Current status of the thick-GEM TPC for the J-PARC E15 experiment - - PowerPoint PPT Presentation

Current status of the thick-GEM TPC for the J-PARC E15 experiment

Fuminori Sakuma (RIKEN)

1

Tum-Riken Kick-Off Meeting @ TUM, May 10-11, 2010.

Contents

 Introduction  TGEM-TPC for the J-PARC E15 exp.  Thick GEM (TGEM)

goal: gain〜104 with stable operation in P10 @ NTP

 Results of TGEMs in P10 @ NTP

• Cu-electrode TGEMs
• C-electrode TGEMs
• C/Cu-electrode Hybrid TGEMs

 Summary

2

3

J-PARC E15 Experiment

search for K-pp bound state using 3He(K-,n) reaction K-

3He

Formation

exclusive measurement by Missing mass spectroscopy

and

I nvariant mass reconstruction

Decay

K-pp cluster

neutron

Λ

p p

π-

Mode to decay charged particles

at at J-PAR ARC

4

J-PARC E15 Setup

1GeV/c K- beam

p π− p n

mass resolution for K-pp

invariant mass σ = 19MeV/c2 (σCDC = 250µm) missing mass (for 1.3GeV/c neutron) σ = 9.2MeV/c2 (σToF = 150ps)

Neutron ToF Wall Cylindrical Detector System Beam Sweeping Magnet

K1.8BR Beam Line

Beam trajectory CDS & target Sweeping Magnet Neutron Counter Beam Line Spectrometer

important to measure not only non-mesonic decay mode but also mesonic decay mode

Thick-GEM TPC improves z-resolution

measurement of mesonic decay-mode of K-pp

mesonic mode is suppressed!

6

TGEM-TPC for the J-PARC E15 exp.

TGEM-TPC for the J-PARC E15 exp.

7

TGEM-TPC is located at the center of Cylindrical Detector System

• located between CDC and target-chamber
• cover the CDC acceptance of AUVA
• minimum materials in the acceptance
• 1mm spatial resolution in the z-direction

TGEM TPC

〜2m

Cylindrical Detector System TGEM-TPC TGEM

filled with P10 gas at atmospheric pressure

Gas connector HV connector

R/O pad size 4mm×20mm # of pad = 4×4×9 = 144 field strip

• double sided
• FPC
• 8mm strip
• 10mm pitch

8mm 10mm 2 cm 28 cm R/O

TGEM

completed TGEM-TPC

4mm

non-necessity of support-structure!

9

making of Field Cages

Large FPC board sticking support frames on the FPC soldering resisters(1MΩ) rolling up the FPC Inner and Outer field cages uniting the two cages completed

10

HV & Readout

HV connector (LEMO, max. 15kV) preamp attachment (test) readout pad readout with TGEM TGEM installation close up view of double-TGEM to reduce detector capacitance,

• ne side of TGEM is

divided into 3 parts

11

Readout Electronics is the same as that of CDC

preamp cards and cables are attached TDC’s in the counting room LVDSECL converters at the exp. hall

• Chip : CXA3183Q

(SONY , low noize ASD IC, τ=16nsec)

• Output : LVDS differential
• Gain : 0.8V/pC at preamp
• 4x4=16ch

8m cables 60m cables

LVDS ECL

τ=80nsec

We measure only time

• info. with the TPC!

12

gas

limit of HV module : 15kV GEM HV : 4kV drift length : 30cm maximum drift-field voltage : ~350V/cm

We choose P10 (Ar/CH4=90/10) for the TGEM-TPC gas

2 2 2 d x eff

C z N σ σ ⋅ = +

σx : total resolution σ0 : resolution w/o diffusion Cd : diffusion constant z : drift distance Neff : effective number of electrons

expected resolution

E = 150V/cm  Cdl = 0.34mm, Cdt = 0.60mm σ0l = 0.5mm σ0t = 0.2mm Neff = 38.7*0.4(cm) = 15.5

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 5 10 15 20 25 30 35

resolution(mm) z(cm)

longitudinal transverse

z Dl Dt σl σt (c (cm) (mm) mm) (mm) mm) (mm) mm) (mm) mm) 0.00 0.00 0.50 0.20 10 1.09 1.89 0.57 0.52 20 1.54 2.67 0.63 0.71 30 1.88 3.27 0.69 0.85

longitudinal transverse

φ-direction resolution is limited by pad size, e.g., 20.0/sqrt(12) = 5.8mm So we use only z-direction info.

13

Thick GEM (TGEM)

goal: gain〜104 with stable operation

What is TGEM ?

TGEM@RIKEN

Drilled hole 300µm Rim 100µm

Garfield Simulation avalanche thickness:400µm ∆VGEM∼1kV

avalanche

TGEM cross-section and drift line

14

Thick-GEM is …

cost-effectively fabricated from double-clad G10 plates, using standard printed circuit board (PCB) techniques

holes are mechanically drilled (and, if necessary, the hole’s rim is chemically etched to prevent discharges) a robust, simple to manufacture, high-gain gaseous electron multiplier easy to operate and feasible to cover large areas, compared to the standard foil GEM

15

 gain〜104 @ P10, NTP (double TGEMs)  stabile operation for a month, with gain fluctuation within ~a few ten % for a month & a few % for a day many groups have reported TGEMs work successfully, but actually it’s NOT so easy to operate TGEM with high gain stably!

• they use small TGEMs, e.g. ~3x3cm2
• most of them don’t discuss stability of TGEM

TGEM prototypes

We have studied basic TGEM behavior and performance.

goal

TGEM prototypes @ RIKEN

produced by REPIC corp. and TOUKAI DENSHI KOUGYOU corp.

16

No. No. Electrod rode Ins nsul ulator Thickn kness[ ss[µm] Hol Hole-diame meter[ r[µm] m] Rim[µm] 1 Cu Cu FR4/U 4/UV 200 200 300 300 50 50 ×2 2 Cu Cu FR4/U 4/UV 200 200 500 500

• ×2

3 Cu Cu FR4/U 4/UV 400 400 300 300

• ×5

4 Cu Cu FR4/U 4/UV 400 400 300 300 30 30 ×2 5 Cu Cu FR4/U 4/UV 400 400 300 300 50 50 ×2 6 Cu Cu FR FR4 400 400 300 300 100 100 ×2 7 Cu Cu FR4/U 4/UV 400 400 500 500

• ×2

8 C FR FR4 400 400 300 300

• ×4

9 C FR4/U 4/UV 400 400 300 300

• ×7

10 10 C G10 400 400 300 300

• ×2

11 11 C CEM EM3 400 400 300 300

• ×2

12 12 C FR FR4 600 600 300 300

• ×2

13 13 C/ C/Cu Cu FR4/U 4/UV 400 400 300 300

• ×4

Tota tal 40 size : 10cm x 10cm

17

many TGEM prototypes

readout pad

test chamber Ga Gas : : P10 a 10 at 1a 1atm, no normal t tem emper eratur ure

DC 1M 2200p 1M 2M 1M 2M 1M GEM1 GEM1 GEM2 GEM2 mesh 20M 20M 20M 20M 11mm 2mm 400um 2mm 400um R/O 20M 72Hz low-pass

HV V div ivid ider w wit ith h res esis istive cha chain in

18

Test bench setup

doub uble TGE TGEMs

 Ratio of ∆VGEM/Etrns/Eind is const. 11mm 2mm 2mm drift mesh TGEM 1 TGEM 2 R/O pad CS preamp

55Fe

Double GEM setup X-ray e- ΔVGEM ΔVGEM Etrns Eind Edrift

(150V/cm)

19

Results of TGEMs

Cu-electrode TGEM

TGEMs with thickness of 400μm and hole diameter

• f 300μm achieve maximum gain of 104

20

No. No. Electrod rode Ins nsul ulator Thickn kness[ ss[µm] Hol Hole-diame meter[ r[µm] m] Rim[µm] Max g gain 1 Cu Cu FR4/U 4/UV 200 200 300 300 50 50 〜10 103 2 Cu Cu FR4/U 4/UV 200 200 500 500

• 3

Cu Cu FR4/U 4/UV 400 400 300 300

• 〜10

104 4 Cu Cu FR4/U 4/UV 400 400 300 300 30 30

• v
• ver

r 2×10 104 5 Cu Cu FR4/U 4/UV 400 400 300 300 50 50

• v
• ver

r 2×10 104 6 Cu Cu FR FR4 400 400 300 300 100 100

• v
• ver

r 2×10 104 7 Cu Cu FR4/U 4/UV 400 400 500 500

• 〜10

103 Rim of 50,100μm : Weizmann method (drilling + masked etching) Rim of 30μm : CERN method (drilling + resist etching) w/o Rim (#3) : w/ hydrogen peroxide - sulfuric acid etching

1.0E+00 1.0E+01 1.0E+02 1.0E+03 1.0E+04 1.0E+05 800 900 1000 1100 1200

effective gain

ΔVGEM[V]

Cu Rim 100μm Cu Rim 50μm Cu Rim 30μm Cu no Rim

goal the limits of gain around 105 is caused by reather limit (source = 55Fe).

Edrift=150V/cm ∆VGEM (V) : Etrans (V/cm) : Einduct (V/cm) 1 : 2.5 : 7.5

dependence on rim size

21

TGEM with larger rims requires higher voltage, but enables higher gain

gain 〜2.5×104

initial drop of gain is caused by charge-up (polarization?) of the insulator? instability of TGEMs with rims is caused by charge-up of the insulator not metalized. mismatch of the center of the etched and drilled holes and incomplete round-shape of rims cause the instability.

gain and resolution stability (24h)

22

energy resolution@5.9keV(σ)

gain

TGEMs with rims ( , , ) are NOT so stable TGEM without rims ( ) is stable

100µm 50µm 30µm No Rim 100µm 50µm 30µm No Rim corrected relative gain resolution(σ) [%]

23

TGEM with 30µm rims can be operated with gain of more than 104 for the long term @ P10, NTP gain stability is within ~50%/week & ~10%/day

long term stability (30µm rims TGEM, 10days)

relative gain

gain=2.5×104

raw data

2 1 0 10days resolution(σ) [%] 40 20 0 10days relative gain

gain~2.5×104

P/T corrected data

2 1 0 10days ∆VGEM is turned up by hand gain=1.0×104

24

C-electrode TGEM

No. No. Electrod rode Ins nsul ulator Thickn kness[ ss[µm] Hol Hole-diame meter[ r[µm] m] Rim[µm] Max g gain 8 C FR FR4 400 400 300 300

• v
• ver

r 2×10 104 9 C FR4/U 4/UV 400 400 300 300

• 10

10 C G10 400 400 300 300

• 〜10

103 11 11 C CEM EM3 400 400 300 300

• v
• ver

r 2×10 104 12 12 C FR FR4 600 600 300 300

• 〜10

102

To avoid the effects of rims, we are developing a new resistive-electrode TGEM (RETGEM ) which has electrodes coated with graphite paint. RETGEMs have an advantage of being fully spark-protected.

Results of the first sample

1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 750 800 850 900 950 1000 gain ΔVGEM[V]

Carbon

ΔVGEM×7.5

gain〜2.5×104

25

gain

energy resolution@5.9keV(σ)

It seemed that C-electrode TGEMs

work excellent!!! However...

Edrift=150V/cm ∆VGEM (V) : Etrans (V/cm) : Einduct (V/cm) 1 : 2.5 : 7.5

goal

• nly first 2 out of 11 samples of RETGEMs work !!!

discharge from burrs arising from drilling process (but these can be removed using antistatic-brush) carbon attachment inside the holes caused by knot of FR4 fiber

26

now, we have been studying another insulator of CEM3 not FR4/G10

carbon TGEMs have no reproducibility at all !!!

cross section of FR4 (400µm) knot of fiber

• nly 2

RETGEMs work burrs

C/Cu-electrode Hybrid-TGEM

In principle, if one side of electrode is resistive then that would be spark-protected.

Cu Cu+Cu electrode Cu C+C electrode C+Cu electrode C C C Cu

spark-protected spark-protected

Hybrid-TGEM would have a possibility of reduction of carbon attachment inside the holes.

27

C+C electrode C+Cu electrode No. No. Electrod rode Ins nsul ulator Thickn kness[ ss[µm] Hol Hole-diame meter[ r[µm] m] drill ill 13 13 C/ C/Cu Cu FR4/U 4/UV 400 400 300 300 Cu→C ×2 C/ C/Cu Cu FR4/U 4/UV 400 400 300 300 C→Cu ×2

28

1.0E+00 1.0E+01 1.0E+02 1.0E+03 1.0E+04 1.0E+05 800 850 900 950 1000 effective gain ΔVGEM[V]

Results of C/Cu-electrode Hybrid TGEM

gain〜2.5×104

The 2 fabrication methods work similarly corrected gain

energy resolution@5.9keV(σ)

We tried 2 drilling directions, i.e. CuC and CCu

Cu C C Cu Edrift=150V/cm

∆VGEM (V) : Etrans (V/cm) : Einduct (V/cm) 1 : 2.5 : 7.5

29

0 25days

hybrid TGEM can be operated with more than gain

• f 104 for the long term @ P10, NTP

gain stability is within ~20%/week & ~5%/day

40 20 resolution(σ) [%]

raw data

blanks in the plots are

• ut of 55Fe source,

but keep to turn on HV gain=2.5×104 gain=1.0×104 relative gain

P/T corrected data

0 25days ∆VGEM is turned up by hand 2 1

long term stability (hybrid TGEM, 25days)

30

• We have been developing a TGEM-TPC for the J-PARC

E15 upgrade

• TGEM-TPC was completed, and commissioning

will be started soon

• Cu electrode TGEM with 30µm rims can be operated

with gain of more than 104 for the long term rather stably @ P10, NTP

• C-electrode TGEM is far from goal...
• C/Cu electrode TGEM can be operated with gain of

more than 104 for the long term stably @ P10, NTP

Summary

31

32

C electrode TGEM with CEM3 insulator

cross section of CEM3

A cross section of CEM3 is very clean compared with that of FR4.

cross section of FR4 knot of FR4 fibers

reduction of carbon attachment inside the holes

Now we are investigating reproducibility of CEM3 RETGEM

A disadvantage of CEM3 RETGEM is its flexibility

1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 800 850 900 950 1000

effective gain ΔVGEM[V]

gain resolution gain〜2.5×104

33

How to make Rims

There are 2 ways

Weizmann method CERN method CERN method Weizmann method masking pattern drilling etching masking off drilling etching masking off (failure) drilling + masked etching Advantage: large rims can be made easily Disadvantage: difficult to center etched and drilled holes drilling + resist etching Advantage: center of etched and drilled holes are the same Disadvantage: difficult to make large rims

It’s known that large rims cause instability of TGEMS, although those enable TGEMs to reach high gain.

Weizmann CERN resist film

drift mesh TGEM 1 TGEM 2 R/O pad ED=150 V/cm

1.0E+02 1.0E+03 1.0E+04 1.0E+05 1000 2000 3000 4000 5000 6000

effective gain ET [ V/cm]

ΔVGEM=900V ΔVGEM=940 ΔVGEM=980V ΔVGEM=1020V ΔVGEM=1060V

ET

1.E+02 1.E+03 1.E+04 1.E+05 2000 4000 6000 8000 10000

effective gain EI [ V/cm]

ΔVGEM=900V ΔVGEM=940V ΔVGEM=980V ΔVGEM=1020V ΔVGEM=1060V

EI

55Fe

1.0E+02 1.0E+03 1.0E+04 1.0E+05 850 900 950 1000 1050 1100

effective gain ΔVGEM [V]

ΔVGEM ΔVGEM

ET, EI dependence (Rim 30µm TGEM)

プリアンプアウト

preamp

35

raw signal

in consideration of TPC

• peration,

effective gain ~ 104 long time stability of gain and resolution

goal for of TGEM study gain 〜2.0×104

Double TGEM #4 P10, 1atm ∆VGEM = 983V

55Fe X-ray

signals and goal of the studies

preamp out ADC spectrum 1mV 100ns 200mV

correction raw gain corrected relative gain

36

P/T correction of gain

correlation P/T correction function P/T

・ ethanol cleaning → ×

・does not improve at all

・ plasma etching → △

・improves a little bit, but it’s not perfect ・does not remove burrs of carbon

・ removing burrs with resist-film and/or antistatic-brush → △

・removes burrs, but does not improve

・ steam cleaning → ×

・does not improve at all

・ polyimide etching → ?

・effects are depend on material of the insulator ・and also depend on etching time

・ change insulator (FR4/UV→CEM3) → ?

・CEM3 TGEMs work good, but that are after polyimide etching (we did not check the without the etching)

37

after-treatment of C-electrode TGEM

We tried many items to make the C-electrode TGEMs work We have to study more

38

Outlook

• Nonagonal TGEMs (w/o rims and Cu-rim-30μm) and

Hybrid-TGEMs for the TPC were produced, and studies have started now.

• Development of Carbon TGEMs will be continued in

the year 2010.

Physics Motivation

39

SPS, RHIC, LHC KEK-PS

W.Weise NPA553, 59 (1993).

E549@KEK-PS FINUDA@DAΦNE DISTO@SATUREN OBELIX@CERN-LEAR

T.Yamazaki, A.Dote, Y .Akiaishi PLB587,167(2004).

deeply-bound kaonic nuclear states exist?

We need conclusive evidence!