Andr Rubbia ETH Zrich

1
Accelerator Experiments for CP Violation

• André Rubbia (ETH Zürich)
• Simulations performed by Paola Sala (ETH
  ZürichINFN)

Second NO-VE International Workshop on
"Neutrino Oscillations in Venice"
3th-5th December, 2003
2
How to experimentally observe the CP-phase ?

• From the unitary mixing matrix
• ones get the freedom of the complex phase d (but
  iff q13 ? 0 !).
• This phase can only be observed in appearance
  mode since disappearance is a T-symmetric process
• The effect for antineutrinos should be opposite
  to neutrinos (d?-d)
• It should have the expected L/E dependence
• But the phase is well hidden, e.g. consider
  oscillations involving ne and nm

CP-even
CP-odd
3
The discriminants

• ??? P(?e?????0) P(?e????0)
• Compares oscillation probabilities as a function
  of E? measured with wrong-sign muon event
  spectra, to MonteCarlo predictions of the
  spectrum in absence of CP violation. This works
  provided we know all other parameters precisely !
• ?CP(?)? P(?e????) P(?e????)
• Compares the appearance of ?? and ?m in two beams
  of both neutrinos and antineutrinos
• ?T(?)? P(?e??? ?) P(????e ?)
• Compares the appearance of ?? and ?e in ne and nm
  beams. This effect can be matter-enhanced for
  long baselines.
• ?T(?)? P(?e??? ?) P(????e ?)
• Same as previous case, but with antineutrinos.
  This effect is usually matter-suppressed with
  respect to the neutrino case for long baselines.

4
Including matter effects
L2900km
L732 2900 km and vacuum
L7400km
dp/2
L730km
L7400km
dp/2
See Nucl.Phys.B631239-284,2002
5
A phase-II experiment

• A CP experiment is a phase-II experiment
• Designed to have ample statistics (for a given
  q13) to precisely determine the oscillation
  probability as a function of energy
• Excellent energy resolution to observe energy
  dependence of oscillation probability and lift
  degeneracy
• A wide band neutrino beam to cover enough
  oscillations peaks or doing counting at
  different neutrino beam energy settings
• Neutrinos and antineutrinos runs to lift
  degeneracy (also in counting mode)

L730 km
dp/2
d0
dp/2
6
Three types of beams for CP ?

• One considers three types of neutrino beams
  produced at accelerators

Select focusing sign
Superbeams
b-beams
Select ion
Select ring sign
7
CP-phase effect at L130 km
??? N(??/2) N(?0)
Compares oscillation probabilities as a function
of E? measured with wrong-sign muon event
spectra, to MonteCarlo predictions of the
spectrum in absence of CP violation
b-beam
conventional
A cross-check !
8
In fact
Superbeams b-beam
?e ? ?m (?) ?m ? ?e (p) ?e ? ?m
(?-) ?m ? ?e (p-)
CP
CP
T
Neutrino factory
?e ? ?m (m) ?m ? ?e (m-) ?e ? ?m
(m-) ?m ? ?e (m)
CP
CP
T
So someone might argue that there is a symmetry
between the two roads superbeta and NF
9
However
Superbeams b-beam
Low energy En GeV e/p0 important µ/p
important Giant detectors (sn?En small)
Neutrino factory
High energy En gtgt GeV m charge with high
purity mandatory e charge to make it worth doing
it t identification for ne?nt Large detectors
(sn?En big)
So from the point of view of experiments,
superbeta and NF require completely different
detector optimizations!!
10
Hence
MINOS, OPERA, ICARUS, JHF, reactors
Superbeta-beams Giant detectors
And search for d?0?
Symmetry is broken
NF Large Magnetized detector
And d?0 detection might be hopeless if sin22q13
is ltlt 103
11
The catalog of detectors and their applications

• Neutrino factory _at_ En gtgt GeV
• Large Magnetized Fe Sampling Calorimeters M
  40kt
• Large Magnetized Liquid Argon detectors M 20kt
• Superbeams, beta-beams _at_ En GeV
• Giant Cerenkov detectors M 1000 t
• Giant Liquid Argon detectors M 100 kt
• Giant scintillator detectors M 30 kt
• Superbeams _at_ En GeV
• Large Low Z Sampling Calorimeters M 50 kt

L. Oberauer
G. Feldman
12
1. Detectors for Neutrino Factory
13
The goal at NF detect m, m, e, e, t, t and
NC !

• Lepton ID ? via CC interactions
• Muons straight-forward, look for penetrating
  particles, but beware p,K and charm decays
• Electrons harder, look for large short
  energy deposition, need good granularity for e/p0
  separation
• Taus hardest, kink or kinematical methods
  (statistical separation), t?hadronsn (Br60)
  look like NC
• Charge ID ? via magnetic analysis
• Muons easy, muon spectrometer downstream or
  fully magnetized target
• Electrons hardest, need to measure significantly
  precisely the bending in B-field before start of
  e.m. shower
• Taus easy for t?mnn (Br18), otherwise difficult

A high statistics study of all oscillation
channels would result in a precise determination
of all parameters, including the d-phase
14
Magnetized Fe sampling calorimeter - MINOS

• Successful construction of MINOS has bolstered
  the case that this is an easy (boring?)
  technology
• could clearly build alonger MINOS
• Golden channels at theNF requires identifying
  muon charge in DIS events

15
F. Sergiampietri, NUFACT 01 (Tsukuba) Based on
ICARUS
70 kton LAr
brute force
16
Signatures in magnetized liquid Argon
See Nucl.Phys.B589577-608,2000,
Nucl.Phys.B631239-284,2002

• A liquid argon TPC embebded in a magnetic field
  provides the possibility to measure both wrong
  sign muons and wrong sign electron samples

?e ? ?m (m) ?m ? ?e (m-) ?e ? ?m
(m-) ?m ? ?e (m)
CP
CP
T
µ
e
17
RD for liquid argon in magnetic field

• Opens new possibility
• Charge discrimination
• Momentum measurement of particles escaping
  detector (e.g. muons)
• MS dominated (Dp/p4 at L12m, B1T)
• Orientation of the field
• Bending in the direction of the drift where
  resolution is the best
• Achieved point resolution in T600 400 µm
• B-field perpendicular to E-field
• Lorentz angle small in liquids a30mrad _at_ E500
  V/cm, B0.5 T
• Required magnetic field strength for charge
  discrimination (xpath in LAr)

3 sigmas discrimination
18
Simulated nm CC events in B0.2 T
If B0,1 T ? xgt4m ? pgt0.8 GeV/c
µ
e
µ
µ
19
Discrimination of the electron charge
x1X0 ?Bgt0,5T
x3X0 ? Bgt0,3T
x2X0 ? Bgt0,4T
B1T
e
2.5 GeV
MC study charge confusionlt103 _at_ B1 T, Elt5 GeV

• Primary electron momentum curvature radius
  obtained by the calorimetric energy measurement
• Soft bremsstrahlung ??s the primary electron
  remembers its original direction ? long effective
  x for bending
• Hard initial bremsstrahlung ??s the energy is
  reduced ? low P ? small curvature radius

See hep-ph/0106088
20
Ongoing RD Test of liquid Argon imaging in
B-field

• Small chamber in SINDRUM-I recycled magnet up to
  B0.5T (230KW) given by PSI, Villigen
• Test program
• Check basic imaging in B-field
• Measure traversing and stopping muons bending
• Charge discrimination
• Check Lorentz angle (a30mrad _at_ E500 V/cm,
  B0.5T)

Width 300mm, height 150mm, drift length 150mm
21
2. Detectors for Super- and beta-beams
22
Giant water Cerenkov

• Perceived widely as a straightforward extension
  of SK (?)
• Many proposals, e.g., Hyper-K, UNO
• Many sites, e.g., Frejus, Kamioka, etc.
• Physics case is broad
• proton decay, neutrino properties, galactic
  supernovae,

e or p0 candidate
K. Nishikawa
23
Liquefied rare gases basic ideas

• Ideal materials for detection of ionizing tracks
• Dense (g/cm3 103 x rgas), homogeneous, target
  and detector
• Do not attach electrons (? long drift paths
  possible in liquid phase)
• High electron mobility (quasi-free drift
  electrons, not neon)
• Commercially easy to obtain (in particular,
  liquid Argon)
• Can be made very pure and many impurities freeze
  out at low temperature
• Inert, not flammable

24
Processes induced by charged particles in liquid
argon
When a charged particle traverses medium

• Ionization process
• Scintillation (luminescence)
• UV spectrum (l128 mn)
• Not energetic enough to further ionize, hence,
  argon is transparent
• Rayleigh-scattering
• Cerenkov light (if fast particle)

UV light
Charge
Cerenkov light (if bgt1/n)
M. Suzuki et al., NIM 192 (1982) 565
25
Comparison Water - liquid Argon
26
Comparison Water - liquid Argon
A new way to look at rare events
27
Extrapolation to underground kton liquid Argon
TPCs a different approach

• The ICARUS collaboration has proposed an
  underground modular T3000 detector for LNGS based
  on the cloning of the T600
• T3000 T600 T1200 T1200
• Design fully proven by t600 technical run
• Ready to be built by industry
• A 10 kton modular liquid argon detector could
  be ordered today (cost 200 M (conservative))
• Following a successful scaling up strategy, one
  could optimize costs and envision building bigger
  supermodules by increasing the dimensions of the
  current T1200 by a factor two in each directions
• However, to reach the wanted mass of 100 kton
  requires nonetheless a large number of
  supermodules (10x10kton 100 kton)
• a single volume appears to be the most attractive
  solution
• Is a strong RD program required to extrapolate
  the liquid argon TPC to the 100 kton scale (in a
  single step?)
• In the following, I will try to address the
  feasibility of a single volume 100 kton liquid
  argon detector

28
100 kton liquid Argon detector
Basic novelties

• Charge imaging scintillation Cerenkov light
  readout for complete event information
• Charge amplification to allow for extremely long
  drifts
• Single 100 kton boiling cryogenic tanker with
  Argon refrigeration

Electronic crates
f70 m
h 20 m
Perlite insulation
29
Access and highway tunnel
highway
Access
h 20 m
f70 m
30
Detector and highway tunnel
Highway tunnel
Detector
31
Open detector
Gas Argon
Liquid Argon
Drift
32
Cryogenic storage tanks for LNG
33
Liquefaction of LNG and transport via ships
Liquefaction plant in Oman
e.g. Nigeria LNG (1010 m3/year)
Filled with LCH4
Up to 145,000m3
34
(No Transcript)
35
Technodyne International Limited Unit 16
Shakespeare Business Center Hathaway Close, 
Eastleigh, Hampshire, SO50 4SR
ROM expected in Q1 2004
36
Summary parameters liquid Argon 100 kton
37
Detector schematic layout
Charge readout plane
GAr
E 3 kV/cm
LAr
Electronic racks
Extraction grid
E-field
E 1 kV/cm
UV visible light readout race track
Cathode (2MV)
(Not to scale)
38
Charge readout

• Detector is running in bi-phase mode
• In order to allow for long drift (20 m), we
  consider charge attenuation along drift and
  compensate this effect with charge amplification
  near anodes located in gas phase
• Amplification operates in proportional mode
• After max drift of 20 m _at_ 1 KV/cm, diffusion
  readout pitch 3 mm

39
Electron extraction in Ar-biphase (ICARUS RD)
Particle produces excitation (Ar) and ionisation
(Ar, e)
Scintillation SC is a result of 1.Direct
excitation 2.Recombination
Electroluminescence EL (proportional
scintillation) is a result of electron
acceleration in the gas
Electric Field
GAr
EL UV light
LAr
e- Ar
SC UV light
Both SC and EL can be detected by the same
photodetector
40
Amplification near wires à la MWPC

• Amplification in Ar 100 gas up to factor G100
  is possible
• GARFIELD calculations in pure Ar 100, T87 K,
  p1 atm
• Amplification near wires, signal dominated by
  ions
• Readout views induced signal on (1) wires and
  (2) strips provide two perpendicular views

Gain vs wire f _at_ 3.5kV
e-
Wire f30mm
102
PCB with strips
41
Extraction and amplification with GEMs
Buzulutskov et al, IEEE transaction on NS,
e-print physics/0308010 Buzulutskov et al,
NIMA513256-259 (2003)
GEM
Gas phase
GEM
GEM
Liquid phase
42
Large Electron Multiplier (LEM)
P. Jeanneret et al., NIMA 500 (2003) 133-143

• A large scale GEM (x10) made with ultra-low
  radioactivity materials (OFHC copper plated on
  virgin Teflon)
• In-house fabrication using automatic
  micromachining
• Modest increase in V yields gain similar to GEM
• Self-supporting, easy to mount in multi-layers
• Extremely resistant to discharges (lower
  Capacitance)
• Cu on PEEK under construction (zero out-gassing)

Chicago-Purdue P.S. Barbeau J.I. Collar J.
Miyamoto I.P.J. Shipsey
LEM bottom (anode) signal
LEM top (cathode) signal
43
LEM with Argon (ICARUS RD)
PRELIMINARY
Detection of charge signal and scintillation
light produced during amplification
400x400 mm2
Holes f 1 mm
44
UV light readout (ICARUS RD)

• Commercial PMT with large area
• Glass-window
• For scintillation VUV l 128 nm
• Wavelength-shifter
• Immersed T(LAr) 87 K

With TPB as WLS
Electron Tubes 9357FLA 8 PMT (bialkali with Pt
deposit) G 1 x 107 _at_ 1400 V peak Q.E. (400-420
nm) 18 (10 cold) Trise 5 ns, FWHM 8 ns
Lally et al., NIMB 117 (1996) 421
45
Cerenkov light readout (ICARUS)

• M. Antonello et al., ICARUS Collab., "Detection
  of Cerenkov light emission in liquid Argon NIMA,
  Article in Press
• Immersed PMT 2 EMI-9814 BQ (sensitivity up to
  160 nm)

Refractive index
Rayleigh scattering
Data consistent with Cerenkov emission
dN/dx(160-600nm) 700 g/cm (b1)
46
A dedicated cryogenic liquid plant for initial
filling phase

• Because of the large amount liquid argon needed
  to fill up the experiment (e.g. 300 ton/day to
  fill in 300 days), liquid argon must be produced
  locally
• One must envision a dedicated cryogenic plant
  located outside the tunnel and connected to the
  detector via km-long vacuum-insulated pipes
• Argon is extracted from the standard process of
  liquefaction from air
• Air mixture is cooled down and cold gas-mixtures
  are separated
• Oxygen, Nitrogen, Argon,
• The Liquid Argon is used to fill the experiment
  (The rest can be sold).

47
Cryogenic parameters initial filling phase
Note initial cooling of tanker not included
48
Cryogenic parameters boiling

• Notes
• Heat loss includes heat input from supports,
  instrumentation (cables), etc.)

49
Cryogenic parameters refilling (refrigeration)

• The dedicated cryogenic plant must hence produce
  liquid argon to refill what has evaporated

50
The dedicated cryogenic complex
Electricity
Air
Hot GAr
W
Underground complex
GAr
LAr
Q
External complex
Joule-Thompson expansion valve
Heat exchanger
Argon purification
LN2,
51
Liquid Argon purification

• Scaling of GAr/LAr purification system developed
  for ICARUS (Air Liquide)

52
What we get for 100 ktons

• Number of targets for nucleon stability
• 6 ? 1034 nucleons ? tp /Br gt 1034 years ?
  T(yr) ? e _at_ 90 C.L.
• Low energy superbeams or beta-beams
• 460 nm CC per 1021 2.2 GeV protons (real focus)
  _at_ L 130 km
• 15000 ne CC per 1019 18Ne decays g75
• Atmospheric
• 10000 atm events / year
• 100 nt CC /year from oscillations
• Solar
• 324000 solar neutrinos / year _at_ Ee gt 5 MeV
• Supernova type-II
• 20000 events _at_ D10 kpc

53
Proton decay
p?Kn
1035
p?ep0
65 cm
1034
p ? K ?e
p425 MeV
1 year exposure !
Nuclear effects in signal fully embedded in
FLUKA nuclear model
54
Atmospheric neutrinos
After 3 months running
Electrons
Muons

• Assumed oscillation parameters
• Dm232 3.5 x 10-3 eV2
• sin2 2Q23 0.9
• sin2 2Q13 0.1
• Electron sample can be used as a reference for no
  oscillation case

Measure L/E of muons electrons Duolith
55
Supernova neutrino detection

• Elastic scattering on electrons (ES)
• Charged-current (CC) interactions on Argon
• Neutral current (NC) interactions on Argon

QneCC 1.5 MeV
-
QneCC 7.48 MeV
QNC 1.46 MeV
Possibility to separate the various channels by a
classification of the associated photons from the
K, Cl or Ar deexcitation (specific spectral lines
for CC and NC) or by the absence of photons (ES)
56
Superbeams beta-beams assumed parameters

• Unless otherwise noted we assume in the following

57
Conventional superbeam E L optimization

• In order to estimate sensitivity to CP-violation
  phase, we define three quantities based on the
  integrated number of events and

??? N(??/2) N(?0)
backgroundintrinsic ne
oscillated
58
Conventional superbeam
Simulations with full focusing optics (see New
J.Phys.488,2002)
p focusing
p- focusing
L730 km
L130 km
intrinsic ne
intrinsic nm
oscillated
Per 1019 pots 100 kton
59
Rejection p0 based on imaging
Single photon rejection (MC)

• Based on full simulation, digitization, noise and
  automatic reconstruction of events
• Algorithm cut for 90 eff. electrons
• Events with vertex conversion within 1cm (3
  wires) of vertex R119
• Single/double mip R230 (preliminary)

Preliminary
cut
1 p0 (MC)
ltdE/dxgt MeV/cm
Imaging provides 2?10-3 efficiency for single p0
60
Rejection p0 based on imaging

• p0 surviving dE/dx separation cut (total 31
  events out of 1000 1GeV p0)
• 21 events Compton scattering
• 5 events Asymmetric decays (partners have less
  than 4 MeV)
• 2 events positron annihilation immediately
• 1 event positron make immediate Bremsstrahlung
  taking gt90 of energy
• p0 rejection improves with energy 5 _at_ 0.25 GeV,
  4 _at_ 0.5 GeV, 3 _at_ 1 GeV, 2 _at_ 2 GeV

Compton electron
Full simulationdigitizationnoise

• Further rejection by kinematical cuts (depends on
  actual beam energy profile)
• E.g. nn ? np0n precise mass reconstruction

? Reduce to negligible level
61
Results conventional beam
3 systematics
L730 km
L130 km
The rules Merit pot ? Eproton and optimal L
hold
62
Beta-beams E L optimization

• In order to estimate sensitivity to CP-violation
  phase, we define three quantities based on the
  integrated number of events and

??? N(??/2) N(?0)
oscillated
Backgroundpions from NC
1
63
Energy integrated rates b-beam
L130 km
L400 km
W/o Cerenkov
With Cerenkov
64
Beta beam charged pion background rejection
Signal
µ
Use combination of charge imaging (?(dE/dx)dx
Tkin) Cerenkov light readout (b)
Background
p
? Reduce to negligible level
W/o Cerenkov optimize neutrino energy to
suppress pion production at the cost of
oscillated event rate (proportional to g)
65
Beta beam charged pion background rejection

• Momentum cut
• Range
• Many pions interact
• Particle stops
• Cerenkov based rejection
• Kinetic energy is measured from deposited charge
• Velocity is measured from Cerenkov photon
  counting
• The two can be combined to discriminate pions
  from muons

66
Baselineenergy optimization b-beam
Ion decays needed to achieve 3s of Dd
L130 km
L400 km
L130 km
With 1 syst.
W/o syst.
67
Sensitivity to CP-violation example I
18Ne g75 L130 km 10 years _at_ 2x1018 ions/yr
1 systematic
68
Sensitivity to CP-violation example II
18Ne g250 L400 km 10 years _at_ 2x1018 ions/yr
1 systematic
69
Beta-beam spectra EL comparison
q13 3
18Ne g250 L400 km
18Ne g75 L130 km
70
Conclusion

• Given the tremendous physics potential of such
  detectors, we invite the community to a deep
  reflection concerning the feasibility of giant
  neutrino detectors and fully compare these two
  options
• Giant 1 megaton H2O
• Giant next-generation 100 kton liquid Argon
  detector, taking advantages of possible advances
  in the LAr TPC technology
• Bi-phase operation with charge amplification for
  long drift distances
• ImagingScintillationCerenkov readout for
  improved physics performance
• Giant boiling cryostat (LNG technology)
• They offer the widest physics output (accelerator
  non-accelerator)
• Coupled to the proper superbeams and beta-beams
  they could greatly improve our understanding of
  the CP-phase in the lepton sector
• International sites with proper depths and
  infrastructure for potentially locating such
  giant detectors should be reviewed and compared
• To build such large/giant detectors for only CP
  seems unconceivable, hence, giant detectors must
  have broad physics programs
• Detectors should be underground (depth to be
  optimized vs backgrounds)