Sterile n search with the ICARUS
T600 in the CNGS beam
R. Sulej
National Centre for Nuclear Research,
Otwock/Swierk Poland
for the ICARUS Collaboration
XV International Workshop on Neutrino Telescope Venice 11-15 March 2013
Outline
1. Sterile n and LSND/MiniBooNE anomaly
2. ICARUS T600 LAr TPC at LNGS
 Detector operation and performance
3. Sterile neutrinos search in the CNGS beam
Experimental search for the „LSND anomaly” with
the ICARUS detector in the CNGS neutrino beam,
Eur. Phys. J. C (arXiv:1209.0122)
 ne signal selection
 Results
4. Conclusions
Neutrino Telescopes, 11-15 March 2013
2
The new frontier
 The discovery of a Higgs boson at CERN/LHC has crowned the
successful Standard Model (SM) and will call for a verification of
the Higgs couplings to the gauge bosons and to the fermions.
 Neutrino masses and oscillations represent today a main
experimental evidence of physics beyond the Standard Model.
 Being the only elementary fermions whose basic properties are
still largely unknown, neutrinos must naturally be one of the main
priorities to complete our knowledge of the SM.
 Albeit still unknown precisely, the incredible smallness of the
neutrino rest masses, compared to those of other elementary
fermions points to some specific scenario, awaiting to be
elucidated.
 The astrophysical importance of neutrinos is immense, so is their
cosmic evolution.
Neutrino Telescopes, 11-15 March 2013
3
Sterile neutrinos ?
 Sterile neutrinos: hypothesized in a seminal paper by B. Pontecorvo
in 1957, as particles not interacting via any of fundamental
interactions of SM except gravity.
 Since per se they may not interact directly, they are extremely
difficult to detect. If they are heavy enough, they may also
contribute to cold dark matter or warm dark matter.
 Sterile n’s may mix with ordinary neutrinos via a mass term. Evidence
may be building up by “anomalies” observed by several experiments:
- sterile n(s) with Dm2≈10-2-1 eV2 from ne observation in nm beam
accelerator experiment, „LSND anomaly”
- n disappearance may have been observed in nuclear reactors and
very intense (megaCurie) e-conversion n sources with maybe
comparable mass differences.
Neutrino Telescopes, 11-15 March 2013
4
Overall evidence
Combined
evidence
s ≈ 3.8
Zhang, Qian, Vogel: „Reactor (…) with known θ13” (arXiv:1303.0900) → s ≈ 1.4 ?
Combined evidence for some possible anomaly:
many standard deviations.
Neutrino Telescopes, 11-15 March 2013
5
Neutrino related anomalies
Allowed LSND
and MiniBooNE
signals
ICARUS T600 search for the LSND like nm→ne signal from the
CNGS nm beam at 730 km and 10 ≤ En ≤ 30 GeV is presented.
Neutrino Telescopes, 11-15 March 2013
6
ICARUS Collaboration
M. Antonelloa, P. Aprilia, B. Baibussinovb, M. Baldo Ceolinb,, P. Benettic, E. Calligarichc,
N. Cancia, S. Centrob, A. Cesanaf, K. Cieslikg, D. B. Clineh, A.G. Coccod, A. Dabrowskag,
D. Dequalb, A. Dermenevi, R. Dolfinic, C. Farneseb, A. Favab, A. Ferrarij, G. Fiorillod, D. Gibinb,
A. Gigli Berzolaric,, S. Gninenkoi, A. Guglielmib, M. Haranczykg, J. Holeczekl, A. Ivashkini,
J. Kisiell, I. Kochanekl, J. Lagodam, S. Manial, G. Mannocchin, A. Menegollic, G. Mengb,
C. Montanaric, S. Otwinowskih, L. Perialen, A. Piazzolic, P. Picchin, F. Pietropaolob, P. Plonskio,
A. Rappoldic, G.L. Rasellic, M. Rossellac, C. Rubbiaa,j, P. Salaf, E. Scantamburloe,
A. Scaramellif, E. Segretoa, F. Sergiampietrip, D. Stefana, J. Stepaniakm, R. Sulejm,a,
M. Szarskag, M. Terranif, F. Varaninib, S. Venturab, C. Vignolia, H. Wangh, X. Yangh,
A. Zalewskag, A. Zanic, K. Zarembao.
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
p
Laboratori Nazionali del Gran Sasso dell'INFN, Assergi (AQ), Italy
Dipartimento di Fisica e INFN, Università di Padova, Via Marzolo 8, I-35131 Padova, Italy
Dipartimento di Fisica Nucleare e Teorica e INFN, Università di Pavia, Via Bassi 6, I-27100 Pavia, Italy
Dipartimento di Scienze Fisiche, INFN e Università Federico II, Napoli, Italy
Dipartimento di Fisica, Università di L'Aquila, via Vetoio Località Coppito, I-67100 L'Aquila, Italy
INFN, Sezione di Milano e Politecnico, Via Celoria 16, I-20133 Milano, Italy
Henryk Niewodniczanski Institute of Nuclear Physics, Polish Academy of Science, Krakow, Poland
Department of Physics and Astronomy, University of California, Los Angeles, USA
INR RAS, prospekt 60-letiya Oktyabrya 7a, Moscow 117312, Russia
CERN, CH-1211 Geneve 23, Switzerland
Institute of Theoretical Physics, Wroclaw University, Wroclaw, Poland
Institute of Physics, University of Silesia, 4 Uniwersytecka st., 40-007 Katowice, Poland
National Centre for Nuclear Research, A. Soltana 7, 05-400 Otwock/Swierk, Poland
Laboratori Nazionali di Frascati (INFN), Via Fermi 40, I-00044 Frascati, Italy
Institute of Radioelectronics, Warsaw University of Technology, Nowowiejska, 00665 Warsaw, Poland
INFN, Sezione di Pisa. Largo B. Pontecorvo, 3, I-56127 Pisa, Italy
Neutrino Telescopes, 11-15 March 2013
7
ICARUS @ LNGS: the first LARGE LAr-TPC
Hall B
cathode
LN2 vessels
1.5m
E
cryogenics
(behind)
readout electronics
T300
E
T300
readout wire arrays
T300 module: two TPCs with the common cathode
 Two identical T300 modules
(2 TPC chambers per module)
 3 readout wire planes at 0°, ±60°, 3mm
plane spacing (for each TPC chamber):
  53000 wires, 3 mm pitch
 LAr active mass 476 t:
 2 Induction planes, 1 Collection
 (17.9 x 3.1 x 1.5 for each TPC) m3;
 drift length = 1.5 m;
 PMT for scintillation light (128 nm):
 Edrift = 0.5 kV/cm; vdrift = 1.6 mm/ms  (20+54) PMTs
 trigger and t0
 Total energy reconstruction of events from charge integration.
 Full sampling, homogeneous calorimeter; excellent accuracy for contained events.
Neutrino Telescopes, 11-15 March 2013
8
ICARUS LAr-TPC detection technique
 2D projection for each of 3 wire planes per TPC
 3D spatial reconstruction from stereoscopic 2D projections
 charge measurement from Collection plane signals
Collection (top view)
Induction 2 (top view)
Induction 1 (frontal view)
CNGS nm charge current
interaction, one of TPC’s shown
Neutrino Telescopes, 11-15 March 2013
9
II MODULE
max drift
I MODULE
60 ppt O2 equiv.
Key feature: LAr purity
Impurities from electro-negative molecules: O2, H2O, CO2
Ultra High Vacuum techniques; highly efficient filters based on Oxysorb/Hydrosorb;
Continuous purification by recirculation in liquid & gas phases;
During data taking: te > 5ms ≈ 60 ppt O2 equivalent impurities  max. 17%
signal attenuation after 1.5m for the longest e– drift length, Edrift = 0.5kV/cm.
Neutrino Telescopes, 11-15 March 2013
10
CNGS data taking: 2010-2012
CNGS data taking (Oct. 2010 – Dec. 2012)
• large sample of n interactions
• superluminal n searches
1. Cherenkov-like e+e– emissions: P. L. B711 (2012) 270
2. timing measurement: P. L. B713 (2012), 17
3. precision measurement: JHEP 11 (2012) 049
detector live-time > 93%
total 8.6 x 1019 pot collected
• n oscillations
n m → n t t - → e– n e n t
nm → ne ”LSND/MiniBooNE” anomaly
2010: Oct. 1 – Nov. 22
5.8 1018 pot
in press: Eur. Phys. J. C, arXiv:1209.0122
2011: Mar. 19 – Nov. 14
4.44
1019
pot
Neutrino Telescopes, 11-15 March 2013
2012: Mar. 23 – Dec.3
3.54 1019 pot
11
LAr TPC performance
Total energy reconstr. from charge integration
 Full sampling, homogeneous calorimeter with
excellent accuracy for contained events
Tracking device
nmCC energy
deposit
 Precise 3D topology and accurate ionization
 Muon momentum via multiple scattering
Measurement of local energy deposition dE/dx
 e/g remarkable separation (0.02 X0 samples)
 Particle identification by dE/dx vs range
1 m.i.p.
Low energy electrons:
σ(E)/E = 11% / √ E(MeV)+2%
Electromagn. showers:
data dE/dx vs residual range
compared to Bethe-Bloch
curves
2 m.i.p.
σ(E)/E = 3% / √ E(GeV)
Hadron showers:
σ(E)/E ≈ 30% / √ E(GeV)
Neutrino Telescopes, 11-15 March 2013
12
Precise particle tracking and identification
CNGS
beam
primary vertex
kaon
cathode
New 2D => 3D approach for LAr TPC
(in press: AHEP, arXiv:1210.5089)
3D object driven by optimization of its 2D projections
(no need for drift matching)
Induction
Collection
3D reconstruction
p
dE/dx based PID
K

μ
Neutrino Telescopes, 11-15 March 2013
13
dE/dx of muon tracks in CNGS
Ek = 2.14GeV, length = 10.0 m
Collection
long muon from nm CC interaction
Induction2
MC:
m=2.37 MeV/cm; RMS=1.37
data:
μ=2.32 MeV/cm; RMS=1.31
Induction1 (a)
muon track:
• reconstructed with
Collection and Induction2
• projected to Induction1
Induction1 (b)
dE/dx in 3 mm track segments (3D), after
removing d rays and e.m. cascades
MC – data agreement on the level of 2%
signal/noise from Landau + gaussian ≈ 10 14
Neutrino Telescopes, 11-15 March 2013
e/g separation: dE/dx in cascade initial part
Ek = 102 ± 10 MeV
θ
• MC: single electrons (Compton)
• MC: e+ e– pairs (g conversions)
• data: EM cascades (from 0 decays)
2 m.i.p.
0 reconstruction:
pπo = 912 ± 26 MeV/c
mπo = 127 ± 19 MeV/c²
θ = 28.0 ± 2.5º
Ek = 685 ± 25 MeV
Collection
• ~0.02 X0 sampling; g’s separated from vtx;
• g rejection increases with energy;
• e selection efficiency const with energy;
1 m.i.p.
2 m.i.p.
1 m.i.p.
MC
Neutrino Telescopes, 11-15 March 2013
15
A search for LSND effects with ICARUS at CNGS
nm→ne signal from the CNGS nm beam at: L = 730 km, 10 ≤ En ≤ 30 GeV
 Differences w.r.t. the LSND experiment:
• L/En ≈ 1 m/MeV at LSND, but L/En ≈ 36.5 m/MeV at CNGS
• LSND-like short distance oscillation signal
averages to: sin2(1.27Dm2new L /E) ≈ ½
and:
<P>nm→ne ≈ 1/2 sin2(2qnew)
 When compared to other long
baseline results (MINOS and T2K)
ICARUS operates in a L/En region in
which contributions from standard
neutrino oscillations are not yet too
relevant.
 Unique detection properties of LArTPC allows to identify individual ne
events with high efficiency.
Neutrino Telescopes, 11-15 March 2013
16
Data sample
• CNGS neutrinos: almost pure nm peaked in the range 10 ≤ En ≤ 30 GeV
– beam associated ne about ~1%
– CNGS events triggered by PMT signal inside SPS proton extraction gate
– empty events rejection, few n interactions/day
• Present sample: 1091 neutrino events (within 6% with MC expectation)
– data from 2010 and 2011, 3.3 x 1019 pot (out of total 8.6 x 1019 pot collected)
– nm CC events identified by L>250 cm track from primary vtx, no hadr interaction
– the signature of the nm→ne signal is observed visually
• Fiducial volume:
– primary vertex min. 5 cm from each side, min. 50 cm from downstream wall
to allow shower identification
• Energy deposition cut: < 30 GeV
– optimized signal/background ratio (ne contamination extends to higher En),
only 15% signal events rejected
– excellent MC/data agreement for deposited energy  negligible syst. error on
signal and bkg expectations
– 10% syst. error due to ne beam component
Neutrino Telescopes, 11-15 March 2013
17
Signal selection
Visibility cuts for nm→ne event identification:
1. Charged track from the vertex, m.i.p. over 8 wires (dE/dx < 3.1 MeV/cm
excluding d-rays), developing into EM-shower
2. Spatial separation (150 mrad) from other tracks at the vertex, at least in one
transverse (±60°) view
Expected conventional ne– like events:
(cuts: vtx in fiducial volume; visible energy ≤ 30 GeV)
3.0 ± 0.4 ev. due to the intrinsic ne beam contamination
1.3 ± 0.3 ev. due to q13 oscillations, sin2(q13) = 0.0242 ± 0.0026
0.7 ± 0.05 ev. of nm→nt oscillations with electron production, 3n mixing predictions.
The total is therefore of 5.0 ± 0.6 expected events
(uncertainty on the NC and CC contaminations included)
Selection efficiency, validated on MC sample of ne events: h =0.74±.05
in a good approximation, h independent from shape of energy spectrum
3.7 ± 0.6 expected background events after cuts.
Neutrino Telescopes, 11-15 March 2013
18
Signal selection efficiency in MC simulation
• ne events generated according to nm spectrum in order to reproduce
oscillation behaviour;
• full physics and detector MC simulation in agreement with data
• 122 events over 171 simulated inside the detector, satisfy fiducial volume
and energy cuts;
• visibility cuts: (3 independent scanners), leading to 0.74 ± 0.05 efficiency;
• < 1% systematic error from dE/dx cut on the initial part of cascade;
• no ne-like events selected among NC simulated sample of 800 events.
Typical MC event
En= 11 GeV, pt= 1.0 GeV/c.
(only the vertex region is shown)
Neutrino Telescopes, 11-15 March 2013
19
Signal selection efficiency check in MC simulation
• automatic cuts mimicking data selection, large sample of MC
C1: inside fiducial volume and Edep < 30 GeV;
C2: no identified muon, at least one shower;
C3: one shower: initial point (or g conversion point) < 1 cm from vtx,
separated from other tracks;
C4: ionisation signal from single mip in the first 8 wires.
Signal selection efficiency (after the fiducial and energy cuts):
0.6/0.81 = 0.74, in agreement with the visual scanning method.
Neutrino Telescopes, 11-15 March 2013
20
2 events observed in data
a
b
(a) Etot = 11.5 ± 1.8 GeV,
pt = 1.8 ± 0.4 GeV/c
(b) vis Etot = 17 GeV,
pt = 1.3 ±0.18 GeV/c
In both events: single electron shower in the transverse plane clearly
opposite
to hadronic component
Neutrino Telescopes, 11-15 March 2013
21
The observation is presently compatible with the
absence of a LSND anomaly
• The limits on no. of events due to the
LSND anomaly are respectively
3.41 (90% CL) and 7.13 (99% CL)
• Given the present data sample of, the
limits to the oscillation probability are:
Pνμ→νe ≤ 5.4 x 10-3 (90% CL)
Pνμ→νe ≤ 1.1 x 10-2 (99% CL)
• The exclusion area is shown for the
plot Dm2 vs sin2(2q). At small Dm2
ICARUS strongly enhances the
probability with respect to
short baseline experiments.
with (Dm2, sin2(2q)) = (0.11 eV2, 0.10) as
many as 30 events should have been seen.
Neutrino Telescopes, 11-15 March 2013
22
Result combined with the low energy excess
1:
1-5:
6:
6-9:
MiniBooNE best fit in the combined 3+1 model
excluded by present ICARUS result (=> excessive osc. P @ large L/En)
best value including ICARUS result
compatible with present ICARUS
Neutrino Telescopes, 11-15 March 2013
23
LSND anomaly surviving area
allowed MiniBooNE
allowed LSND 90%
allowed LSND 99%
limit of
KARMEN
present ICARUS
exlusion area
ICARUS result strongly limits the window of possible parameters for LSND
anomaly indicating a narrow region (Dm2, sin22q) = (0.5 eV2, 0.005) where
there is an overall agreement (90% CL):
• the present ICARUS limit
• the limits of KARMEN
• the positive signals
of LSND and MiniBooNE Collaborations
Neutrino Telescopes, 11-15 March 2013
24
Conclusions
• Successful operation of ICARUS T600 on CNGS beam.
• LAr TPC allows unambiguous identification of individual ne
events with high efficiency.
• ICARUS result strongly limits the window of possible
parameters for LSND anomaly indicating a narrow region:
(Dm2, sin22q) = (0.5 eV2, 0.005)
where there is an overall agreement (90 % CL) between:
– the present ICARUS limit
– the limits of KARMEN
– the positive signals of LSND and MiniBooNE
• Definitive answer to sterile n search: ICARUS T600 + anew 150 t
detectors exposed to new 2GeV n beam at CERN SPS at 1.6 km and
460 m baseline.
Neutrino Telescopes, 11-15 March 2013
25
Thank you!
Neutrino Telescopes, 11-15 March 2013
26
ICARUS at CERN
New CERN SPS 2 GeV neutrino facility in North Area
100 GeV primary proton beam fast extracted from SPS in North Area:
C-target station next to TCC2 + two magnetic horns, 100 m decay pipe,
15 m of Fe/graphite dump, followed by m stations.
Interchangeable n andn focussing.
Far position: 1600 m
ICARUS-T600 detector +
magnetic spectrometer
Near position: 330 m
150t LAr-TPC detector to be build
+ magnetic spectrometer
Neutrino Telescopes, 11-15 March 2013
27
ICARUS at CERN
● Clarify the surviving LSND/MiniBooNE (Dm -sin 2q) region
2
2
 far: ICARUS-T600 + near: T150 at CERN SPS
 shorter distances: 1600 m and 330 m from target
 lower neutrino energies (~ 2 GeV) than in CNGS beam.
→ increased event rate, reduced overall event multiplicity, enlarge
the angular range improving substantially ne selection efficiency
proton recoil
minimum ionizing
● In absence of oscillations: n spectra at different distances equal
 independent of specific experimental event signatures
 without any MC comparisons
Neutrino Telescopes, 11-15 March 2013
28
ICARUS at CERN: mass and mixing angle
● Dual method, unlike LNSD and MiniBooNE, determines both the mass
difference and the value of the mixing angle.
● Events rates shown both for [background] and [oscill.+background] at
d=330 m and d=1600 m and the Dm2 = 0.4 eV2, sin2(2q) = 0.01
● For d=1.6 km, En < 5 GeV and 4.5 1019 pot (1 y)
a ne oscillation signal of
≈1200 events is expected above 5000 backgr. events
T150 near detector
Neutrino Telescopes, 11-15 March 2013
T600 far detector
29
ICARUS at CERN: expected sensitivities
e-appearance:
1 year nμ beam (left)
2 year anti-nμ beam (right)
for 4.5 1019 pot/year,
3% syst. uncertainty on n
energy spectrum
nμ→ne
nμ→ne
LSND allowed region fully
explored in both cases
e-disappearance:
1 year nμ beam
1 year nμ + 2 years anti-nμ beams
(right)
ne→ne
combined “anomalies”:
from reactor ns, Gallex and
Sage experiments.
Neutrino Telescopes, 11-15 March 2013
30
L/En nm
ICARUS window
Neutrino Telescopes, 11-15 March 2013
31
Precision study, e.g.: nm CC Quasi-elastic interactions
CNGS data
Collection
p
m
proton:
visible part Ek = 121±24 MeV
inelastic interaction at: Ek = 45±9 MeV
init direction w.r.t. beam axis:
cos:(-0.601, -0.779, -0.179) ± 3o
Induction2
precision study possible thanks to:
 particle identification
 energy, momentum reconstruction
 vertex region analysis
QE: very clean events
• single visible particle identified as a stopping/interacting proton
• events with high energy and higher particle multiplicity and
reinteractions
Neutrino Telescopes, 11-15 March 2013
32