Development of an Aerogel-based Photon Detector

1
Development of an Aerogel-based Photon Detector

• T. Nomura (Kyoto Univ.)

2
Motivation, Challenge, and Solution
A part of veto detector in rare KL experiment
Located in intense neutral beam
Aerogel-based photon detector
3
Motivation

• Necessary in KLgp0nn measurement
• Importance of KLgp0nn
• CP violating process
• Branching ratio proportional to Im(Vtd)2,
  BR10-11
• Very small theoretical uncertainty (1)
• Play an important role to explore BSM (like SUSY)
  
• Difficulty of KLgp0nn experiment
• All neutrals in initial and final states
• Event signature 2g (from p0) nothing (2n)
• We have to prove nothing in order to suppress
  backgrounds
• KL?p0 p0p0 (BR 21), p0 pp- (BR 13), p0 p0 (BR
  10-3),

4
Motivation
To prove nothing, we needs Hermetic veto
system
Need to catch photons escaping through
the beam-hole
What we want to develop !!
5
Challenge

• In-beam environment
• High intensity neutral beam (necessary to
  observe gt100 KL?pnn events in 2-3 years)
• A vast amount of neutrons (a few10GHz)
• Produce protons, pions, (g and e/e-) in the
  detector
• Most KLs survive after decay region (100MHz)
• Decay into p, g ,e/e- in the detector
• ? These secondary particles fire the counter
  and disturb its primary function !!

6
Solution

• Utilize Cherenkov radiation Aerogel (low
  refractive index 1.05) radiator
• Avoid detection of slow particles from neutron
  interactions Slow p, p and other hadrons cannot
  emit lights.
• Use direction information Segment the detector
  into many modules and require coincidence
  along the beam direction
• Catch forward photons only Reduce fake signal
  due to g from secondary p0 (neutron interaction
  and KL decay in the detector)

7
Design of In-Beam Photon Detector

• Module
• Pb (g converter) Aerogel (Cherenkov
  radiator)
• Light collection with Flat mirror Winston
  cone
• Sparse sandwich detector
• modules array

8
Proof of Principle (I)

• Weve made three generations of prototypes
• Prototype 1 (2001-2)
• Simple structure
• 11cm x 11cm tiles
• Flat mirror
• Read by 5-inch PMT
• Exercise to use aerogel detector
• light yield

9
Proof of Principle (II)

• Prototype 2 (2002-3)
• Sophisticated optics
• 11cm x 11cm tiles
• 2-axis parabolic mirror
• Read by 5-inch PMT
• light yield
• response to proton (as substitute for neutron)

10
What we learned from Prototype 2

• Response to Proton(analogous to neutrons)
• Single module efficiency
• Find N2 gas scintillation(problematic in 1p.e.
  region)
• Two layers coincidence
• Good agreement with MC

11
Proof of Principle (III)

• Prototype 3 (2004-5)
• Base design
• 30cm x 30cm area(3x3 of 10cm sq. aerogel tiles)
• Flat mirror
• Winston cone
• Read by 5-inch PMT
• light yield
• position / angular dependence

12
Elements of Prototype 3

• Elements
• 3x3 10cm sq tiles, stacked 5 layers
• Winston cone
• made by thin Al
• AlSiO2 evaporated on inner surface

45 Tiles made by Matsushita
Made by Yokohama-Kiko
13
What we learned from Prototype 3 (i)

• Position dependence
• Global structure, reproduced well by MC
• Edge effect between tiles
• surface deterioration by trimming process
  (water jet)

By Winston cone entrance (x6cm)
By tiles edge effect (x5cm)
14
What we learned from Prototype 3 (ii)

• Angular dependence
• Global structure, reproduced well by MC
• Winston cone deformed
• stressed by joint or support

By cones acceptance (q7 deg)
Reflection anglemeasured by laser
By cone deformation (q5 deg)
15
Practice in KEK E391a experiment

• E391a-III had run
• with this Prototype 3
• called APC (Aerogel Photon Catcher)
• Used as in-beam g tagger

BA (Beam Anti) E391a in-beam detector PWO
Quartz sandwich
16
Application
To KOPIO experiment at BNL
To KLgp0nn experiment at JPARC
17
KOPIO experiment
Terminated (2005 August)

• Planned KLgp0nn experiment at BNL
• (Construction 2006-, Run 2010-)
• High intensity proton beam
• 100 TP/spill
• Soft KL beam
• 0.5-1.0 GeV/c
• Horizontally wide beam
• 4mrad x 90mrad
• Measure g direction as well as energy
• Sensitivity 40 SM events (S/N2), or 200 SM
  events (S/N0.3)

18
KOPIO detector Beam Catcher
19
KOPIO In-beam Photon Detector

• In-beam Aerogel Detector
• Module size 30cm x 30cm
• Module array
• Number of modules 420
• 12-21 in horizontal with beam divergence
• 25 layers along beam(8.3 X0 in total)
• Z gap between layers 35cm

20
Expected Performance by MC (1)
In KOPIO case

• Photon efficiency
• Soft KL in KOPIO
• Relatively low energy g
• Relatively small shower
• Low threshold

Coincidence condition 4 p.e in A, 2 p.e. in B
21
Expected Performance by MC (2)
In KOPIO case
Hit probability for Neutrons
Hit probability for KLs
Dominated by decays in the detector
0.3 _at_ 800MeV
22
Application to Experiment at J-Parc

• KLgp0nn experiment plan at J-Parc
• 30GeV proton, 100 TP/spill
• Small production angle, relatively hard KL beam
• Pencil beam
• ( a few 10 ) mstr, 10cmf at the detector
• Step 1 with (modified) KEK E391a detector
• 10 SM events, Discovery phase
• Step 2 with new, optimized detector
• Precision measurement, 100 SM events

23
In-beam Detector for J-Parc experiment

• In contrast with KOPIO case
• Pencil beam
• A series of modules along beam direction
• Relatively high energy g (neutrons)
• Detection threshold can be (has to be) higher
• 3 consecutive hits, 424 p.e.

24
Expected Performance
In J-Parc case

• Neutron Hit probability
• Level of 0.1 _at_ 4.0GeV/c

• Photon efficiency
• 90 efficiency _at_1GeV
• 99 efficiency _at_2GeV
• At high energy limit, inefficiency O(10-3)

25
Expected Performance
In J-Parc case
With various threshold
26
Summary

• Weve developed Aerogel-based Photon
  Detector to use in intense neutral beam
• One of the key detectors in KLgp0nn experiment
  to explore physics beyond the SM
• New concept
• Pb (converter) Aerogel (Cherenkov radiator)
• Sparse sandwich detector
• Proof-of-Principle done with 3 generations of
  prototypes
• Originally, it was designed for KOPIO
  experiment.Now, we are considering to use this
  detector at J-PARC

27
Signal Loss due to False Hit
In KOPIO case

• Accidental hit due to neutrons may kill K?pnn
  signal

• Total false hit probability was found 0.4
  events / m-bunch
• Integrated over the duration consistent with
  the arrival time of g from our signal KL
• If we set the time window to be 3ns,signal loss
  due to false hit will be 4.6
• Calculation based on random effect
• Detailed studies by MC under way

Apply timing cut
28
False Hit Rate due to neutron (Step1)
In J-Parc case

• Accidental hit due to neutrons (In case
  Step 1)
• Integrate over neutron momentum
• Enough low rate (1MHz) even with lowest thres.

29
False Hit Rate due to neutron (Step2)
In J-Parc case

• Accidental hit due to neutrons (In case
  Step 2)
• Very high rate (10MHz) even with highest
  thres.

30
Signal / Background (Step2)
In J-Parc case

• To gain background rejection power
• Prefer lower threshold
• To prevent acceptance loss due to false
  veto
• Prefer high threshold