Neutron Imaging Detector based on the PIC Joe Parker Cosmic Ray - - PowerPoint PPT Presentation

Neutron Imaging Detector based on the µPIC

Joe Parker Cosmic Ray Group, Kyoto University

MPGD2011, 30 Aug 2011

Neutron Imaging Detector based on the µPIC

Prototype system and basic operation. Demonstration measurements. Future improvements.

KYOTO UNIVERSITY, COSMIC RAY GROUP

J.D. Parker, K. Hattori, S. Iwaki, S. Kabuki, Y. Kishimoto, H. Kubo,

• S. Kurosawa, K. Miuchi, H. Nishimura, T. Sawano, T. Tanimori, K. Ueno

JAPAN ATOMIC ENERGY AGENCY, MATERIALS AND LIFE SCIENCE FACILITY DIVISION

• M. Harada, T. Oku, T. Shinohara, J. Suzuki

MPGD2011, 30 Aug 2011

Neutron imaging detector prototype (µNID)

Ar-C2H6-He3 (up to 2 atm total pressure). Gas gain < 1000 for neutron imaging. TPC measures 3D proton-triton tracks. Compact, high-rate FPGA-based DAQ. Energy deposition estimated by time- above-threshold method. Efficiency up to ~30%, position res. of ~120 µm, time res. of ~1 µs.

10 cm 400 µm

DRIFT CAGE µPIC ALUMINUM VESSEL

9.0 cm 3 2 . 8 c m 2.5 cm ALUMINUM DRIFT PLANE (0.3 mm)

Prototype with top removed. 10-cm µPIC

• mfr. by DNP

X (strips)

10 20 30 40 50 60

Relative clock pulse

5 10 15 20 25 30 35 40 45 50

Position

X (strips)

10 20 30 40 50 60

Time-above-threshold (clocks)

5 10 15 20 25 30

Energy Deposition

DAQ and FPGA logic

µPIC

FPGA encoder 32 bits:

• rientation,

time, position, edge VME memory PC

Digital out (256 ch ! 2) VME bus 33-bit LVDS (! 2) Amplifier-Shaper-Discriminators External gate

Threshold Threshold

1

µPIC µPIC ASD ASD

DATA ENCODING

Two words per pulse. ‘edge bit’ saved with each data word.

Time-above-threshold (∝ energy deposit)

Simultaneous measurement

• f position and ‘energy

deposition’ at high rates.

(ATLAS, KEK)

PROTON-TRITON TRACKS

Y (strips)

10 20 30 40 50 60

Relative clock pulse

5 10 15 20 25 30 35 40 45 50

Position Y (strips)

10 20 30 40 50 60

Time-above-threshold (clocks)

5 10 15 20 25 30

Energy deposition

DAQ and FPGA logic

µPIC

FPGA encoder 32 bits:

• rientation,

time, position, edge VME memory PC

Digital out (256 ch ! 2) VME bus 33-bit LVDS (! 2) Amplifier-Shaper-Discriminators External gate

Threshold Threshold

1

µPIC µPIC ASD ASD

DATA ENCODING

Two words per pulse. ‘edge bit’ saved with each data word.

Time-above-threshold (∝ energy deposit)

Simultaneous measurement

• f position and ‘energy

deposition’ at high rates.

(ATLAS, KEK)

PROTON-TRITON TRACKS

DAQ and FPGA logic

µPIC

FPGA encoder 32 bits:

• rientation,

time, position, edge VME memory PC

Digital out (256 ch ! 2) VME bus 33-bit LVDS (! 2) Amplifier-Shaper-Discriminators External gate

Threshold Threshold

1

µPIC µPIC ASD ASD

DATA ENCODING

Two words per pulse. ‘edge bit’ saved with each data word.

Time-above-threshold (∝ energy deposit)

X (strips)

10 20 30 40 50 60

Relative clock pulse

5 10 15 20 25 30 35 40 45 50

Position

X (strips)

10 20 30 40 50 60

Time-above-threshold (clocks)

5 10 15 20 25 30

Energy Deposition

Simultaneous measurement

• f position and ‘energy

deposition’ at high rates.

(ATLAS, KEK)

PROTON-TRITON TRACKS

DAQ and FPGA logic

µPIC

FPGA encoder 32 bits:

• rientation,

time, position, edge VME memory PC

Digital out (256 ch ! 2) VME bus 33-bit LVDS (! 2) Amplifier-Shaper-Discriminators External gate

Threshold Threshold

1

µPIC µPIC ASD ASD

DATA ENCODING

Two words per pulse. ‘edge bit’ saved with each data word.

Time-above-threshold (∝ energy deposit)

X (strips)

10 20 30 40 50 60

Relative clock pulse

5 10 15 20 25 30 35 40 45 50

Position

X (strips)

10 20 30 40 50 60

Time-above-threshold (clocks)

5 10 15 20 25 30

Energy Deposition

Simultaneous measurement

• f position and ‘energy

deposition’ at high rates.

PROTON TRITON NEUTRON

(ATLAS, KEK)

PROTON-TRITON TRACKS

Materials and Life Science Facility (MLF)

Test experiments at J-PARC

J-PARC

Tokai, Ibaraki

Experiments in Nov. 2009, June 2010, and Feb. 2011. Beam power ~120 kW. Carried out at NOBORU beam line. Fill gas: Ar-C2H6-3He (63:7:30) at 2 atm, efficiencies ~28%(5 cm), ~13%(2.5 cm).

Moderator-to-detector distance of ~14.5 m.

• Max. beam size: 10 × 10 cm2.

25 Hz pulse rate, 10 Å band- width.

De Detect ector

• r

position position

Rotary collimator Bandwidth chopper Adjustable B4C slits

NOBORU (BL10)

NOBORU BEAM LINE

Long term operability and 3He usage

SAMPLE HOLDER µNID

AMPLIFIER-SHAPER- DISCRIMINATORS (ASD) Prototype in experimental area at NOBORU.

Same gas filling used for first two experiments (separated by 8 months). No degradation in performance seen in June experiment. Gain recovered by increasing anode voltage. Detector remained operable after more than 1 year on single gas filling.

Time after filling Gain

(% of initial)

1st Exp (2009) 2nd Exp (2010) Dec 2010 0 months 100 8 months 67 13 months 30

Strategies to extend operation

Annealing of vessel and µPIC against outgassing. Careful selection of materials. Gas purification system (c.f. Nakamura’s talk, 16:45 today).

DAQ performance at NOBORU

Time-averaged data rates from 200 kHz ~ 9.4 MHz. Neutron rate of 80~100 kHz. Large dead time (40 ~ 85%).

FPGA encoder 32 bits:

• rientation,

time, position, edge VME memory PC

VME bus 33-bit LVDS (! 2) External gate

DAQ BOTTLENECKS

Reduction in incoming data means fewer VME readouts. Effectiveness depends on details of TOF distribution and gate. Useful for Bragg transmission, resonance absorption.

EXTERNAL TOF GATE

Time (ms)

10 20 30 40 50 60 70 80 90

100

Counts

10

10

10

40 ms Neutron pulses

• n
• ff

VME-to-PC data transfer creates dead time. Encoder FIFO buffers limit DAQ rate. * LIMITATIONS CAN BE REDUCED WITH FURTHER DEVELOPMENT OF DAQ HARDWARE.

Neutron-gamma separation

Both neutrons and γ’s are detected (γ efficiency ~10-3). Neutrons selected by cuts in total time- above-threshold and 3D track length. Fraction of detected γ’s surviving neutron cuts < 10-6 (effective gamma sensitivity of < 10-9).

SAMPLE TOF DISTRIBUTIONS

Data taken at NOBORU, J-PARC in June 2010.

TOF (ms)

5 10 15 20 25 30 35 40

• No. of events

1 10

10

10

10

• No. of neutrons and gammas

Bi filter n γ

TOF (ms)

5 10 15 20 25 30 35 40

• No. of events

1 10

10

10

10

10

• No. of neutrons and gammas

Polyethylene degrader n γ γ/n < 10-9 γ/n < 10-7

Length (cm)

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

Time-above-threshold (clocks)

100 200 300 400 500 1 10

10

10

Time-above-threshold vs. Track length

Escape events Fully-contained neutrons Event pile-up, scattered protons n γ

Position resolution with PID

Y (strips)

10 20 30 40 50 60

Time-above-threshold (clocks)

5 10 15 20 25 30

Energy deposition

Cd TEST CHART

No PID With PID

Proton direction from shape of distribution Track length from end-points

Position from mid- point of track. Resolution: ~1 mm (σ) Resolution with PID: 349 ± 36 µm (σ) (Includes beam dispersion.)

Data taken at NOBORU, J-PARC in Nov. 2009.

5 cm 2 mm slits Preliminary

Refining position resolution

Two methods: End-Point Extrapolation (EPE) and Peak Interpolation (PI). Combining both methods produces best result of σ = 118.4 ± 0.2 µm.

Data taken at NOBORU, J-PARC in Feb. 2011.

NO REFINEMENT (σ = 315 µm) EPE ONLY (σ = 182 µm) EPE + PI (σ = 118 µm)

Y (strips)

10 20 30 40 50 60

Time-above-threshold (clocks)

5 10 15 20 25 30

Energy deposition

Track length from extrapolation Track length from peaks

Pulse width

5 cm 0.5 mm slits Preliminary

Image of a wristwatch

IMAGING PLATE (200 MIN.) µPIC (29 MIN.)

Courtesy of Ohi, J-PARC Preliminary

Bin size: 200 µm × 200 µm. Data taken at NOBORU, J-PARC in Feb. 2011 (µPIC).

Bin size can be decreased with higher statistics. Image processing techniques could improve image.

~3.5 cm

Demonstration measurements

Small-angle neutron scattering. Resonance imaging. Bragg-edge transmission.

MPGD2011, 30 Aug 2011

Small-angle neutron scattering

~1 cm

Radial distance (mm)

2 4 6 8 10 12 14 16 18 20

• 3

10

• 2

10

Distance from beam center

Arbitrary units

7 Å 8 Å Preliminary Spherical SiO2 nanoparticles (diameter ~200 nm). Sample-to-detector distance of 1666 mm. Exposure time of 35 min.

Radial position of peak depends on wavelength but is constant in momentum transfer, q. Expected pattern for spherical particles seen in q.

Momentum transfer vs. wavelength

Preliminary

Data taken at NOBORU, J-PARC in Nov. 2009.

Small-angle neutron scattering

~1 cm

Radial distance (mm)

2 4 6 8 10 12 14 16 18 20

• 3

10

• 2

10

Distance from beam center

Arbitrary units

7 Å 8 Å Preliminary

)

• 1

Å q (

0.004 0.008 0.012 0.016 0.02

• 4

10

• 3

10

• 2

10

Arbitrary units

q projection (6 < λ < 10 Å)

Preliminary Spherical SiO2 nanoparticles (diameter ~200 nm). Sample-to-detector distance of 1666 mm. Exposure time of 35 min.

Radial position of peak depends on wavelength but is constant in momentum transfer, q. Expected pattern for spherical particles seen in q.

Data taken at NOBORU, J-PARC in Nov. 2009.

Resonance absorption

Sheets of In, Ta, Ag, Mo, and Mn. Typical area of 10 cm × 10 cm. Thicknesses from 10 µm to 1 mm. Large samples to accumulate statistics quickly (~16 min/sample). Good time resolution and background rejection allows us to see resonances near beginning of pulse.

Time (ms)

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Transmission

0.2 0.4 0.6 0.8 1

Mn Mo Ag Ta In

0.04 0.08 0.6 0.8 1

Preliminary

Time (ms)

0.5 0.6 0.7 0.8 0.9 1

Transmission

0.2 0.4 0.6 0.8 1

Preliminary Indium compared with ENDF/B-VII.0

Data taken at NOBORU, J-PARC in Feb. 2011.

Resonance imaging

Assorted metals. DAQ rate of 2.96 MHz (neutron rate of ~30 kHz). Exposure time of 5.5 min.

Image around TOF=90.9 µs

Preliminary

s) Time (μ

50 100 150 200 250 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 / ndf

2

χ 78.14 / 42 p0 0.0097 ± 0.7814 p1 0.23 ± 90.86 p2 0.29 ± 11.28

Transmission

Transmission for 59Co

Neutron TOF (µs)

Preliminary Neutrons at resonance energy for selective imaging.

59Co resonance observed at 90.86 ± 0.23 µs.

Matches known resonance at 132 eV (TOF of 90.9 µs). Co Mo

Mn/Cu

Mo Na Cu ~2 cm

Data taken at NOBORU, J-PARC in Nov. 2009.

Bragg-edge transmission

Time (ms)

2 4 6 8 10 12 14 16 18

Transmission

0.15 0.2 0.25 0.3 0.35

Edge spacing is consistent with expected BCC crystal structure. Precise measurement of edge positions determines lattice parameter.

Fe powder (>99% pure, grain size < 325 µm). Sample thickness of 1.6 cm. DAQ rate of 2.94 MHz (neutron rate of ~30 kHz). Exposure time of 40 min. (110) (200) (211)

Data taken at NOBORU, J-PARC in Feb. 2011.

Preliminary 1.6 cm

Bragg-edge transmission

78 × 40.5 × 10 mm3 TIG-WELDED 316L STAINLESS STEEL PLATE 19 mm 78 mm

) Å ( λ

1.5 2 2.5 3 3.5 4 4.5 5 5.5

Transmission

0.2 0.25 0.3 0.35 0.4 0.45 Transmission vs. neutron wavelength

Preliminary

(111) (200) (220) (311)

Weld area Steel plate 2 Steel plate 2

50 100 150 200 250 50 100 150 200 250 3.5 4 4.5 5 5.5 6 6.5

Welded steel plate (6 < TOF < 8 ms)

Preliminary Weld area

50 100 150 200 250 50 100 150 200 250 3.5 4 4.5 5 5.5

Welded steel plate (9 < TOF < 14 ms)

Preliminary Weld area

Edge spacing is consistent with FCC crystal structure.

Data taken at NOBORU, J-PARC in June 2010.

Bragg-edge transmission

Edge positions related to spacing of crystal planes perpendicular to beam. Variation in position may be related to internal strain. Full strain tensor requires measurements from multiple directions.

Data taken at NOBORU, J-PARC in June 2010.

x ( s t r i p s )

100 110 120 130 140 150 160

y (strips)

60 80 100 120 140 160 180

3.56 3.58 3.6 3.62

) Å Edge position (

(200)

Preliminary

Weld center-line

x (strips)

100 110 120 130 140 150 160

y (strips)

60 80 100 120 140 160 180

2.15 2.16 2.17 2.18

) Å Edge position (

(311)

Preliminary Divide image into 4.8 × 4.8 mm2 ‘pixels’ and fit edge positions*.

* Fit procedure based on Santisteban, et al. (2001)

d = λ 2

d-spacing from wavelength strain component in beam direction

ε = d − d0 d0

Future improvements

Optimization of gas mixture. Smaller pitch µPIC. New ASICs and encoder for more compact DAQ.

MPGD2011, 30 Aug 2011

Gas optimization and pixel pitch

Pressure

(atm)

Drift velocity

(µm/ns)

Transverse diffusion

(µm/cm1/2)

Longitudinal diffusion

(µm/cm1/2)

Expected improvement in resolution

Ar:C2H6:3He (63:7:30) Ar:C2H6:3He (63:7:30) Xe:C2H6:3He (50:20:30) Ar:CO2:3He (50:20:30)

2 23.1 273 169 (118 µm) 3 23.4 231 126 ~15% 2 29.4 183 125 ~15% 2 22.5 107 114 ~15%

Shorten p-t track lengths by increasing pressure

• r changing to gas with higher stopping power.

Reduce diffusion of drift electrons. Moderate reductions in pixel pitch produce corresponding reduction in position resolution.

Gas parameters determined by MAGBOLTZ. Resolutions estimated with GEANT4.

400 µm 350 µm

REDUCE PIXEL PITCH

Gas optimization and pixel pitch

Pressure

(atm)

Drift velocity

(µm/ns)

Transverse diffusion

(µm/cm1/2)

Longitudinal diffusion

(µm/cm1/2)

Expected improvement in resolution

Ar:C2H6:3He (63:7:30) Ar:C2H6:3He (63:7:30) Xe:C2H6:3He (50:20:30) Ar:CO2:3He (50:20:30)

2 23.1 273 169 (118 µm) 3 23.4 231 126 ~15% 2 29.4 183 125 ~15% 2 22.5 107 114 ~15%

Shorten p-t track lengths by increasing pressure

• r changing to gas with higher stopping power.

Reduce diffusion of drift electrons. Moderate reductions in pixel pitch produce corresponding reduction in position resolution.

Gas parameters determined by MAGBOLTZ. Resolutions estimated with GEANT4.

400 µm 350 µm

REDUCE PIXEL PITCH

DAQ improvements

µ-­‑PIC

CR board

To memory board

Replace ASDs with CMOS chips (developed with KEK). 16 channels/chip (increased from 4). Power per channel reduced by factor of more than 3.

4 mm

New CMOS ASIC

Combine CMOS chips with FPGA on single board. Four boards replace ASD racks, encoder, cables. Each board writes to memory, increasing max. data rate. New boards now under testing (cf. Iwaki, poster 67).

ASDs FPGA ENCODER

10-cm prototype in lab at Kyoto Univ.

~25 cm ~35 cm

Analog Signal 128ch Analog Data & Trig. Control

118mm × 220mm FPGA CMOS ASICS ADC

Summary

TPC based on micro-pattern gaseous detector and FPGA DAQ system.

Position resolution of 118 µm; time resolution of ~1 µs. Compact DAQ with high data rates. Strong rejection of gammas and fast neutrons.

Detector remains operable over long time.

Annealing to reduce outgassing for increased long-term stability. Gas filtration system could extend operation considerably.

Continuing studies to improve detector performance with aid

• f GEANT4 simulation.

Gas mixture and pixel pitch optimization.

Setting up 20-cm neutron imaging detector for use at Kyoto University.

µPIC sizes up to 30 × 30 cm2 are currently available.