arXiv:0804.2111v1 [hep-ph] 14 Apr 2008
April 14, 2008
A Possible Future Long Baseline Neutrino and Nucleon
Decay Experiment with a 100 kton Liquid Argon TPC
at Okinoshima using the J-PARC Neutrino Facility
A.Badertscher1, T.Hasegawa2, T.Kobayashi2, A.Marchionni1,
A.Meregaglia1∗, T.Maruyama2,3, K.Nishikawa2, and A.Rubbia1
(1) ETH Z¨ urich, (2) KEK IPNS, (3) University of Tsukuba
In this paper, we consider the physics performance of a single far detector composed
of a 100 kton next generation Liquid Argon Time Projection Chamber (LAr TPC)
possibly located at shallow depth, coupled to the J-PARC neutrino beam facility with
a realistic 1.66 MW operation of the Main Ring. The new far detector could be located
in the region of Okinoshima islands (baseline L ∼ 658 km). Our emphasis is based
on the measurement of the θ13and δCP parameters, possibly following indications for
a non-vanishing θ13in T2K, and relies on the opportunity offered by the LAr TPC
to reconstruct the incoming neutrino energy with high precision compared to other
large detector technologies. We mention other possible baselines like for example J-
PARC-Kamioka (baseline L ∼ 295 km), or J-PARC-Eastern Korean coast (baseline
L ∼ 1025 km). Such a detector would also further explore the existence of proton
∗Now at IPHC, Universit´ e Louis Pasteur, CNRS/IN2P3, Strasbourg, France.
The“Quest for the Origin of Matter Dominated Universe” is a long standing puzzle of our
physical world (see e.g. ). Answers to this question might come from further exploration
• the Lepton Sector CP Violation by precise testing of the neutrino oscillation processes
– measure precisely the δCP and the mixing angle θ13;
– examine matter effect in neutrino oscillation process and possibly conclude the
mass hierarchy of neutrinos.
• Proton Decay:
– Search for p → ν K+and p → e π0in the range 1034÷35years
with assuming non-equilibrium environment in the evolution of universe.
The primary motivation of the T2K experiment∗ is to improve the sensitivity to the
νµ→ νeconversion phenomenon in the atmospheric regime by about an order of magnitude
in the mixing angle sin22θ13compared to the CHOOZ experimental limit . The discovery
of a non-vanishing θ13angle would ascertain the 3 × 3 nature of the lepton flavor mixing
matrix. The measurements in T2K might indicate that the θ13angle is in a region where the
simultaneous determination of neutrino mass hierarchy and the CP violating phase becomes
The J-PARC accelerator complex which includes the 180 MeV LINAC, the 3 GeV Rapid
Cycling Synchrotron (RCS) and the 30-50 GeV Main Ring Synchrotron (MR) is planned to be
commissioned in 2008. The J-PARC neutrino beam facility, under construction for the T2K
experiment, is foreseen to begin operation in 2009. The final goal for the T2K experiment is
to accumulate an integrated proton power on target of 0.75 MW×5 × 107seconds. Within
a few years of run, critical information, which will guide the future direction of the neutrino
physics, will be obtained based on the data corresponding to about 1÷2 MW × 107seconds
integrated proton power on target (roughly corresponding to a 3σ discovery at sin22θ13>
0.05 and 0.03, respectively) .
It is well documented in the literature (see e.g. Refs. [5, 6, 7]) that the most challenging
task for next generation long baseline experiments is to unfold the unknown oscillation
parameters sin22θ13, δCPand mass hierarchy, sgn(∆m2
energy dependence of the oscillation signal to resolve the so-called problems of “correlations”
The most important experimental aspects here are the beam profile (e.g. the ability to
cover with sufficient statistics the 1stmaximum of the oscillation, the 1stminimum, and the
2ndmaximum), the visible energy resolution of the detector, with which the neutrino energy
31), and from the measurement of the
∗We call T2K the approved Tokai-To-Kamioka experiment using Super-Kamiokande as far detector. New
investigations or performance upgrades beyond this approved phase will presumably have a different name.
can be reconstructed, and the spectrum of the misidentified background (e.g. π0spectrum,
One possible approach, advocated in the literature [8, 9], is to locate a very massive
coarse Water Cerenkov detector at very long baselines (L ≥ 2500 km) from a new multi-
GeV neutrino source and to detect more than one oscillation maximum. In this kind of
configuration, although the oscillation peaks are well separated in energy, an important
issue is the ability to subtract the backgrounds, in particular that coming from misidentified
neutral current events with leading π0’s which introduce new sources of systematic errors. In
order to alleviate this problem, a different approach assuming two similar far Water Cerenkov
detectors, one located at Kamioka (L ∼ 295 km) and the other in Korea (L ∼ 1000 km)
at the same OA2.5◦from the J-PARC neutrino beam, was discussed in Ref. . The two
detectors would see the same sub-GeV neutrino beam (in fact the same as in T2K) but
the different baselines would allow to study E/L regions corresponding to the 1st and 2nd
The approach of our paper is different. We consider the physics performance of a single
100 kton next generation liquid Argon Time Projection Chamber (LAr TPC) located in the
region of Okinoshima islands corresponding to a (relatively modest) baseline L ∼ 658 km.
A detector in this location will automatically see the J-PARC neutrino beam at an off-axis
angle of ∼ 0.8◦. As we show below, this configuration allows to study both 1st and 2nd
oscillation maximum peaks with good statistics. We rely on the realistic opportunity offered
by the increase of intensity of the J-PARC MR and on the higher precision of the LAr TPC
than other large detectors to separate the two peaks. In addition, the π0background is
expected to be highly suppressed thanks to the fine granularity of the readout, hence the
main irreducible background will be the intrinsic νecomponent of the beam.
At this stage of our investigations, the emphasis is placed on the measurement of the
θ13and δCP parameters, possibly following indications for a non-vanishing θ13in T2K. We
also compare the possibility to determine the sin22θ13-δCPparameters with a single detector
configuration at other possible baselines like for example the T2K (Tokai-Kamioka, OA2.5◦,
L ∼ 295 km) or the Tokai-Eastern Korean coast (L ∼ 1025 km) with an off-axis OA1.0◦[11,
12]. More complete physics performance studies will be presented elsewhere.
2 A possible next step beyond T2K
If a significant νesignal were to be observed at T2K, an immediate step forward to a next
generation experiment aimed at the discovery of CP-violation in the lepton sector would be
recommended with high priority. We have conducted a case study for the discovery of lepton
sector CP violation based on the J-PARC MR power improvement scenario. Naturally,
next generation far neutrino detectors for lepton sector CP violation discovery will be very
massive and large. As a consequence, the same detector will give us the rare and important
opportunity to discover proton decay. Thus, we also discuss the proton decay discovery
potential with a huge underground detector.
Compared with the T2K experimental conditions, lepton sector CP violation discovery
• an improved neutrino beam condition (intensity increase, broader energy spectrum,
possibly re-optimization of the focusing optics, ...);
• an improved far neutrino detector (by improvements we primarily mean increased signal
reconstruction efficiency, better background separation, better energy resolution, ...,
and not only its volume).
Detector improvements include
• detector technology;
• its volume;
• its baseline and off-axis angle with respect to the neutrino source.
As for the neutrino beam intensity improvement, a realistic first step power improvement
scenario at J-PARC Accelerator Complex has been recently analyzed and proposed (LINAC
energy is recovered to be 400 MeV, h=1 operation at RCS and 1.92 seconds repetition cycle
operation at MR). As for the detector technology, we assume a 100 kton Liquid Argon Time
Projection Chamber for our case study.
The effects of the CP phase δCP appear either
• as a difference between ν and ¯ ν behaviors (this method is sensitive to the CP-odd term
which vanishes for δCP= 0 or 180◦);
• in the energy spectrum shape of the appearance oscillated νe charged current (CC)
events (sensitive to all the non-vanishing δCP values including 180◦).
Since antineutrino beam conditions are known to be more difficult than those for neutrinos
(lower beam flux due to leading charge effect in proton collisions on target, small antineutri-
nos cross-section at low energy, etc.), we concentrate on the possibility to precisely measure
the νeCC appearance energy spectrum shape with high resolution during a neutrino-only
3The liquid Argon Time Projection Chamber
The liquid Argon Time Projection Chamber (LAr TPC) (See Ref.  and references therein)
is a powerful detector for uniform and high accuracy imaging of massive active volumes. It
is based on the fact that in highly pure Argon, ionization tracks can be drifted over distances
of the order of meters. Imaging is provided by position-segmented electrodes at the end of
the drift path, continuously recording the signals induced. T0is provided by the prompt
In this paper, we assume a LAr TPC detector with a mass of the order of 100 kton,
for example of a kind based on the GLACIER concept . Unlike other liquid Argon
TPCs operated or planned which rely on immersed wire chambers to readout the ionization
signals, a double phase operation with charge extraction and amplification in the vapor phase
is considered here in order to allow for very long drift paths and for improved signal-to-noise
ratio. We assume that successful application of such novel methods will be an important
milestone in the R&D for very large LAr TPC detectors in the range of 100 kton. At this
stage, a ton-scale prototype based on this scheme has been developed and is under test .
Challenges to realize liquid Argon TPC’s with a scale relevant to this paper have been
reviewed e.g. in Ref. .
We refer to previous physics performance studies assuming such a detector configura-
tion [12, 17] and recall that since Liquid Argon TPC has advantages on
• good energy resolution/reconstruction,
• good background suppression,
• good signal efficiency
it is suitable for a precision measurement of the neutrino energy spectrum to extract CP
information. Thus we concentrate on the νeappearance energy spectrum shape measurement
to extract leptonic CP phase information.
4The choice of far location: Okinoshima
The J-PARC neutrino beam line was designed and constructed in order to allow an off-axis
angle with respect to Super-Kamiokande (SK) between 2.5◦and 3◦. A beam setup yielding
an OA2.5◦at SK was chosen for the T2K experiment. In this configuration, the center of the
T2K neutrino beam will go through underground beneath SK, and will automatically reach
the Okinoshima island region with an off-axis angle ∼ 0.8◦and eventually the sea level east
of the Korean shore with an off-axis angle ∼ 1◦. Larger off-axis angles are obtained moving
inland Korea (either north, south or west). Figure 1 shows these baseline options using the
same beam configuration as the T2K experiment. Parameters for different baselines and
beam axes are summarized in Figure 1 and Table 1.
The neutrino flavor oscillation probability including atmospheric, solar and interference
terms, as well as matter effects, can expressed using the following equation [18, 19, 11]
P(νe→ νµ) ∼ sin22θ13· T1+ α · sinθ13· (T2+ T3) + α2· T4.(1)
baseline), Okinoshima (middle-baseline) and Korea (long-baseline).
Possible baseline candidates for 100 kton LAr TPC detector, Kamioka (short-
T1 = sin2θ23·sin2[(1 − A) · ∆]
(1 − A)2
T2 = sinδCP· sin2θ12· sin2θ23· sin∆sin(A∆)
·sin[(1 − A)∆]
(1 − A)
·sin[(1 − A)∆]
(1 − A)
T3 = cosδCP· sin2θ12· sin2θ23· cos∆sin(A∆)
T4 = cos2θ23· sin22θ12sin2(A∆)
where α ≡
the mixing angle of the 1st and 3rd generations, while θ12is that for 1st and 2nd, and θ23is
31, ∆ ≡
, A ≡
. ∆m31= m2
1, ∆m21= m2
Same beam configuration as T2K experiment
Off-axis angle 2.5◦
Korea eastern shore
Parameters for baseline and axis with respect to beam of the candidates are
that for 2nd and 3rd generations.
The term which includes T1is the “atmospheric term”, those including T2and T3are
“interference terms”, and the one that includes T4is the “solar term”. The “interference
terms” are sensitive to the δCP phase and play therefore an important role to extract the
For definite calculations, we use the following parameters (we assume that most of these
parameters will be precisely measured within the timescale of the one discussed here):
• θ23is π/4.
• θ12is 0.572904 rad.
31| is 2.5×10−3eV2.
• Earth density for matter effects are 2.8 g/cm3.
• normal hierarchy is assumed unless mentioned otherwise.
Figure 2 shows the oscillation probability as a function of the E(GeV)/ L(km) (neglecting
for the moment matter effects). In the plot, the mixing angle sin22θ13was assumed to be 0.1.
If the distance between source and detector is fixed, the curves can be easily translated to that
for the expected neutrino energy spectrum of the oscillated events. As can be seen, if the
neutrino energy spectrum of the oscillated events could be reconstructed with sufficiently
good resolution in order to distinguish first and second maximum, useful information to
extract the CP phase would be available even only with a neutrino run.
We note that the following observables depend on leptonic CP phase:
• position/height of the first oscillation maximum peak;
• position/height of the second oscillation maximum peak.
The position of the first oscillation minimum is unaffected by it (in this point all terms of
the oscillation probability vanish except the solar term). To effectively experimentally study
these observables, we point out that:
sin22θ13= 0.1 (for other oscillation parameters see text). Black curve shows the case for the
δCP=0, red shows that for δCP=90◦, blue shows that for δCP=270◦, respectively.
Probability for νµ → νe oscillations as a function of the E(GeV)/L(km) for
• the second oscillation maximum peak should end up at sufficiently high energy in order
to be measurable;
• the beam should have a sufficiently wide energy range to cover the 1st and 2nd maxi-
• the neutrino flux should be maximized in order to increase as much as possible the
statistical significance of the first and second maxima.
In order to cover a wider energy range, we accordingly favor a detector location which is near
on-axis. If one assumes that the second oscillation maximum has to be located at an energy
larger than about 400 MeV, the oscillation baseline should be longer than about 600 km. In
addition, in order to collect enough statistics, we choose a baseline which is not too much
longer than above stated.
Taking into account all of the above mentioned considerations, we privilege the Oki-
noshima region: placing a detector in an appropriate location on the island†will probe
neutrino oscillations at a baseline of ∼ 658 km away from the source at an off-axis angle of
5 Neutrino Flux and expected Event Rates
We assume realistic parameters of the J-PARC beam after all accelerator complex up-
grades described in the KEK road-map are accomplished, as follows :
• The average beam power is reaching 1.66 MW;
• A total of 3.45 × 1021POT is delivered on target per year;
• The optimal kinetic energy of the incident protons is 30 GeV.
To remain conservative, we focus on an analysis which uses a neutrino run only during
five years under the best J-PARC beam assumption. An antineutrino beam (opposite horn
polarity) might be considered in a second stage in order to cross-check the results obtained
with the neutrino run (in particular see Section 8 on the mass hierarchy problem). The
parameters of the Okinoshima location and the assumed beam are therefore as follows:
• Distance from J-PARC is 658 km;
• The axis of the beam is off by 0.76◦;
• five years operation with horn setting to neutrino run.
The expected neutrino flux calculated under these assumptions is shown in Figure 3
where the curves correspond to one year run (3.45×1021POT). The black, red, green, blue
lines show νµ, ¯ νµ, νe, ¯ νefluxes, respectively.
The interacting neutrino cross section on Argon was computed using the NUANCE pro-
gramme . We use the followings parameters for the cross section calculation:
• Number of protons is 18, and that of neutrons is 22.
• Medium density is 1.4 g/cm3.
• Nucleus Fermi Gas model with binding energy of 30 MeV and Fermi momentum of
†The exact location of the potential experiment in Okinoshima has not yet been investigated, but we note
that the island has several hills.
0 0.511.522.53 3.54
Figure 3: Calculated neutrino flux under the
assumptions (3.45×1021POT). Black shows
νµ, red shows ¯ νµ, green shows νe, blue shows
¯ νefluxes, respectively.
section, unit by pb. Nuclear effects are taken
Calculated neutrino-Argon cross
Taking these into account, we obtain the neutrino cross section shown in Figure 4. Table 2
shows that total number of charged current (CC) events at Okinoshima for the null oscillation
case and the number of CC νeevents from νµ→ νeoscillations for three different mixing
angles, and in various δCP scenarios, normalized to five years neutrino run at 1.66 MW and
100 kton fiducial mass.
The number of expected events is already a good indicator of the signal, but for this
analysis we use binned likelihood fitting for the energy spectrum, therefore all information
including total number of events is taken into account, as described in the next section.
6 Expected Energy Spectrum of νeCC events
Images taken with a liquid Argon TPC are comparable with pictures from bubble cham-
bers. As it is the case in bubble chambers, events can be analyzed by reconstructing 3D-tracks
and particle types for each track in the event image, with a lower energy threshold of few
MeV for electrons and few tens of MeV for protons. The particle type can be determined
from measuring the energy loss along the track (dE/dx) or from topology (i.e. observing
the decay products). Additionally, the electronic readout allows to consider the volume as a
calorimeter adding up all the collected ionization charge. The calorimetric performance can
Events for 100 kton at Okinoshima normalized to 5 years at 1.66 MW beam power
Beam components (null oscillation)82000
Table 2: Number of CC events at Okinoshima: beam components for null oscillation, and
oscillated events in various δCP scenarios (normalized to five years neutrino run)
be excellent, as we will show in Section 10, depending on event energy and topology.
In order to understand the effect of resolution on physics performance, we show in this
section the νecharged current (CC) event energy spectra using different simple models. One
is the perfect resolution case as a reference, and others are 100 MeV/200 MeV Gaussian
resolution cases for more realistic cases. This “resolution” should include neutrino interac-
tion effects as well as detector resolution. Possible smearing or backgrounds affecting the
measurement of the neutrino energy spectrum are listed below subdivided into 4 classes:
• Neutrino interaction:
– Fermi motion and nuclear binding energy,
– Nuclear interactions of final state particles within the hit nucleus (FSI),
– Vertex nuclear remnant effects (e.g. nuclear break-up signal),
– Neutral Current (NC) π0event shape including vertex activity.
• Detector medium:
– Ionization processes,
– Scintillation processes,
– Correlation of between amount of charge and light,
– Charge and light quenching,
– Hadron transport in Argon and secondary interactions,
– Charge diffusion and attenuation due to impurity attachment.
• Readout system including electronics system:
– signal amplification or lack thereof,
– signal-to-noise ratio,
– signal shaping and feature extraction.
– Pattern recognition
– Background processes (NC π0, νµCC, ...) and their event shape
– Particle identification efficiency and purity
Very few real events in liquid Argon TPCs have so far been accumulated to allow a full
understanding of these complex effects and their interplay. The only sample comes from a
small 50 lt chamber developed by the ICARUS Collab. and exposed to the CERN WANF
high energy neutrino beam which collected less than 100 quasi-elastic events .
It is clear that the energy resolution will depend on several detector parameters, including
the readout pitch, the readout method chosen and on the resulting signal-over-noise ratio
ultimately affecting the reconstruction of the events. We therefore stress that significantly
improved experimental studies with prototypes exposed to neutrino beams of the relevant
energies and sufficient statistics are mandatory to assess and understand these effects. In
the meantime, we give in Section 10 preliminary estimates for potentially achievable energy
resolutions in a liquid Argon medium. The results are based on full GEANT3  simulation
of the energy deposited by final state particles in the detector volume, however do not include
all possible contributing effects.
Figure 5 shows the energy spectra of electron neutrino at the cases of δCP equal 0◦, 90◦,
180◦, 270◦, respectively. Shaded region is common for all plots and it shows the background
from beam νe. Here perfect resolution is assumed for reference to later cases.
If we smear the energy spectra shown in Figure 5 with Gaussian of sigma equals to 100 or
200 MeV independent from original neutrino energy, we obtain spectra shown in Figures 6
and 7. As seen easily, an energy resolution below 100 MeV is crucial since the robustness of
the neutrino oscillation is directly determined by the visible second oscillation peak around
400 MeV in the energy spectrum. In 200 MeV resolution case, the second peak is hidden by
the smearing of the 1st oscillation maximum peak.
7Oscillation parameters measurement from energy
Assuming all others were measured very precisely, there are only two free parameters
sin22θ13and δCP to be fitted using the energy spectra. As shown in Figure 2, the value of
δCPvaries the energy spectrum, especially the first and the second oscillation peaks (heights
and positions), therefore comparison of the peaks determine the value δCP, while the value
of sin22θ13changes number of events predominantly.
To fit the free parameters, a binned likelihood method is used. The Poisson bin-by-bin
probability of the observed data from the expected events (for an assumed pair of values
00.51 1.522.53 3.54
Neutrino Energy (GeV)
Neutrino Energy (GeV)
Neutrino Energy (GeV) Neutrino Energy (GeV)
Number of events
Number of events
Number of eventsNumber of events
case, but δCP = 0◦(top-left), 90◦(right-top),
180◦(left-bottom), 270◦(right-bottom) cases.
Energy spectra at sin22θ13=0.03
Neutrino Energy (GeV)
Neutrino Energy (GeV)
Neutrino Energy (GeV)Neutrino Energy (GeV)
Number of events
Number of events
Number of eventsNumber of events
Figure 6: Energy spectra assuming Gaussian
100 MeV smearing. δCP = 0◦(top-left), 90◦
(right-top), 180◦(left-bottom), 270◦(right-
(sin22θ13, δCP)) is calculated. The fit procedure is validated by testing the result on a
pseudo-experiment. Figure 8 shows one typical pseudo-experiment for the perfect resolution
case (with δCP equals to 0, and sin22θ13equals to 0.1). Best fit gives reasonable result.
At this stage, we only take statistical uncertainty into account for the fitting, thus other
systematic uncertainties like far/near ratio, beam νeshape, energy scale, and so forth are
not considered. Also, oscillated signal and beam νeare only accounted in the fit, i.e. other
background like neutral current π0or beam ¯ νecontamination is assumed to be negligible
compared to beam νecontamination.
As a reference, we first extract allowed regions in the perfect resolution case (See Fig-
ure 9). Twelve allowed regions are overlaid for twelve true values, sin22θ13=0.1, 0.05, 0.02,
and δCP=0◦, 90◦, 180◦, 270◦, respectively. The δCP sensitivity is 20∼30◦depending on the
true δCP value.
Allowed regions are then extracted for 100 and 200 MeV resolution cases. The used
energy spectra are same as Figures 6 and 7. Results are shown as Figures 10 and 11.
One obvious but important issue to be pointed out is the robustness of the fitting.The
fit procedure shows that results could also be extracted with the 200 MeV resolution: this
result is as expected statistically; however, we stress that in this case there is no second
oscillation maximum peak visible in Figure 7. Hence, we think it is mandatory to keep an
energy resolution less than 100 MeV as goal for the credibility of this experiment.
Neutrino Energy (GeV)
Neutrino Energy (GeV)
Neutrino Energy (GeV)Neutrino Energy (GeV)
Number of events
Number of events
Number of eventsNumber of events
Figure 7: Energy spectra assuming Gaussian 200 MeV smearing. δCP = 0◦(top-left), 90◦
(right-top), 180◦(left-bottom), 270◦(right-bottom) cases.
8Investigation of mass hierarchy
The influence of matter on neutrino oscillations was first considered by Wolfenstein .
As is well-known (although never directly experimentally verified), oscillation probabilities
get modified under these conditions. Matter effects are sensitive to the neutrino mass or-
dering and different for neutrinos and antineutrinos. As mentioned earlier, we consider in
this paper the possibility of a neutrino-only run. Hence, we briefly address in this section
the question of normal hierarchy (NH) versus inverted hierarchy (IH). Since we are focusing
on the potential discovery of CP-violation in the leptonic sector, our discussion is geared
towards possible ambiguities that would arise if the mass hierarchy was unknown.
Indeed, Figure 12 illustrates the results of a fit of a pseudo-data with NH by both NH
(black) and IH (red) hypotheses, assuming only the neutrino run. The best fit likelihood
value with one assumed hierarchy is used to calculate the likelihood variation ∆L for both
hierarchies. One could claim that CP-violation is discovered if the experimental results of a
given experiment exclude the δCPphase to be either 0 or 180◦. Hence, the danger is that the
lack of knowledge of the mass hierarchy (or rather the “wrong” choice in hypothesis when
selecting the hierarchy in the fit of the data), gives a result for δCPconsistent with either 0 or
180◦. On the other hand, if fits of a given experiment with both assumed hierarchies provide
neither 0 nor 180◦, one can declare “discovery” although the mass hierarchy could not be
Number of events
shows the best-fit (best oscillation shape + background), and the crosses show the pseudo-
Typical one pseudo-experiment to show the validity of the fitting. Histogram
determined. Alternatively, if it turned out in the actual experiment that one of results of
the fit is consistent with δCP= 0 or 180◦, one would still have the possibility to consider an
anti-neutrino run in order to solve this ambiguity.
Our results indicate that CP-violation would be unambiguously discovered under our
assumption for true values of δCP’s in the region of 90◦and 270◦. Alternatively, in case
of true values of δCP’s near 0◦or 180◦an antineutrino flux would help untangle the two
solutions if the neutrino mass hierarchy was unknown.
9Expected sensitivity to sin22θ13
Although this paper has focused on the possibility to measure the sin22θ13 and δCP
oscillation parameters, it is instructive to estimate the sensitivity of the potential setup in
the case of negative result from T2K. The corresponding sensitivity to discover θ13in the
true (sin22θ13,δCP) plane at 90% C.L. and 3σ is shown in Figure 13, where we assumed 5
years of neutrino run comparing the 100 kton LAr at Okinoshima with increased statistics
in SK at Kamioka and a potential 20 kton LAr at Kamioka.
In order to discover a non-vanishing sin22θ13, the hypothesis sin22θ13 ≡ 0 must be
excluded at the given C.L. As input, a true non-vanishing value of sin22θ13is chosen in the
Figure 9: Allowed regions in the perfect res-
olution case. Twelve allowed regions are over-
laid for twelve true values, sin22θ13=0.1, 0.05,
0.02, and δCP=0◦, 90◦, 180◦, 270◦, respec-
0 0.02 0.040.060.08 0.10.12
resolution (Gaussian sigma) case. Twelve al-
lowed regions are overlaid for twelve true val-
ues, sin22θ13=0.1, 0.05, 0.02, and δCP=0◦, 90◦,
180◦, 270◦, respectively.
Allowed regions in the 100 MeV
simulation and a fit with sin22θ13= 0 is performed, yielding the “discovery” potential. This
procedure is repeated for every point in the (sin22θ13,δCP) plane.
At the 3σ C.L. the sensitivity of the T2K experiment is 0.02 . Continuing to collect
data at SK with an improved J-PARC neutrino beam for another 5 years would improve
the T2K sensitivity by a factor ∼ 2. In comparison, our simulations indicate that a 20 kton
LAr TPC is expected to perform better than this although the masses are comparable, since
the signal efficiency is higher than SK and the NC background is assumed to be negligible
contrary to SK. Even better, a 100 kton LAr TPC at Okinoshima would further improve
the sensitivity by about a factor six compared to SK at Kamioka (or a factor 10 compared
to T2K), thanks to its bigger mass, increased cross-section at higher energies and reduced
off-axis angle, although the neutrino fluxes at the same off-axis angle would be reduced by
a factor ≃ 5 because of the longer baseline.
allowed regions are overlaid for twelve true values, sin22θ13=0.1, 0.05, 0.02, and δCP=0◦,
90◦, 180◦, 270◦, respectively.
Allowed regions in the 200 MeV resolution (Gaussian sigma) case. Twelve
10Neutrino energy resolution in a LAr medium
As mentioned previously, the neutrino energy resolution expected in LAr medium depends on
several parameters and their interplay that ultimately need to be measured experimentally.
In principle, energy and momentum conservation allow to estimate the incoming neutrino
energy in the detector via a precise measurement of decay products, with the exception of
the smearing introduced by Fermi motion and other nuclear effects (nuclear potential, re-
scattering, absorption, etc.) for interactions on bound nucleons. In this section, we try to
estimate the smearing introduced by these effects with the help of exclusive neutrino event
final states distributed with the Okinoshima flux and generated with the GENIE MC 
subsequently fully simulated with GEANT3.
Nuclear effects in neutrino interactions can be roughly divided into those of the nuclear
potential and those due to reinteractions of decay products. Bound nucleons and other
hadrons in nuclei are subject to a nuclear potential. The Fermi energy (or momentum) must
be calculated from the bottom of this nuclear potential well, and the removal of a nucleon
from any stable nucleus is always an endothermic reaction. When hadrons are produced in
the nucleus, some energy is spent to take it out of this well: for a nucleon, the minimum
energy is given by the nucleon separation energy (around 8 MeV), and corresponds to a
sin 2 θ13
cp = 0
cp = 180
sin 2 θ13
sin 2 θ13
sin 2 θ13
Figure 12: Mass hierarchy investigation with neutrino run only. If fits with both hierarchy
hypotheses provide neither 0 nor 180◦, one can declare discovery of CP violation in the
leptonic sector. If any of the fits results in a δCP of 0 or 180◦, then an anti-neutrino run
could be envisaged.
θ13Sensitivity - 3σC.L.
SK - 22.5 kton
at Kamioka - 1.66 MW
20 kton LAr
at Kamioka - 1.66MW
100 kton LAr
at Okinoshima - 1.66MW
Figure 13: Sensitivity to sin2(2θ13) for 5 years of neutrino run comparing 100 kton LAr at
Okinoshima with SK at Kamioka and 20 kton LAr at Kamioka.
nucleon at the Fermi surface. In this case, the daughter nucleus is left on its ground state.
More deeply bound nucleons, leaving a hole in the Fermi sea, correspond to an excitation
energy of the daughter nucleus, and an additional loss of energy of the final state products.
This energy is then spent in evaporation and/or gamma deexcitation. Thus, the energy of
the final state products is expected to be always slightly smaller than for interactions on free
nucleons, and spread over a range of about 40 MeV. Correspondingly, the Fermi momentum
is transferred to the decay products and compensated by the recoil of the daughter nucleus.
Additional momentum distortions come from the curvature of particle trajectories in the
Reinteractions in the nuclear medium also play an important role. Interaction products
can lose part of their energy in collisions, or even be absorbed in the same nucleus where they
have been created. This is particularly true for pions, that have an important absorption
cross section on nucleon pairs, while kaons have smaller interaction probability.
Once final state products have exited the nucleus, they will propagate in the medium
detailed event simulation (see text) for neutrino oscillations with δCP = 0◦: perfect recon-
struction (top-left), LAr using deposited energy(right-top), ibid but with charged pion mass
correction (left-bottom) using final state lepton only (like in WC) (right-bottom) cases.
Full GEANT simulations of the reconstructed neutrino energy spectra from
with the possibility that further interactions occur.
In the case of the liquid Argon TPC, the medium is fully homogeneous and the detector
is fully active. All deposited energy in the medium (above a certain threshold) will be
eventually collected. Several methods can be adopted to reconstruct the neutrino energy
and we list in the following three:
1. Final state lepton: this method, traditionally used in large Water Cerenkov detectors
like Kamiokande, IMB or SK, relies on the precise measurement of the energy and
direction of the outgoing lepton and kinematically constrains the incoming neutrino
energy by assuming a quasi-elastic configuration:
M − Eℓ+ pℓcosθℓ
where Eℓ, pℓand cosθℓare the energy, momentum and scattering angle of the outgoing
lepton. This method is sensitive to the final state configuration and to Fermi motion
(up to ≃ 240 MeV/c) which randomizes the direction of the outgoing lepton.
2. Momentum conservation: neglecting the incoming neutrino mass, one obtains
Eν= |?Pν| = |? pℓ+
where ? pℓis the momentum of the outgoing lepton, ? phis the momentum of the outgoing
hadron h and?PF is the Fermi motion of the hit nucleon (up to ≃ 240 MeV/c). This
method is also sensitive to Fermi motion since the recoiling remnant hit nucleus (with
momentum −?PF) is not measured‡.
3. Energy conservation: using a calorimetric approach, one can sum the deposited
energies of all outgoing particles. In a tracking-calorimeter this is obtained by summing
the dE/dx measurements along each ionizing track to obtain the associated kinetic
energies T ≡?(dE/dx)dx. One should identify final state particles in order to take
into account their rest masses. This method is sensitive to the nuclear binding energy
of the decay products and is intrinsically more precise than the above method relying
on momenta for the energies considered here. Mathematically one can write this result
= Etot− M = Eℓ+
Eh− M = Eℓ+
(Th+ mh) − M
= Eℓ+ TN+
(Tπ± + mπ±) +
All methods are sensitive to potential re-interactions of the outgoing hadrons within the
nucleus. In order to estimate the potential of a fully homogeneous and sensitive medium
like the liquid Argon TPC, we performed full simulations of neutrino interactions: the event
4-vector are generated with the GENIE MC and final state particles (after nuclear reinter-
action) were propagated through the liquid Argon medium with GEANT3. The geometry of
the setup is an infinite LAr box and at this stage we have neglected charge quenching (i.e.
‡We point out that if one uses the knowledge on the direction of the incoming neutrino, the total (missing)
transverse momentum can be kinematically constrained to zero, thereby providing a handle compensate for
the transverse component of the Fermi motion. The longitudinal component of the momentum, which is not
negligible at low neutrino energies, remains however undetermined.
we assume a linear response to dE/dx). We also assume a 100% efficiency to identify final
state charged pions.
The achievable incoming neutrino energy resolution in the liquid Argon medium has
been estimated using νeCC events distributed with the νµflux expected at the Okinoshima
location. The results, using the energy conservation method, are shown in Figure 14.
11 Other baseline options
Although Okinoshima looks a very good candidate as mentioned so far, the favorable
geography would allow, in principle, for a few baseline candidates for the detector locations;
Kamioka for a relatively ”short” baseline, Okinoshima for a medium baseline and Korea for
the longest baseline. It is worthwhile to compare the physics potential at Okinoshima with
other scenarios with the same strategy and same assumed beam conditions, i.e. five years
neutrino run at 1.66 MW.
Figure 15 illustrates the analyzed allowed regions using each option. A perfect resolu-
tion was assumed here to directly compare the physics potential of the various sites. As
seen, sensitivity of δCP is similar between Okinoshima and Korea, but that in Kamioka is
much worse than others. This indicates the analysis strategy using only neutrino run is not
suitable for Kamioka case, as was expected since the OA2.5◦available at Kamioka does not
allow to cover 1st and 2nd oscillation maxima peaks. As for the sensitivity of the sin22θ13,
Okinoshima is better than Korea, due to the increased statistics at the shorter baseline.
00.020.040.06 0.080.10.12 0.14
Okinoshima or Korea, assuming the T2K beam optics is kept. Perfect resolution is assumed
to illustrate the physics potential from each site selection.
Allowed regions if the location of the 100 kton LAr TPC were at Kamioka,
12Proton Decay Discovery Potential
Grand Unification of the strong, weak and electromagnetic interactions into a single unified
gauge group is an extremely appealing idea [25, 26] which has been vigorously pursued
theoretically and experimentally for many years. The detection of proton or bound-neutron
decays would represent its most direct experimental evidence. The physics potentialities of
very large underground Liquid Argon TPC was recently carried out with detailed simulation
of signal efficiency and background sources, including atmospheric neutrinos and cosmogenic
backgrounds . It was found that a liquid Argon TPC, offering good granularity and energy
resolution, low particle detection threshold, and excellent background discrimination, should
yield very good signal over background ratios in many possible decay modes, allowing to
reach partial lifetime sensitivities in the range of 1034− 1035years with exposures up to
1000 kton×year, often in quasi-background-free conditions optimal for discoveries at the
few events level, corresponding to atmospheric neutrino background rejections of the order
of 105. Multi-prong decay modes like e.g. p → µ−π+K+or p → e+π+π−and channels
involving kaons like e.g. p → K+¯ ν, p → e+K0and p → µ+K0are particularly suitable, since
liquid Argon imaging provides typically an order of magnitude improvement in efficiencies
for similar or better background conditions compared to Water Cerenkov detectors. Up to
a factor 2 improvement in efficiency is expected for modes like p → e+γ and p → µ+γ
thanks to the clean photon identification and separation from π0. Channels like p → e+π0or
p → µ+π0, dominated by intrinsic nuclear effects, yield similar efficiencies and backgrounds
as in Water Cerenkov detectors. Thanks to the self-shielding and 3D-imaging properties of
the liquid Argon TPC, the result remains valid even at shallow depths where cosmogenic
background sources are important. In conclusion, a LAr TPC would not necessarily require
very deep underground laboratories even for high sensitivity proton decay searches.
Once again, we stress the importance of an experimental verification of the physics po-
tentialities to detect, reconstruct and classify events in the relevant GeV energy range. This
experimental verification will require the collection of neutrino event samples with high statis-
tics, accessible e.g. with a detector located at near sites of long baseline artificial neutrino
In this paper, we have reported our case study for a new future long baseline neutrino
and nucleon decay experiment with a 100 kton Liquid Argon TPC located in the region of
Okinoshima using the J-PARC Neutrino Facility.
Our study assumes a realistic upgrade of the J-PARC Main Ring operation yielding an
average 1.66 MW beam power for neutrinos coupled to the new far detector at Okinoshima,
which would then see the beam at an off-axis angle of ≃ 1◦and at a baseline of ≃ 658 km.
In this first discussion, we focused on the possibility to measure the θ13 and δCP pa-
rameters. Our strategy is based on the precise measurement of the energy spectrum of
the oscillated events, and in particular on the comparison of the features of 1st maximum
oscillation peak with those of the 2nd maximum oscillation peak. In order to efficiently
determine and study those features, we rely on the excellent neutrino event reconstruction
and incoming neutrino energy resolution, as is expected in the case of the fully-sensitive and
fine charge imaging capabilities of the liquid Argon TPC.
Beyond this case study, we intend to perform more detailed simulations of the perfor-
mance of the experiment including for example different sources of systematic errors and
sources of backgrounds which have been up-to-now neglected.
The construction and operation of a 100 kton liquid Argon TPC certainly represents
a technological challenge at the present state of knowledge of the technique.
become a realistic option, we stress the importance and necessity of further R&D and of
dedicated experimental measurement campaigns. At this stage, we intend to pursue our
investigations on a ton-scale prototype based on the novel double-phase readout imaging
method. Options to operate small devices in the J-PARC neutrino beam are being assessed
in parallel. In addition, we have started to address the possibility of an “intermediate”
prototype, presumably in the range of 1 kton of mass, based on a similar but scaled down
design of the potential 100 kton detector.
For it to
Part of this work was supported by the Swiss National Foundation and ETH Zurich.
 G. Steigman Ann.Rev.Astron.Astrophys, 14(1976)339; A.D. Sakharov Pisma Zh.ETF
 Y. Itow et al., “The JHF-Kamioka neutrino project,” arXiv:hep-ex/0106019.
 M. Apollonio et al. [CHOOZ Collaboration], “Limits on neutrino oscillations from the
CHOOZ experiment,” Phys. Lett. B 466, 415 (1999)
 T. Hasegawa, Talk given at the 4th International Workshop on Nuclear and Particle
Physics at J-PARC (NP08). Slides available at http://j-parc.jp/NP08/
 J. Burguet-Castell, M. B. Gavela, J. J. Gomez-Cadenas, P. Hernandez and O. Mena,
“On the measurement of leptonic CP violation,” Nucl. Phys. B 608, 301 (2001)
 H. Minakata and H. Nunokawa, “Exploring neutrino mixing with low energy super-
beams,” JHEP 0110, 001 (2001) [arXiv:hep-ph/0108085].
 V. Barger, D. Marfatia and K. Whisnant, “Breaking eight-fold degeneracies in neu-
trino CP violation, mixing, and mass hierarchy,” Phys. Rev. D 65, 073023 (2002)
 W. J. Marciano, “Extra long baseline neutrino oscillations and CP violation,”
 M. V. Diwan et al., “Very long baseline neutrino oscillation experiments for precise
measurements of mixing parameters and CP violating effects,” Phys. Rev. D 68 (2003)
 M. Ishitsuka, T. Kajita, H. Minakata and H. Nunokawa, “Resolving neutrino mass
hierarchy and CP degeneracy by two identical detectors with different baselines,” Phys.
Rev. D 72, 033003 (2005) [arXiv:hep-ph/0504026].
 K. Hagiwara, N. Okamura and K. i. Senda, “Physics potential of T2KK: An extension
of the T2K neutrino oscillation experiment with a far detector in Korea,” Phys. Rev.
D 76, 093002 (2007) [arXiv:hep-ph/0607255].
 A. Meregaglia and A. Rubbia, “Neutrino Oscillations With A Next Generation Liq-
uid Argon TPC Detector in Kamioka or Korea Along The J-PARC Neutrino Beam,”
 S. Amerio et al., ”Design, construction and tests of the ICARUS T600 detector”, Nucl.
Instrum. Meth. A 527 (2004) 329 and references therein.
 A. Rubbia, “Experiments for CP-violation: A giant liquid argon scintillation, Cerenkov
and charge imaging experiment?,” arXiv:hep-ph/0402110.
 A. Rubbia, “ArDM: A ton-scale liquid argon experiment for direct detection of dark
matter in the universe,” J. Phys. Conf. Ser. 39 (2006) 129 [arXiv:hep-ph/0510320].
 A. Marchionni, Talk given at the 4th International Workshop on Nuclear and Particle
Physics at J-PARC (NP08). Slides available at http://j-parc.jp/NP08/
 A. Meregaglia and A. Rubbia, “Neutrino oscillation physics at an upgraded CNGS
with large next generation liquid argon TPC detectors,” JHEP 0611, 032 (2006)
 M. Freund, “Analytic approximations for three neutrino oscillation parameters and Download full-text
probabilities in matter,” Phys. Rev. D 64, 053003 (2001) [arXiv:hep-ph/0103300].
 A. Cervera, A. Donini, M. B. Gavela, J. J. Gomez Cadenas, P. Hernandez, O. Mena
and S. Rigolin, “Golden measurements at a neutrino factory,” Nucl. Phys. B 579, 17
(2000) [Erratum-ibid. B 593, 731 (2001)] [arXiv:hep-ph/0002108].
 F. Arneodo et al. [ICARUS-Milano Collaboration], “Performance of a liquid argon time
projection chamber exposed to the WANF neutrino beam,” Phys. Rev. D 74, 112001
 R. Brun et al., GEANT3 Reference Manual,. CERN Program Library Long Writeup.
W5013, DD/EE/84-1, 1987. For hadronic secondary interactions, we select GFLUKA.
 L. Wolfenstein, “Neutrino oscillations in matter,” Phys. Rev. D 17 (1978) 2369.
Act.Phys.Pol.B Vol.37,8:2349, 2006
Nucl.Phys.B Proc.Suppl. 159:217-222,2006;C.Andreopoulos,
 J. C. Pati and A. Salam, “Unified Lepton-Hadron Symmetry And A Gauge Theory Of
The Basic Interactions,” Phys. Rev. D 8, 1240 (1973).
 H. Georgi and S. L. Glashow, “Unity Of All Elementary Particle Forces,” Phys. Rev.
Lett. 32, 438 (1974).
 A. Bueno et al., “Nucleon decay searches with large liquid argon TPC detectors at
shallow depths: Atmospheric neutrinos and cosmogenic backgrounds,” JHEP 0704,
041 (2007) [arXiv:hep-ph/0701101].