SLIDE 1
Neutrino Oscillation Experiments at Reactors and Accelerators - - PowerPoint PPT Presentation
Neutrino Oscillation Experiments at Reactors and Accelerators - - PowerPoint PPT Presentation
Neutrino Oscillation Experiments at Reactors and Accelerators Gaston Wilquet IIHE-Universit Libre de Bruxelles IV mes Rencontres du Vietnam, Hanoi, July 2000 Contents CHOOZ and PALO VERDE long baseline experiments Search for e
SLIDE 2
SLIDE 3
To-morrow Byron Lundberg will give a seminar at FermiLab on "Results from DONUT: First Direct Evidence of the Tau Neutrino” This was expected and did no happen explicitly in Sudbury at Neutrino 2000
SLIDE 4
Long base line experiments at nuclear power plants of Chooz and Palo Verde Motivation: νe disappearance in (∆m2 , sin2 2 θ) parameter space indicated by νµ disappearance in atmospheric experiments νµ → νe ?
Neutrino Oscillation Experiments at Nuclear Reactors
SLIDE 5
Neutrino Oscillation at Reactors: pros and cons
MeV E MeV E
e thresh
e
3 8 . 1 ≈ > < =
+
ν
- Eν ≈ few MeV ⇒ Access to low ∆m2 at medium L
⇒ Below µ, τ thresholds: only disappearance
- High flux, but small σ
- 4π source ⇒ detector mass ÷ L2
- Disappearance ⇒ good knowledge of absolute ν flux and e+ energy spectrum
⇒ or multi-L experiment ( ≥ 2 detectors or reactors) ⇒ no sensitivity at high ∆m2 (not serious problem with ∆m2 ≤ 0.01 eV2
- Cheep and well known ν source
Calculated and measured ν flux and energy spectrum at L=0 known to ~ 2% (Bugey 1995)
e x
ν → ν
2 2
E 3 MeV m 0.003 eV L 1000 m ≈ ∆ ≈ ≈ ≈
6.5 10-42 cm2
ν Interaction spectrum
1 20 1 20
10 6 10 200 3 . 3 rate Fission 3 . 3 1
− − ⇒
≈ ≈ ⇒ = s s MeV GW GW GW P
e therm elec
ν
SLIDE 6
Detection of neutrinos from nuclear reactors
1953 : F.Reines and C.L. Cowans discover the neutrino at Savannah River nuclear power plant
Detectors
- vessel filled with liquid scintillator
doped with neutronphage
- shielding (bunker, underground)
+ active veto: cosmic rays, reactor n, natural radioactivity
n e p
+
→ +
e
ν
s c i n t i l l a t i o n ' s e
- e
a n n i h i l a t i o n : 2
- f 0 . 5 1 1 k e V
+ −
γ γ
known) E ( s ' capture nuclear
∑
→
γ
γ
Space and delayed time correlation Signal
MeV E MeV E
e thresh
e
3 8 . 1 ≈ > < =
+
ν
Cerenkov light
SLIDE 7
The Long Base Line CHOOZ Experiment
- Phys. Let. B466 (1999) 415
SLIDE 8
CHOOZ detector
- 1 detector - 2 reactors (8.5GW) : L= 998, 1114m
∆L=116.7m
- rock overburden: 300 m water equivalent
0.4 cosmic µ m-2 s-1
- 5 tons Gd-doped liquid scintillatior (0.09%)
- 17 tons liquid scintillator : contain γ from n
PMT radioactivity shield
- 90 tons active cosmic-ray muon veto
E 8MeV
γ =
∑
5t 17t 90t
SLIDE 9
: full power: Event rates 24.7±0.7 eve reactors off: 1.2 even nt ts s/day /day
Data taking: April 1997 - July 1998 Reactor 1 ON 2058.0 h 8295 GWh Reactor 2 ON 1187.8 4136 Reactors 1 & 2 ON 1543.1 8841 Reactors OFF 3420.4 Background estimates Response calibration: γ, n and γ-n radioactive sources (60Co, 252Cf, Am/Be) En
abs time dependence monitoring ( ) with n from cosmic : σE = 0.5 MeV
E 8MeV
γ =
∑
SLIDE 10
No event selection
Reactor ON Reactor OFF e+-like E (MeV) e+-like E (MeV) n-like E (MeV) n-like E (MeV)
ν selection
@ > 30 cm from wall, n - e+ distance < 100 cm n - e+ delay in (2-100) µs E(e+-like) in (1.3 - 8) MeV E(n-like) in (6-12) MeV Main background fast spallation n in rock + p from n scattering (e+ like) + n capture 2991 candidates (287 reactors off) Efficiency: 69.8% n-like E (MeV) n-like E (MeV) ν region ν region ν region ν region
SLIDE 11
Ee+ spectrum
+
- inverse β-decay cross-section
- simulation of detector response
Ee+ (MeV) Ee+ (MeV) Ee+ (MeV)
reactor ON OFF
- data
— MC
background subtracted
e e
E spectrum measured R E spectrum expected
+ +
=
R
e flux known to 1.4%
ν
- daily evolution of core isotopic evolution
- instantaneous fission rate from thermal power
- ν yield from measured β spectra of main isotopes
R 1.010 0.028 (stat) 0.027 (syst) = ± ±
No oscillation signal
SLIDE 12
Analysis Methods
A - Compare unfolded Ee+ absolute spectra of both reactors to expectation Systematic uncertainty on absolute normalisation: ~2% Two “independent” measurements B - Ratio of spectra Most systematic cancel No sensitivity at large ∆m2 C - Compare unfolded Ee+ spectra shapes of both reactors to expectation Intermediate sensitivity
SLIDE 13
Chooz exclusion plot
θ 2 sin2
4
10 . 7
−
⇐ 1 . ⇓
) (
2 2
eV m ∆
10
- 4
10
- 3
10
- 2
10
- 1
1 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
sin2(2θ) δm2 (eV2) analysis A 90% CL Kamiokande (multi-GeV) 90% CL Kamiokande (sub+multi-GeV) νe → νx analysis B analysis C
A — absolute spectra B — spectra ratio C — spectra shape Kamiokande 90%
SLIDE 14
The Long Base Line Palo Verde Experiment
G.Gratta Neutrino 2000 F.Boehm et al. hep-ex/000322
SLIDE 15
Palo Verde detector
E 8 M eV
γ =
∑
- 1 detector - 3 reactors (11.6 GW) : L= 750, 890m
∆L= 110m
- rock overburden: 32 m water equivalent
22 cosmic µ m-2 s-1
- 11.3 tons Gd-doped liquid scintillation (0.1%)
- oil and 105 tons water buffer: γ and n shield
shield PM radioactivity
- optically segmented detector (900x12.7x25.4 cm3)
⇒ background suppression
SLIDE 16
Analysis based on the knowledge of the flux form the known reactors power ⇒ True expected event number compared to Observed number of candidates corrected for detector efficiencies (MC) Difficulty : No period with all reactors off to measure simply the reactors off background. efficiency 0.075 0.077 0.112 0.111 Unknowns :
- Background
- Overall normalisation
within systematic uncertainty
R = 1.04 + 0.03 (stat) + 0.08 (sys) ⇒ No oscillation
SLIDE 17
Run till end Summer 2000 2 new reduced power periods Not likely to do better than Chooz
SLIDE 18
Three neutrinos families analysis
(at least) 3-flavour analysis Reactor experiments exclude two-family νµ → νe oscillation in parameters region where νµ deficit in atmospheric experiments favours two-family νµ → ντ (or νs)
SLIDE 19
3-flavour mixing parametrization
3 1 2 1 3 2 2 1 3 1 3 2
2 6 (8) parameters 1 3 Dirac (Majorana) 1 (3)phases
' '
k k k , e kk' k k k , . kk k ,k k
U m U e, , U m
U
α α µ α α τ
ν ν ν ν ∆ ν ν α µ τ ν ν ∆
= = = ≠
= = = = =
∑ ∑ ∑
CKM-like matrix standard parametrization)
12 13 12 13 13 23 13 23 13
1 3 genreation nunbers
i ij ij ij ij
c c s c s e c cos s c i, j , s sin c c
U
δ
θ θ
−
= = = =
SLIDE 20
3-family flavour at the strong mass hierarchy approximation
2 2 2 2 2 3 1 3 2 2 2 2 2 1 2 2
- 3
2
- 6
2
3 1 2
if e.g. 10 eV atmospheric neutrinos e.g. 10 eV solar neutrinos m m m m m m m m m m
m m ,m
∆ = − ≈ − δ = − ∆ δ
- 2
2
- 3
2 2
L/E region where m E/ L causes oscillation (e.g. atmospheric neutrinos m 10 eV , E 1GeV, L=1000km) and m E/ L
P(
α
⇓ ∆ ∆ = = δ ≈ ⇓
ν
2 2 2 3 3 2 2 eff 3 3
) 4 U U sin (1.27 m E/L) sin 2 4 U U
α β≠α β α αβ β
→ ν ≈ ∆ θ = Physics governed by:
- ∆m2
- flavour contents of ν3
- effective 2-flavour
like oscillation
2
mν
1
ν
2
ν
3
ν
2
m ∆
2
m δ
SLIDE 21
Effective 2-family atmospheric νµ disappearance in 3-family mixing
2 eff 2 2 2 2 e 3 e3 13 23 2 eff 2 2 2 4 3 3 23 13 2 e3 2 e3 2 2 3 3 2 2 3 3 2 2 3 3
sin 2 4 U U sin 2 sin sin 2 4 U U sin 2 cos U small (reactors) U 0.1 ? 4 U U 1 (full mixing atmospheric) U U 0.5 ? U U 1
µ µ µτ µ τ µ τ µ τ µ τ
θ = = θ θ θ = = θ θ < ≈ ⇒ ≈ ≈ + ≈
E.Lisi, Neutrino 2000
E.Lisi, G.Fogli, ...
SLIDE 22
H.Sobel Neutrino2000 Space is left for U2
e3 ≠ 0
SLIDE 23
Conclusions:
- No evidence for νe disappearance in LBL reactor experiments
- Reactor + Atmospheric neutrino experiments
+ in 3-flavour strong mass hierarchy model room left for a small νe contents in ν3
- No more constraining data to be expected from reactors in near
future
SLIDE 24
SLIDE 25
Compare somehow conflicting results from two similar experiments: KARMEN2: no signal LSND: statistically significant signal
2 2
Search for
- scillation at rather large
0 1
e
m ~ . eV
µ
ν ν ∆ → >
LSND: G.Mills Neutrino 2000 KARMEN2: K.Eitel
Neutrino Oscillation Experiments at Low Energy Accelerators (Beam Stoppers)
SLIDE 26
Conceptual design:
p target shielding ν detector p 800 Mev π ν 30 m
<Eν> ~ 30 MeV
2 2
1eV E L m ≈ ≈ ∆
SLIDE 27
( )
µ
π µ ν
µ
ν
+ +
→
- (
)
µ
π µ ν
µ
ν
−
→
( )
e
e
µ
µ ν ν
µ
ν
+ +
→ ( )
e
e
e
µ
µ ν ν
ν
+ +
→
( ) e
e
µ
π ν
ν
+ +
→
- (
)
e
e
e π
ν
ν
−
→
( )
e
e
e
µ
µ ν ν
ν
− −
→ ( )
e
e
µ
µ ν ν
µ
ν
− −
→
- scillation
e µ
ν ν →
ν from Decays in Flight
LSNDenergy spectra
spectra are for LSND
( )
µ
π µ ν
µ
ν
+ +
→
( )
e
e
µ
µ ν ν
µ
ν
− −
→
( )
e
e
e
µ
µ ν ν
ν
+ +
→ ( )
e
e
µ
µ ν ν
µ
ν
+ +
→
( )
e
e
e
µ
µ ν ν
ν
− −
→
( )
e
e
e π
ν
ν
+ +
→
( X) N
µ
µ ν
µ
ν
−
→
- scillation
e µ
ν ν →
ν from Decays/Captures at Rest
53MeV
53 e E MeV
ν
µ
ν ν
<
→
53 e E MeV
ν
µ
ν ν
>
→
53MeV
The ν Energy spectra
SLIDE 28
Detectors
vessel filled with oil + liquid scintillator doped with neutronphage several light signal by arrays of PMT’s
n e p
+
→ +
e
ν
s c i n t i l l a t i o n ' s e
- e
a n n i h i l a t i o n : 2
- f 0 . 5 1 1 k e V
+ −
γ γ
known) E ( s ' capture nuclear
∑
→
γ
γ
Space and delayed time correlation
Signal
MeV E MeV E
e thresh
e
3 8 . 1 ≈ > < =
+
ν
Cerenkov light
KARMEN2
H Gd
KARMEN2
E =8MeV
γ
∑
E =2.2MeV
γ
SLIDE 29
Main Karmen2 pro: Time structure of ISIS p source
20 ms
from π + from µ +
[ ] [ ]
background measured in 600 and from signal recorded in 600 10 6 with 2.2 s slope cosmic background measure
e e
t ,t ns t ns,t . s
µ µ µ
ν ν ν ν ν µ µ
- +
- →
+ +
- [
]
d in 10 6 20 small duty cycle small cosmic background t . s,t ms µ + +
- →
- Segmented detector: better n-e- space correlation
- High scintillator concentration: × 4 better E resolution
and signal from QE from
prompt
- scillation
e
e
p
e
µ
ν
ν ν
+
→
SLIDE 30
Main LSND pro: Electron ID and direction Homogeneous detector + low scintillator concentration ⇒ Cerenkov ring as e+ signature
- ×3 Mass
- L=18m (instead of 30m) ⇒ lower ∆m2
- 3% of DIF π+ → µ+ νµ (instead of 0.1%)
⇒ higher energy beam component ⇒ νµ → νe oscillation via νe C → e- N
SLIDE 31
LSND detector at LAMPF, Los Alamos KARMEN-II detector at ISIS, RAL
SLIDE 32
Statistical analysis difficulties
KARMEN-2
- no signal and very low expected background : place an Upper Limit
- (for long: 0-event observed sample, 3 expected background)
- non physical max likelihood : sin2 2θ < 0
LSND
- signal region in parameter space computed from rapidly oscillating
likelihood function with many local maxima
Profusion of recent papers and workshops
- n our to fix C.L. limits from likelihood functions
(starting G.J.Feldman & R.D.Cousins, Phys Rev D57(1998)3873)
SLIDE 33
KARMEN-2 results e+ e+ n n
All backgrounds measured except intrinsic
SLIDE 34
KARMEN-2 exclusion plot
+ Unified frequentist approach (F.-C.)
2 3 3 2
sensitivity =1.7 2 1 3 10 10 @ large m sin . ∆ θ
− −
⋅ < ⋅
2 2 11 1
2 1
- sc
- sc
promt delayed prompt k k back
- sc
back promt delayed prompt k exp
- isson
back back
L(sin , m ) { r f ( E ,E ,t , t, r ) ( r r ) f ( E ,E ,t , t, r ) } P ( r | r ) θ ∆ ∆ ∆ ∆ ∆
=
= ⋅ + = − ⋅ ×
∏
SLIDE 35
New LSND global analysis of all event categories with a common Electron trigger and Ee in [20-200] MeV
from decay ( 2.2 s) from decay (16 from capture (186 s)
GS
e N ms ) n µ µ β γ µ prompt from decay (16 from capture (186 s)
GS
e N ms ) n β γ µ
e- trigger
SLIDE 36
Global fit ♦to all relevant distributions
- E (e,β,γ,µ)
- ∆t (e - β,γ,µ)
- ∆r (e - β,γ,µ)
- θ (ν − e- )
- R : ratio of likelihood of prompt (e - ) and delayed events (γ)
to be correlated/accidental ♦ for all electron trigger events categories ♦ with parameters:
- π+/π− production ratio
- all DAR and DIF π and µ
- efficiencies µ, e, β, γ
Oscillation signal : “ e γ ” events with large R
New LSND global analysis of all event categories with a common Electron trigger and Ee in [20-200] MeV
SLIDE 37
Variables entering in R R for no-oscillation channel ∆t (e - γ) ∆r (e - γ) PMT hits ÷ E
SLIDE 38
The oscillation signal in Ee in [20-60] MeV
Excess of events at R > 10 (large e-γ correlation likelihood) beam on beam off expected ν excess of total background background events 83 33.7 16.6 32.7 ±9.2
Event excess is Ee dependent
R>10 Ee(MeV) L/Eν (m/MeV)
Fit to full R distribution Posc=0.0025±0.0006 ±0.0004
SLIDE 39
LSND signal region
2 2 3 3
sensitivity =1.7 10 KARMEN2 @ larg 2 1 3 1 e m sin . ∆ θ
− −
⋅ < ⋅
2 2 1 1 1
2
Nevent
- sc
- sc
e k k back i i e k i back exp normal i i i
L(sin , m ) { r f ( E ,R,L ,cos ) r f ( E ,R,L ,cos ) } N ( r N | N )
ν ν ν ν
θ ∆ θ θ
= = =
= ⋅ + ⋅ × ⋅
∏ ∑ ∏
Relaxing cuts:
- Ee in [20-200] MeV
- R>0
+ +
Cut at ∆L= 2.3 (90%) 4.6 (99%)
Compatible 1993-95 & 1996-98 signals
SLIDE 40
Joined likelihood 3 “sensible” common favoured regions 90% 99% LSND alone KARMEN2, NOMAD,BUGEY excluded
- K. Eitel hep-ex-990906
and New J. Phys. 2 (2000)1
Preliminary joined KARMEN2-LSND analysis
SLIDE 41
Conclusions
- LSND signal in ∆m2 ~ eV2 is one of the 3 oscillation signals
- No evidence that result is wrong
- Allowed LSND parameters space domain will not be fully covered by
KARMEN2 (⇒ spring 2001): not enough statistics given background
- Need for a joined analysis based on a common likelihood function
based on the final data,
- Need new experiment(s) with higher sensitivity:
MiniBOONE approved at FERMILAB from 2001 I216 proposal at CERN PS 2001
SLIDE 42
2
3
e s
m
µ τ
∆ ν ν ν ν
In the mean time either
- r
A. De Rujula Nu2000
SLIDE 43
+
S e a r c h f o r
- s c i l l a t i o n s
f r o m d e c a y s a t r e s t
e
e
µ
ν ν µ
+
→ →
The neutrino production chain (numbers are for LSND)
p+N 800 Mev
11%
π+
DAR 97% DIF 3%
+
µ ν µ
DAR 100%
+
µ ν µ
DAR 100%
+
e
e
ν ν µ
+
e
e
ν ν µ
89%
π-
Absorbed 95% DIF 5%
−
µ ν µ
Absorbed 88%
−
e
e
ν ν µ
DAR 12%
SLIDE 44
Neutrino Oscillation Experiments at High Energy Accelerators
CHORUS and NOMAD short baseline experiment Search for νµ−ντ oscillation ντ appearance in same high energy ντ free νµ beam at the CERN SPS Wide Band Neutrino Beam
Motivation (early 1990’s) Search for “hot dark matter” candidates with mν > 1 eV with ~50 times better sensitivity than E531: Posc(νµ−ντ) > 10-4
SLIDE 45
CERN Wide Energy-Band Neutrino Beam Line
1
relative
E [GeV] 27 0.056 19 0.009 ~40 0.002 ~32 3 < >
e e ν µ µ τ
Φ ν ν ν ν ν
- 6
10 ~43
Irreducible prompt ντ background from DS → τ ντ less than 0.1 event in 4 years well below sensitivity
2 2
Maximum sensitivity @ 27 50 0 6 GeV m eV . km ∆ ≈ ≈
SLIDE 46
- M. Mezzetto Neutrino 2000
P.Astier at al. CERN-EP-2000-049)
SLIDE 47
ντ signal extraction technique: excess of events in kinematics box ⇒ precise energy/momentum & good particle ID
SLIDE 48
Examples of sensitive kinematics variables
µ
ν
17 4 huge CC backgroun Not used d . %
µ τ
τ µ ν ν
− −
→
e
- +
- τ
ν ν
Decay channels τ π π π (nπ )ν 15.2% To 49 5 kinematics very 17 8 small CC backgrou NC 82 5 d tal n
e
h ( n ) . e % . % % .
τ τ
τ π ν τ ν ν
− − − −
→ → → ≠
SLIDE 49
- Precise simulation of kinematics of signal and background
In kinematics box where signal expected: background known to O O (10-5)
Data Simulator
Most systematic (hadron shower simulation, Fermi motion, …) cancel out
Replace in CC Data (DS) and MonteCarlo (MCS) samples by Monte-Carlo (backgound NC) (signal CC (background CC
e
- S,B
Data
) e )
µ τ
ν ν ν
µ ν τ
ε
−
- =
S,B s,b DS MC S,B MCS
ε ε ε ντ signal extraction technique also requires
SLIDE 50
Analysis technique
(signal/background) ) ( ) is the probability for Event classification based on l an event described by kinemati
- g likelyhood r
cs to be signal (background) c i h at o
S i S B B i i i
( X ( X n n ) X X l l λ
- =
- L
L L L
- osen as most discriminating variables
product of (quasi)independent multivariate , ,...) Tail of (large signal, small background probs) devided into (independent) signal bins
i j
pdf ( X X ln λ
- L
L Binning obtained from maximum sensitivity Decay channels and s (based on MC à la Feldman-Cousins) à la Feldman-Cousins (unified frequentist approach) Bli ignal bins combine nd analys : data d is in
- potential signal region not looked at untill analysis fully defined
SLIDE 51
- e
e
τ
τ ν ν
−
→
Background from CC (1.5% of CC Transverse momentum imbalance
e
e )
µ
ν ν −
Background and from NC isolation e e - γ
−
Signal region “blindly” selected divided in 6 bins 6 events in bins of DATA box found after box definition
Event selection: efficient e- ID (20%) e-/π- rejection ~ 106 e-/π0 rejection ~ 104
SLIDE 52
- n
inclusive (B.R.=49.5%) h ( ) τ τ π ν
−
→
0-γ sample
Very new event selection: h- selection since CERN-EP 2000-049 shown at Nu2000 π 0 likelihood (2 γ ) ρ likelihood (>1 γ ) h- candidate likelihood better e-/µ− rejection
e- and h- channels contribute similarly to sensitivity
SLIDE 53
Search Summary
Events expected: 55.2±5.2 Events found: 58 Nτ = 14937 = expected number of signal events if Posc(νµ−ντ) = 1
Channels/Bins with very low background
SLIDE 54
New CHORUS (different statistical method)
2 4 2
2 4 06 10 for large m sin .
µτ
θ ∆
−
> ⋅
@90% C.L. Unified frequentist approach
SLIDE 55
More to expect from NOMAD:
- and
still being improved c hannel e
τ
µ τ τ τ π π π ν
ν ν ν ν
+
→
→ →
polarization in N X
µ
Λ ν µ Λ
−
→
talk by R.Petti in PS6
in pr g ress
- e
µ
ν ν →
talk by Minh-Tam Tran in PS6
SLIDE 56
Collaboration Collaboration
Belgium (Brussels, Louvain-la-Neuve), CERN, Germany (Berlin, Münster), Israel (Haifa), Italy (Bari, Cagliari, Ferrara, Naples, Rome, Salerno), Japan (Toho, Kinki, Aichi, Kobe, Nagoya,Osaka, Utsunomiya) , Korea (Gyeongsang), The Netherlands (Amsterdam), Russia (Moscow), Turkey (Adana, Ankara,Istanbul)
CHORUS experiment CHORUS experiment Search for Search for ν νµ
µ →
→ ν ντ
τ oscillation
- scillation
L.Ludovicci Neutrino 2000 E.Eskut at al. CERN-EP-2000-0??)
SLIDE 57
ντ Direct detection technique
Observation of the τ-lepton track produced in CC ντ interactions in 770 kg nuclear emulsion target : “kink” topology
“Kink”
ττ= 2.9 10−13 < βγcττ > ≈ 1.5 mm
0)
( ( ) B.R. 18% 50% n 14% not yet in u e n l d d c h
τ µ τ τ
π π π µ ν ν π ν ν π
− − − + −
Prompt ντ background <~ 0.1 event in 4 years
SLIDE 58
CHORUS target
4 stacks of emulsion interleaved with fibre trackers and emulsion interface trackers
emulsion stack
trackers
- 1 stack : 142 x 144 x 2.8 cm3
= 36 plates of 790 µm 80 80 µ µm m 100 100 µ µm m
MIP MIP ~35 ~35 grains / 100 / 100 µ µm m
emulsion 350 µm emulsion 350 µm base 90 µm Grain size ~ 1. µm Grain measurement ~ 0.3 µm Angular resolution ∼ 1.5 mrad
1 plate
SLIDE 59
Spaghetti Calorimeter Veto plane
µ
- µ
- h-
h-
Emulsion target Scintillating fibre tracker Air core magnet hadron spectrometer Muon spectrometer CHORUS detector : event kinematics measurement Hadron Sign and momentum
Air-core magnet hadron spectrometer
∆p/p =√(0.035.p(GeV/c)+0.222)
Muon ID, sign and momentum
Iron-core muon spectrometer ∆p/p~10%-15% (p<70 GeV)
Showers energy, missing Pt
Lead&fibers “spaghetti” calorimeter
∆E/E=32%/√E (hadrons) ∆E/E=14%/√E (electrons) ∆qhadr~60 mrad @10 GeV
SLIDE 60
Automatic Emulsion Data Taking (K.Niwa and Nagoya University
0 µm 21 µm 54 µm View size: 30x40 mm2 Focal depth : ~3 mm
ν beam
100 µm 16 S
SLIDE 61
Analysis strategy
150 2
pos
m mrad
θ
σ µ σ = =
- Event reconstruction and loose kinematics selection
1 µ− with pµ<30 GeV/c no µ− and at least 1 h- with 1<p<20 GeV/c
- Track predictions at emulsion trackers for
tracks reconstructed in scintillating fibre trackers
- Tracks found and followed by automatic microscopes
in 3 successive interface emulsion trackers up to stack entry
- Followed back plate by plate in target to find vertex
- Automatic search for a “kink” decay topology: 3% of events
- Events with “kink” are analysed manually: 1% of selected
events retained as candidates
- Precise kinematics analysis of candidates
τ µ
µ ν ν
−
(n ) h
τ
ν π
−
SLIDE 62
Automatic Vertex Location
- Follow-up track, plate by plate to the vertex
- 100 µm most upstream of each target plate are scanned
- Vertex defined by the first plate out of two consecutive plates
where a track segment is not found
SLIDE 63
Kink Finding - Parent Search (Large Angle-Long Path kinks)
100 µm most upstream of the vertex plate are searched for all track segments in a cone of width ∝ 1/P Segments with small impact parameters w.r.t. the follow-up track → Candidate track parent track → Manual microscope inspection (3% of scanned events) Manual candidate event selection:
- “clean” 1-prong kink: no sign of nuclear interaction
1% of inspected events
SLIDE 64
Year
1994 1995 1996 1997
All POT / 1019
0.81 1.20 1.38 1.67
5.06
Good emulsion
97% 73% 100% 100%
~93%
Expected Ncc / 103
120 200 230 290
840
Emulsion trigger / 103
422 547 617 719
2,305 1µ to be scanned
66,911 110,916 139,527 151,105
468,459 1µ scanned so far
88% 55% 81% 83%
77% 1µ vertex location and kink search
20,400 21,610 41,558 52,789
136,357 0µ to be scanned
19,846 29,350 37,143 36,073
122,412 0µ scanned so far
60% 58% 79%* 67%*
67% 0µ vertex location and kink search
3,024 4,424 8,704 7,054
23,206
Data
SLIDE 65
Background evaluation and reduction
- “White kink
- hadron elastic scattering with no sign of nuclear activity
- badly known rate is measured at large distance from vertex
- number within τ- decay path computed by MC
WK
τ
- Charm decays and white kinks reduction in the h- channel
- Ph dependent τ candidate decay path cut
such that 80% of the true τ are retained ∀ Ph
- Φ(τ-Hadron shower) in transverse plane > 90 °
Cuts optimised for maximum sensitivity “without looking at data”
- π and K decays
- Pt (daughter-parent) > 250 MeV/c : reject 100 %
- Charm background
- primary lepton not identified and, if D+ , charge of secondary wrong
- D- produced by νµ/νe beam component
SLIDE 66
Background
<0.05 1µ 0µ 0.11 0.03 0.03 0.30
0.8
0.05 0.05
- +
D from CC with primary missed
e /
/ e
µ
ν ν µ
+
- from
CC with primary missed and wrong charge for decay
e
D / / e / h
µ
ν ν µ µ
+ − + +
CC and NC associated missed and + neutrals D / D D / h µ
+ − − −
→
"White kinks" elastic scattering with no nuclear activity h−
Prompt beam from decays
S
D
τ
ν
negligible
SLIDE 67
branching Nb cross--section location efficiency kink finding ratios events ratios ratios efficiency
1 5014 2004 7018
- sc
max CC ,h max
- sc
i i i i CC i i
P N / N A N ( P ) BR N A
ντ µ
µτ τ τ µ τ τ µτ τ µ ν
σ ε σ
− −
=
= = = ⋅ < >= + =
∑
Results
Channel Observed Expected background 0 0.1 5 014
max
N h
τ
µ −
−
1.1 5 004
Upper limit on 90% 2 4 (T.Junk, NIM A434 (1999) 435)
t
N @ C.L. . =
4
3 4 10
- sc
P .
µτ −
< ⋅
SLIDE 68
4 4 4 4 2 2 4 4 4 2 4 4
NOMAD CHORUS CHORUS (Feldman-Cousins) (Junk) (F.-C.) sensitivity (F.C.) 2 6 "to compare" 3 4 10 2 0 10 10 3 7 10 3 7 10 2 03 10 2 @ large 4 06 10 6 8 10 4 0 10
- sc
. . . P . sin m . . . m . ( .
ντ
θ ∆ ∆
− − − − − − − − −
⋅ ⋅ ⋅ ⋅ ⋅ ⋅ ⋅ ⋅ ⋅
2 )
@ full mixing 0 6 0 6 0 6 eV . . .
µ τ
ν ν →
2 2
m ( eV ) ∆
2 2
sin θ
Junk 1.2 expected background 90% C.L. given Feldman_Cousins 0 observed 2 4 1 4 N @ N . .
τ τ
< <
see talk by R.Petti in PS6
SLIDE 69
e τ
ν ν →
2 2
m ( eV ) ∆
2 2
sin θ
There are 0.9% of events in the beam
4
2 6 10
- sc
e
P .
τ −
< ⋅
SLIDE 70
More to expect from CHORUS: 2 years Phase-2 analysis launched
Among other things:
- improved kink (τ) finding efficiency
- 3-hadron and e- decay channels
in emulsion thanks to new upgrade in automatic microscope technology
Reach Posc < 10-4
0,001 0,010 0,100 1,000 10,000 100,000 1994 1996 1999 2000
#frames/s
*+ s
Other Physics CHARM physics in emulsion (D
- bservation published)
and in calorimeter as target (J/ production submitted) Form factors Trident ( ) production Search for heavy neutral l ψ µ µ ν
+ −
- eptons ...
SLIDE 71
≥2 segments connected 1.5 mm 1 . 5 m m all segments not passing through small impact parameter