p e+e+  e-e- e+e+   511 keV n n p  2.2 MeV ~200 s e e + annihilates with e - of liquid: MeV n captured by Gadolinium: 8 MeV of photons emitted within 10’s of  sec. Delayed Coincidence of 2 signals

Proposed experiments ExperimentLocation SitesSystematicsLimit Double CHOOZFranceNear/Far0.6%0.02 BraidwoodUSANear/Far0.3%0.005 Daya BayChinaNear/Mid/Far % Example: Double CHOOZ 1% 0.4% Importance of systematics CHOOZ systematics Was 2.7%

Future (Accelerators) T2K (Japan) 295km NO A (NUMI beam) 810km Both projects are Long Baseline Off-axis projects. They search for  ~ e oscillations by searching for e appearance in a  beam. Determine that    is non-zero  Measure it? Mass hierarchy?

OFF-AXIS Technique Most decay pions give similar neutrino energies at the detector: The Neutrino Energy Spectrum is narrow: know where to expect e appearance Can choose the off-axis angle and select the mean energy of the beam. ( Optimizes the oscillation probability)  Target Horns Decay Pipe Super-K.    2 o  3 o

T2K  0.7 GeV  e from K (hashed) and  decays 0.4 %  background at peak e  New 40 GeV Proton Synchrotron (JPARC)  Reconstructed Super-K  Near detector to measure unoscillated flux distance of 280 m (Maybe 2km also)  JPARC ready in 2008  T2K construction  Data-taking starting in 2009

 disappearance:  m 23 2 and  23. Position of dip  m 23 2 to an accuracy of ~ eV 2 Depth of dip Sin 2 2  23 to an accuracy of 0.01 Factor of 10 improvement in both

Measurement of  13. e appearance

Sensitivity, correlations, degeneracies But, the limit on sin 2 2   is much worse if we take into account correlations and degeneracies Sin 2 2  13 ~ 0.04  CP 150

T2K II: Hyper-Kamiokande One megaton Water Cerenkov and 4MW accelerator.   +150 o -150 o sin 2 2  13 Improvement by more than an order of magnitude on  13 sensitivity All degeneracies included

T2K II: Sensitivity to  CP Definition: For each value of sin 2 2  13 : The minimum  for which there is a difference Of 3  between CP and NO CP violation Limited by statistics CP violation asymmetry (  bar) decreases with increasing sin 2 2  13 Sin 2 2  o 50 o 

NO A Detector Given relatively high energy of NUMI beam, decided to optimize NO A  for resolution of the mass hierarchy Detector placed 14 mrad (12 km) Off-axis of the Fermilab NUMI beam (MINOS). At Ash River near Canadian border (L = 810km) : New site. Above ground. Fully active detector consisting of 15.7m long plastic cells filled with liquid scintillator: Total mass 30 ktons. Each cell viewed by a looped WLS fibre read by an avalanche photodiode (APD) cells TiO 2 Coated PVC tubes

NO A The quantum efficiency of APD’s is much higher than a pm’s: ~80%. Especially at the higher wave lengths surviving after traversing the fibre. Measured Photoelectrons Per muon track 30 pe Asic for APD’s: 2.5 pe noise  S/N ~ 12 After15.7m still 30 photoelectrons/mip. with looped fibre. Coating:15% TiO 2

Avalanche Photodiode  Hamamatsu 32 APD arrays  Pixel size 1.8mm x 1.05mm (Fibre 0.8mm diameter)  Operating voltage 400 Volts  Gain 100  Operating temperature: -15 o C (reduces noise) Photon Asic for APD’s: 2.5 pe noise  S/N ~ 30/2.5 = 12

APD response Measured with light equivalent to one and two mip’s Noise Signal well separated from noise pe

The Beam PROTONS: 6.5 x protons on target per year. Greatly helped by  Cancellation of BTeV  Termination of Collider programme by A gain of a factor of > 2 in numbers of protons delivered. Longer term: Construction of an 8 GeV proton driver: x x protons on target per year is the goal.

The Beam: Same NUMI beam as MINOS 14 mrad Can select low, medim and high energy beams by moving horn and target Best is the Medium energy beam

Beam spectra Signal Sin 2 2  13 = 0.04 Beam e background ~ 0.5%   

  e separation Electrons (shower) Electrons (shower) Muons Low energyHigh energy  o in NC also a problem. Signal e efficiency: 24%.  CC background 4 x  NC background 2 x 10 -3

Location

3  discovery limits for  13 = 0 Discovery limit is better than 0.02 for ALL  ’s  and BOTH mass hierarchies. 2.5 years each and . 5 years

3  discovery limits for  13 = 0 Comparison with Proton Driver 2.5 years each and .

3  discovery limits for  13 = 0 Comparison with T2K and 2 Reactor experiments Braidwood Double Chooz T2K

Resolution of mass hierarchy  Fraction of  over which the mass hierarchy can be resolved at    qual amounts of neutrino and antineutrino running: 3 years each assuming Phase I.  Near the CHOOZ limit the mass hierarchy can be resolved over 50% of the range of .  T2K Phase I can only resolve the hierarchy in a region already excluded by CHOOZ. Because of its lower energy.  Some small improvement if we combine T2K and NO A results CHOOZ limit T2K

Looking further ahead  With a proton driver, Phase II, the mass hierarchy can be resolved over 75% of  near the CHOOZ limit.  In addition to more protons in Phase II, to resolve hierarchy a second detector at the second oscillation maximum can be considered:   atm = 1.27  m 2 32 (L/E) =   L/E = 1485, a factor of 3 larger than at 1 st max.  For ~ the same distance, E is 3 times smaller:  matter effects are smaller by a factor of 3  50 kton detector at 710 km.  30km off axis (second max.)  6 years (3  + 3 ) Determines mass hierarchy for all values of  down to sin 2 2  13 = 0.02

CP reach  To look for CP violation requires the proton driver.  But combining with a second detector is what really becomes SIGNIFICANT. Proton driver Proton driver + 2 nd detector

Near Detector to understand the beam 262 T 145 T totally active 20.4 T fiducial (central 2.5 x 3.25 m) 8-plane block 10.6 T full 1.6 T empty Muon catcher 1 m iron Target region Veto region 9.6 m 5 m 3.5 m Shower containment region

Near detector locations M Test MINOS Surface Building NuMI Access Tunnel

Cost and schedule  Total cost (Far and near detectors, building, admin etc…) 164 M$ (including 50% contingency) Status  Approved by Fermilab Program Advisory Committee  Going through reviews Schedule  Assumption: Approval in  Building ready: May  First kiloton: October  Completion: July  Possible CERN participation.  European groups already in NO A: Athens, College de France, Tech. Univ. Munich, Oxford, Rutherford

Status of NUMI/MINOS: Near detector They get ~ 2.5 x protons/spill Spill: either 2.2 or 3.8 secs. (Depends on antiproton cooling) Delivered so far: 0.8 x With 2.5 x and 2 sec spill  2.5 x /year With a factor of ~ 2 from stopping collider ~ 5 x Not Far off NOvA target…! Event time in Spill Structure: Booster batches Energy of CC events For running in 3 configurations: Only target moved Direction of track In CC events Relative to nominal beam

Status of NUMI/MINOS: Far detector Blind analysis. Plots are for 1 week running. Should have about 500 CC events if no oscillations Event time in Spill Structure: Booster batches Cosmic rays: from above 180 o Beam events: 90 o

Conclusions  The neutrino oscillation programme is very rich.  The smallness of neutrino masses is fascinating.  The mass hierarchy must be determined.  Is there any CP violation in the neutrino sector?  The road to these is the observation of a non-zero     The NUMI beam is functioning well.  NO A has a well-developed long term research programme.

Near Detector in MINOS Surface Building 45,000   CC events2,200 e CC events 6.5 x pot in 75 mrad off-axis beam Kaon peak

Neutrino spectra at near and far detectors  CC events e CC events Far Detector x 800 Site 1.5 Site 2

3  Determination of CP Violation