Electromagnetic Calorimeter for T2K 280m OffAxis Detector

1
Electromagnetic Calorimeter for T2K 280m Off-Axis
Detector

• Kevin McFarlandUniversity of Rochester
• on behalf of the 280m working group

2
Outline

• Physics motivation
• Requirements for ECAL
• Possible Designs and Performance
• Plans

3
Physics Goals
4
Physics Goals Involving EM Events

• Background Events for Electron Neutrino
  Appearance at Super-K
• primarily relies on ability to measure inclusive
  and exclusive p0 production processes
• this is most critical measurement in which ECAL
  plays a role
• we heard from Nakaya-san that current
  uncertainties are at least 20-30
• want eventually better than 10

5
Physics Goals Involving EM Events

• Electron Neutrino Flux in Beam
• measure electron neutrino energy distribution as
  a test of beamline simulation
• this is important, but a cross-check
• Photon Background Events at Super-K
• can also be measured by studying exclusive D
  production at 280m in other final states

6
Super-K Electron Backgrounds

• What causes backgrounds? (n.b. D?gN not included)

study by Shiozawa (Oct 2003 280m meeting)
7
Why are these p0 Background?

• Many, but not all, are asymmetric
• Lose 2ndphoton

Higher-energy g
Lower-energy g
ltEggt480MeV
ltEggt40MeV
study by Hiraide (Kyoto)
8
Example p0 Background Event
study by Shiozawa(Oct 2003 280m meeting)
9
Background p0 Kinematics

• p0s in background events are typically harder and
  more forward than average p0
• but noticethat theyare notvery
  highenergyeither!
• by designof cuts

p0 momentum
p0 direction
ltPpgt0.54GeV/c
ltqpgt49deg
study by Hiraide (Kyoto)
10
So what events must 280m Observe?

• Inclusive (mostly NC) p0 production
• 300-1000 MeV momentum
• not in exotic corners of phase space
• but need to map out this phase space
• helpful but not essential to be efficient at 280m
  for asymmetric decays, say to Eg30 MeV
• want to know about other particles that might or
  might not fail to make a 2nd ring
• soft muons (high y CC events)
• additional pions

11
p0 Background Prediction at 280m

• measure p0 rate and kinematics at 280m
• need to correct for far/near
• use CC production of p0 and p where
  neutrinoenergy is known to find Endependence of
  production
• form predictions for visible (2-ring) and
  background p0 at Super-K
• the former serves as a cross-check

(flux) x (distance)2
12
Energy Spectrum Correction

• See 30 correctionsto far/near ratio across
  the neutrino spectrum
• Note that MC says thatneutrinos producing p0
  fallin a broad range of energies
• Correction cannot beignored. C.f., 2km

CC NC-1p0 NC-coherent NC-DIS
study by Hiraide (Kyoto)
nm induced (after all cuts)
13
Single EM Events

• As previously noted, want to also observe
• Key issue here is reducing backgrounds from
  single p0
• again, need to maintain good efficiency for
  second low energy photon from asymmetric decays
• full reconstruction of exclusive final states may
  help, e.g.

14
Requirements Summary
15
ECAL Requirements

• Reconstruct a pure sample of inclusive p0
• in a sample where other final state particles can
  be noted (e.g., in FGD)
• efficiency for asymmetric decays is secondary
• Observe exclusive states with p0 and e-
• low backgrounds, e.g., p0?e- fakes, for these
• Good energy and angle resolution for p0,e-

16
ECAL in 280m Detector
17
Conceptual Off-Axis 280m Detector
Magnet (and side MRD)
Magnet (and side MRD)
Tracker (TPC or chambers)
Fine Grained detector w/ or w/o water target
Iron shield for m-ID
Scintillator
18
Why is an ECAL needed?

• Simply, the radiation length of fully active
  detectors is too large to contain e, g
• X0 of polystyrene extruded scint. is 42.4 cm
• X0 of water is 36.1cm
• example

NC-1p0,En0.82GeV ,Pp00.45GeV/c
Even with a limited fiducial volume(one X0 from
edge of FGD) Two g FGD conversions 51 One g
FGD conversion 35
FGD (Scint.)2mx2mx2m
g
e-
p0
e
g
study by Hiraide (Kyoto)
EM calorimeter (Pb)
19
Containment of Photons

• No surprise full containment requires a lot of
  material
• 10-15 cm Pb equiv.
• Design must be compact, so layered approaches
  studied
• low density tracking
• then EM calorimetry
• increasingly coarse at back of shower?

study by Uchida (Imperial)
20
Possible ECAL Designs
21
Designs Studied

• Internal radiators (interspersed in FGD)
• McGrew, Yanigasawa
• KOPIO-like design (shashlyk modules)
• Kudenko, Konaka et al.
• Strawman Pb/Strawtube design
• Uchida, Boyd, Barker, Wark
• MINERvA-like design (radiators in edge of FGD)
• Chvojka, McFarland

22
Internal Radiators (TASS)

• EM radiators are internal to the design
• photons convert near vertex for best resolution

study by McGrew
23
Bars and Shashlyk Modules

• Shashlyk Modules, very fine sampling calorimeter
  (0.3mm radiators)
• Long bars parallel to beam of Pb plus
  scintillator
• well suited to external ECAL mounted to magnet

Kudenko talk, 280m pre-meeting
24
Bars and Shashlyk Modules
Kudenko talk, 280m pre-meeting
Segment of ?-catcher
?

• long bars with dual end readout also provides
  precise TOF measurement
• additional vertex constraint from this under study

25
UK Strawman

• Not so much a design as a departure point for
  simulations
• Design has a total of 150mm of Pb
• layers of 1mm, 3mm or 10mm radiator
• gaps of 9, 12 or 15 mm(gaseous Ar detector
  volume)
• straws of various granularities
• Study, e.g., the effectof using or not
  usingpulse height on pointingresolution

26
MINERvA-like Design

• Extends FGD into ECAL region
• with appropriateaddition ofradiators asrings
  or plates
• Design and simulation fairly mature (stale?)
• biggest concern is at 90 degrees relies
  significantly on using MRD (5cm iron granularity)
  as a leakage catcher
• Design is straightforward and nearly free

ECAL
FGD
27
Common Features

• Containment requires absorber layers
• Shower fluctuation effects in tracking photons
  are non-trivial
• E.g., event at righthas two disjointtracks,
  but truephoton directionsplits the difference
• Leads tosurprisingly badpointing resolution

FGD/Pblayers
FGD (Scintillator)
study by Chvojka (Rochester)
28
Pointing Resolution

• Limitations in pointing resolution are
  surprisingly independent of detectorif detector
  is compact

Detector thickness 18X0 (9 modules 2X0, 10 cm
width)
MINERvA-like
study by Kudenko
study by Chvojka(Rochester)
29
Coping with Pointing Resolution?

• Impact poor p0 mass and vertex resolution (in
  absence of p0 mass constraint)
• KOPIO design
• very sparse material tracker after first
  radiator
• but this adds a lot of space
• TASS design
• internal radiators keep conversion point near
  vertex so short lever arm for mistakes
• Just swallow hard and gut it out?
• not been shown that backgrounds to p0 are a
  concern!

30
Kinematic Resolutions Still Good

• Energy resolution still excellent
• e.g., MINERvA-like 6/sqrt(E) 2mm Pb/17 mm Scint
• Angular resolution for p0 still good
• because dominated by high energy photon whereas
  mass reconstruction is dominated by low energy
• For the physics goals we have set out, this is
  what is important
• if backgrounds to samples can be kept low

31
p0 Kinematics Example

• An example of a challenging analysis is coherent
  p0 production
• S/B separationbased on angle
• See little or noimpact of photonangular
  resolution
• sE/sqrt(E)4 for p0
• so 4-7 in the region of interest

MINERvA-like
study by Chvojka(Rochester)
32
Another Common Feature

• All of these ideas need more work
• optimization of design parameters, particularly
  thickness of radiators
• current resources required and schedule for
  construction not sufficiently evaluated

33
Plans and Next Steps
34
Decision Tree

• Is full coverage of ECAL necessary in downstream
  FGD?
• proposed motivations (multi-pion events, oxygen
  target in TPC) need further study to quantify
• Detector capabilities met for proposed designs?
• kinematics of inclusive p0, quasi-elastic e-
• backgrounds for same (fake photons, lost photons)
• hopefully studied with unified simulation!
• Cost, person-power and schedule for optimized
  designs

35
Conclusions

• Physics requires need an ECAL capable of good
  kinematic measurement of p0,e-
• Backgrounds to signatures
• inclusive NC and exclusive CC p0 production
• quasi-elastic ne
• must be kept low
• We have several designs which are capable of
  providing this
• need optimization, resource schedule evaluation
• complete by Dec. 2004 280m meeting

36
Acknowledgments

• Thanks to
• Clark McGrew, Yoshi Uchida, Katsuki Hiraide, Yuri
  Kudenko, Masato Shiozawa, Jesse Chvojka for
  slides and plots
• Dave Wark and Yoshi Uchida for helping to sharpen
  the physics arguments for the ECAL
• my 280m co-conveners Federico Sanchez, Akira
  Konaka and Tsuyoshi Nakaya