phenix high pt upgrades recent progress yasuo miake for high pt upgrade team highpt
TRANSCRIPT
PHENIX High pt Upgrades
Recent Progress
Yasuo MIAKE
For High pt Upgrade Teamhttp://utkhii.px.tsukuba.ac.jp/~highpt/
2 Yasuo MIAKE, Univ. of Tsukuba
Plan Guestimate
• Coverage of 4 TOF panel equivalent as a first stage– 150 modules
– 300 PMT’s ; 12,000,000 yen
– 400 liter; 20,000,000 yen
– Total ; ~40,000,000 yen
• Since no space on the East, we install on the West.– Remove 2 lower TOF panels from
the East after successful pp measurements
– 1 spare panel
– 1 panel to be constructed
• TOF 1 panel ; 13,000,000 yen
TOF 1 panel ; 0.5 m x 2 m
If we install on the East,
3 Yasuo MIAKE, Univ. of Tsukuba
Additional Cherenkov
• RICH– CH4;1.00044– CO2;1.00041
th= 34• Pion ; 4.7
GeV/c
• Aerogel– Best match at
1.005 – Commercially
available for 1.007 - 1.07
• Samples for test– 1.007– 1.01– 1.015
Pion
Kaon
4 Yasuo MIAKE, Univ. of Tsukuba
Extended PID
Pion-Kaon
separation
Kaon-Protonseparation
TOF ~100 ps 0 - 2.5 - 5
RICH n=1.00044
th~34
5 - 17 17 -
Aerogel n=1.007
th~8.5
1 - 5 5 - 9
0 4 8
0 4 8
0 4 8 0 4 8
0 4 8
0 4 8
Aerogel together with TOF can extend the PID capability upto 10 GeV/c(Without TOF, no K-proton separation at < 5 GeV/c)
5 Yasuo MIAKE, Univ. of Tsukuba
Aerogel Cherenkov at KEK-B BELLE
• Non-imaging Cherenkov
• 960 modules at Barrel, 228 at Endcap.
• Index n = 1.01 - 1.03
• 12 x 12 x 12 cm3
• 2 Fine-mesh PMT (1.5T)– R6683(3”) for n = 1.01
– R6682(2.5”) for n = 1.015
– R6681(2”) for n = 1.02
• ~ 10 - 20 p.e.
Journal of Non-Crys.225(1998)369
6 Yasuo MIAKE, Univ. of Tsukuba
KEK Beam Test Setup
• Particle Identification– Gas C for electron– TOF for hadron ID
• KEK pi-2 channel– 1 - 4 GeV/c for momentum scan
Gas C for electron tagging
TOF for hadron PID
KEK-PS T496 Dec. 12 - 21, 2001
7 Yasuo MIAKE, Univ. of Tsukuba
Photos (1)
• Many vistors, many students! It was fun!
8 Yasuo MIAKE, Univ. of Tsukuba
Photos (2)
M. KonnoT. Takagi
T. TakagiT. Ohki
M. Ono
9 Yasuo MIAKE, Univ. of Tsukuba
PMT selections
• PMT Requirements– Photon counting
• Gain; > 10**6• Low dark current
– Larger diameter– Cost
• Hamamatsu recommended– 2” ; R6231– 3” ; R6233
• UV PMT for the KEK test– R2059 ( 160 - 650 nm)
• Russian PMT’s ?
R329
R6231 R6233
No. Cost A few >400 >800
R6231 45,000 40,500 31,500 29,200
R6233 52,000 46,800 36,400 33,800
Nucl.Inst.Meth. 406(1998)213
10 Yasuo MIAKE, Univ. of Tsukuba
Pulse height distr.of single p.e.’s
• 2” R6231
• Clean single photoelectron peak seen.
• Gain of 2x10**6 will be obtained.– Done by
Hamamatsu
– Will be tested at Tsukuba
11 Yasuo MIAKE, Univ. of Tsukuba
Index of Aerogel
• Commercially available– Matsushita-Denko
• ~50k yen / litter• For the test experiments,
n=1.007 -8, 1.015, 1.020 have been purchased for the tests.
• Index measurements done.– Masahiro Konno
No. Index Measured
SSP-30D
1.007 - 8 1.007
+-0.002
SP-15A 1.015 1.018
+-0.002
SP-15B 1.015 1.017
+-0.003
SP-20A 1.020 1.020
+-0.001
SP-20B 1.020 1.021
+-0.001
SP-25 1.025
SP-30 1.030
12 Yasuo MIAKE, Univ. of Tsukuba
Measurement of Refractive Index
• Surface condition of the sample dominates the error.
Masah
iro K
onn
o
n=sin(φ2+
α2)
sin(φ2)
13 Yasuo MIAKE, Univ. of Tsukuba
Measurements of Transmission
• Shorter transmittance for shorter wave length.
• Shorter transmittance for smaller index.
Laser 355nm415nm532nm
355 nm415 nm
532 nm
Index 1.007 1.018 1.02
355 nm 0.6 1.0 1.3
415 nm - 2.0 4.4
532 nm 6.4 5.6 8.3
Measured Transmittance [cm]
14 Yasuo MIAKE, Univ. of Tsukuba
Calibration of PMT
• Using solid state laser pulser (415 nm), PMT’s were calibrated before and after the KEK test.
• Measurements at various HV’s give consistent results.
HV 1800V
4.2 pe
)exp(!
)()( n
n
nnP
n−=
Picosec Pulser(Hamamatsu)
415 nm
PMT
15 Yasuo MIAKE, Univ. of Tsukuba
Two Type of Designs
• Belle Type – Collect scattered photons
• Non-directional lights
• Area of photocathode/cell size
• 2 PMT per cell– more expensive
Belle Type Mirror Type
• Mirror Type– Collect direct photons
• Directional lights• Efficient way to get light!?
• Sophisticated mirror design?!• Easy to get larger cell size• 1 PMT per Cell
– Cheaper
16 Yasuo MIAKE, Univ. of Tsukuba
Aerogel Signal for protons and pions
• Clear separation of protons and pions observed.
β = 0.9989
β = 0.9544
PID by TOF successful
+3 GeV/cPions
Protons
n=1.017
17 Yasuo MIAKE, Univ. of Tsukuba
Features of Cherenkov Emission
• It is Cherenkov Emission.
⎟⎟⎟
⎠
⎞
⎜⎜
⎝
⎛
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛ +−=
2
22
2max1
1p
pm
nNN
Belle; Tyvek
m = 0.14
p = 3.0
Belle ;Goretex
m = 0.14
n = 1.017
Np.e. vs. Momentum Np.e. vs. Index
€
N photonvisible ≈ 500sin2ϑ c [1/cm]
ϑ c = Cos−1(1
nβ)
€
Nmax = 500 ×q.e.×12 [cm]
≈1200 p.e.
18 Yasuo MIAKE, Univ. of Tsukuba
Belle Type (1) Reflector
• Three types of reflector.– Tyvek
– Millipore
– Goretex
• Offline optical measurements at BNL says “Tyvek is good”
• Effect of reflector– Without reflector, < 1/4
• Reflector is essential for Belle Type
ReflectorPMT1 (p.e.)
PMT2 (p.e.)
Total (p.e.)
Goretex 12 13 25
Millipole 9 10 19
Tyvek 8 8 16
Black Paper
2 3 5
3 GeV/c pions
n = 1.017R6233(3”,non-UV)
19 Yasuo MIAKE, Univ. of Tsukuba
Belle Type (2) Position Dep.
• With Goretex, >25 p.e. obtained everywhere.
• In each PMT, exponential behaviour is observed.– Goretex ; λ= 7.0 cm
– Tyvek ; λ= 5.5 cm
• Exponential shape may not be trivial issue?
3 GeV/c pions
n = 1.017R6233(3”,non-UV)
Goretex
€
e−x−xPMTλ
x
20 Yasuo MIAKE, Univ. of Tsukuba
Belle Type (3) Angular Dep.
• Seems to be proportional to the thickness of the aerogel.– Scattered photons have no
directionality.
3 GeV/c pions
n = 1.017R6233(3”,non-UV)
θ
21 Yasuo MIAKE, Univ. of Tsukuba
Belle Type (4) Area of Photocathode
• Put Iris-plate in front of the photocathode.
• Roughly proportional to the area of photocathode, S– Scattered photons have no directionality.
• Then, saturate with larger PMT coverage
3 GeV/c pions
n = 1.017R6233(3”,non-UV)
IrisS
22 Yasuo MIAKE, Univ. of Tsukuba
Belle Type (5) Thickness Dep.
• Photons proportional to the thickness of the aerogel.– Consistent with;
• Angular dependence
• No directionality
3 GeV/c pions
n = 1.017
R6233(3”,non-UV)
L
23 Yasuo MIAKE, Univ. of Tsukuba
Larger Cell of Belle Type
• Using exponential behaviour, we can estimate the performance for larger cell of Belle Type.– Smallest at the center
• We can estimate performance of Belle Type for any size, any thickness.
3 GeV/c pions
n = 1.017
R6233(3”,non-UV)
D
D
Npe at
theCenter
24 Yasuo MIAKE, Univ. of Tsukuba
Mirror Type (1) Shape of Mirror
• Flat vs Parabola Mirror– Aluminized mylar sheet with
styro foam backing (hand made)
• No significant difference at the center as expected from the cone angle of 10 degree.
Mirror Shape Npe
Flat 9
Parabora 9
Mirror Type
3 GeV/c pions
n = 1.017R6233
(3”,non-UV)
25 Yasuo MIAKE, Univ. of Tsukuba
Mirror Type (2) Position Dep.
• Expected Position Dep.– Relation of Cone angle ~ 10.1
deg. and PMT size
– Expected diameter ~ 3” PMT
• Broad peak at the center
• No significant difference between Flat and Parabola mirrors.
3 GeV/c pions
n = 1.017R6233
(3”,non-UV)
26 Yasuo MIAKE, Univ. of Tsukuba
Mirror Type (3) Thickness Dep.
• Saturate!– Difficult to get more p.e.
– Due to short transmittance
3 GeV/c pions
n = 1.017R6233
(3”,non-UV)
L
Mirror Type
27 Yasuo MIAKE, Univ. of Tsukuba
Bell vs Mirror
• Even with Mirror type, reflector is important !
Mirror Shape Npe
Flat (Tyvek) 9
Parabora (Tyvek) 9
Parabora (Goretex) 14
3 GeV/c pions
n = 1.017R6233
(3”,non-UV)
ReflectorPMT1 (p.e.)
PMT2 (p.e.)
Total (p.e.)
Goretex 12 13 25
Millipole 9 10 19
Tyvek 8 8 16
Black Paper
2 3 5
3 GeV/c pions
n = 1.017R6233(3”,non-UV)
28 Yasuo MIAKE, Univ. of Tsukuba
Directional vs Scattered Photons
• Is the worse results with mirror type due to bad mirrors?– Note that it is hand made with
aluminized mylar sheet.
• Other way to measure Directional/Scattered photons.
• Rotate Belle Box and put the beam straight to aerogel and PMT.
• To remove signals from PMT window, subtract (b) from (a).
PMT
Aerogel Empty
Beam
(a) (b)
(a)
(b)
29 Yasuo MIAKE, Univ. of Tsukuba
Measurements with 2” PMT
• Assume uniform angular distribution of scattered photons.
• Out of 17 p.e. (straight beam geometry), 8 p.e.’s are scattered photons.
17 p.e.
8 p.e.
6 p.e.
n=1.0172” PMT3 GeV/cTyvek
€
BackSide = exp(−λz)
0
10
∫ dz exp(−5λ )0
10
∫ dz
≅1.2 with λ =1/5 [1/cm]
30 Yasuo MIAKE, Univ. of Tsukuba
Measurements with 3” PMT
• Guestimate of scattered components in straight geometry– 8 p.e. x *(8 p.e./6 p.e.) = 11 p.e.
• Thus, direct photons in straight geometry are 11 p.e., which is similar to direct photons observed with 2” PMT.– Consistent with cone angle of 10 degree.
22 p.e.
8 p.e.
n=1.0173” PMT3 GeV/cTyvek
Not mea
sured
!
31 Yasuo MIAKE, Univ. of Tsukuba
Use of UV PMT
• Assuming cherenkov from uniform radiator, use of UV pmt will gain a factor of 2.
• Increase of 25 % observed with the straight beam measurements.– Not uniform at the level of
200 nm?
Non-UV (R6231)160 - 650 nm
17 p.e.
UV (R2059)
300 - 650 nm21 p.e.
Tyvek, n=1.015
32 Yasuo MIAKE, Univ. of Tsukuba
Use of wavelength shifter
• Wavelengthshifter may increase the overall photon yield.
• Thanks to Prof. Peressedov, we could try POPOP at KEK.– Powder on aerogel surface
• Larger tail appeared with POPOP for pions!– Note POPOP localized only
on the surface of Aerogel.
!!
33 Yasuo MIAKE, Univ. of Tsukuba
Use of POPOP powder
• Additional tail also appeared for protons.– Scintillation ?
• But, it is smaller tail than for those of pions.
• Need to investigate a way to uniformly distribute POPOP over the large volume not only the surface of Aerogel.
34 Yasuo MIAKE, Univ. of Tsukuba
Novosibirsk vs Matsushita
With Non-UV PMT (R6231) With UV PMT (R2059)n=1.0072” PMT
35 Yasuo MIAKE, Univ. of Tsukuba
Summary of KEK test
• Belle Type– It works; > 20 p.e.
– We can estimate performance of Belle Type with any cell size, any thickness.
• We understand pretty well.
• Mirror Type– ~ 10 p.e.
– Direct photons at most 10 p.e.
• Is there a better way to combine direct and scattered photons? Otherwise, Belle Type seems to be the best.
• Needs to be studied– Better mirror? (I’m pessimistic)
– UV vs. non-UV PMT
– PMT selections
• Gain too low?
• Russian PMT?
– Test of electronics
• Use of PMT amp?
36 Yasuo MIAKE, Univ. of Tsukuba
GEANT Calculation
• Belle Type Aerogel counter installed in PISA.
T. Takagi
PHENIX
East ArmWest Arm
37 Yasuo MIAKE, Univ. of Tsukuba
GEANT results
• Occupancy of 5 % seen in central Au+Au collisions (HIJING)
Backgrounds
z (c
m)
x (cm)
magnet
RICH mirrorAerogel
beam line
Number of fired cells per event
38 Yasuo MIAKE, Univ. of Tsukuba
Optical Model
39 Yasuo MIAKE, Univ. of Tsukuba
Summary of Status
(1) In December 2001, test experiment for the Aerogel Cherenkov have been carried out at KEK with visitors from Dubna.
(2) Both Belle type (aerogel w. 2 PMT's on the side) and Mirror type (mirror followed by 1 PMT behind the aerogel) have been tested. So far, best results (>20 p.e.) obtained from Belle type with 3" PMT's and Goetex as reflector, which is our backup solution now.
(3) We will continue R&D and test experiments for both mirror and Belle type. (Next test beam will be before the summer.)
(4) For the deltailed analysis of optical properties of the Aerogel, such as wavelength dependences of absorption length & scattering length, study with a laser and spectrometer is also in progress.
(5) Comparison of Aerogel from Matsushita and Novosibirsk gives interesting results; while similar results obtained when non-UV PMT was used, Novosibirsk Aerogel provides 20 - 30 % larger signal than that of Matsushita with UV PMT (Quartz window). As is reported in literature, Novosibirsk aerogel seems to have better optical transmission in UV. (According to Sumiyoshi at KEK, master of aerogel, more complicated process is adopted for production of aerogel at Novosibirsk.)