Chiara Casella Matthieu Heller Oliver Holme 1 Dualsided readout studies Difference in Left and Right light yield is used to estimate gamma interaction position along z axis Original AXPET concept was based on dualsided readout ID: 515749
Download Presentation The PPT/PDF document "Dual-sided readout studies" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Dual-sided readout studies
Chiara CasellaMatthieu HellerOliver Holme
1Slide2
Dual-sided readout studies
Difference in Left and Right light yield is used to estimate gamma interaction position along z axis
Original AXPET concept was based on dual-sided readout
Initial studies showed insufficient z-axis resolution (~cm)
New results from A. Goertzen indicated better z resolution (~3.5mm)Investigation started to find best achievable resolutionFollowing A. Goertzen methodUsing digital SiPM for a compact axial module
L
R
z-axis
0
2Slide3
Laboratory setup
Philips digital SiPM tiles – L, R and TaggerL and R tiles are cooled to ~14˚C to reduce dark countOnly events with 3 tile coincidence are stored
Acts to collimate a beam of gamma photons
Automated data taking with Philips and LabVIEW software
Scans through several position along z-axis
3Slide4
Crystal scanning
3x3x100mm LYSO crystalsCrystal aligned with one SiPM pixelAll other cells were inhibited
Typical scan properties
11 data taking runs at different positions along z axis
Typical run properties~15 minutes (with 3.7 MBq source)Up to 9000 events stored to fileUp to 1300 events after photopeak cuts
4Slide5
Data processing for each z position
Data from SiPMs is converted to ROOT filesLight yield is processed:Cut on photopeaks (all three tiles)
Correct for SiPM cell saturation
Correct for SiPM intra-tile cross talk
Obtain photopeak central values
Raw data
Gaussian fit
Corrected data
5Slide6
Data processing for complete scan
z position estimator calculated as
for each event
Mean
for each z position is plotted against z
Fit function is applied to plot
Fit gives a relationship between any
value and z position
6
Error bars show
σ
of (R-L)/(R+L)
(R-L)/(R+L)
(R-L)/(R+L)
z
position [mm]Slide7
Data processing for complete scan
z axis resolution
FWHM
of
values for each z position mapped to FWHM of zBeam width is subtracted to obtain final
resolution
Beam size calculated as 1.7mm
7
R
z
FWHM
Error bars show
σ
of (R-L)/(R+L)
(R-L)/(R+L)
P
osition [mm]
FWHMSlide8
Crystals and surface treatments
Coatings/wrappings:BareWhite TiO
2
paint
Teflon (2 layers)Enhance Specular Reflector (ESR)Surface treatments (one or multiple faces):Polished (from manufacturer)Depolished (with ~10 µm powder)Machined strips (with CNC diamond tool)
8Slide9
Machined strips
Two patterns of machined stripsEvery 5mm, aligned on four faces
Every 5mm, staggered by 1.25mm on four faces
Machined strips are ~0.7mm wide
9Slide10
Results
10Slide11
Non-exponential LY versus z
Crystals with less attenuation
LY versus z is not exponential
Crystals with greater attenuation
LY versus z is more similar to exponential
Polished, teflon wrapped
Two faces depolished, teflon wrapped
11Slide12
Inter-tile cross talk
SiPMs have cross talk between cells in a tileIntra-cell cross talkCaused by avalanches producing secondary photons
The secondary photons
are emitted isotropically
They can also enter and propagate through the crystalInter-tile cross talkInter-tile cross talk reduces L & R differenceDegrades z resolutionThis degradation can be measured by capturing L and R tiles independently, with the other tile off
Not possible in real applicationsPhysically filtering the cross talk photons might be possible12Slide13
Photopeak shifting
LEFT
RIGHT
SUM
13
Machined strips, 5mm (aligned), teflon wrapped
z-axisSlide14
Light yields and energy resolution
14
Machined strips, 5mm (aligned), teflon wrappedSlide15
z resolution
15
Machined strips, 5mm (aligned), teflon wrappedSlide16
Different surface treatments
Resolution down to ~3mm has been achievedTreatments give different energy resolutions due to different LY sum
Polished + TiO
2 paint
Machined strips (four faces, aligned) + ESR Depolishing (two faces) + ESR
Energy resolution
Energy resolution [%]
16Slide17
Machined strips
Fine scans reveal discrete structureGives stepped LY patternSeen in LY sum and energy resolution too
Aligned strips, light yield - RIGHT
Staggered strips, light yield - RIGHT
Fitted photopeak position [pe]
Fitted photopeak position [pe]
Position [mm]
Position [mm]
Staggered strips give a pseudo-continuous behaviour
17Slide18
Depolishing
One face depolished + ESR
Two faces depolished + ESR
18Slide19
Wrappings
Four faced machined strips (aligned)
ESR wrapped
λ
eff ≈ 70 mm
Teflon wrapped
λeff ≈ 40 mm
Unwrapped λeff ≈ 28mm
Light yield - LEFT
Light yield - RIGHT
Light yield - SUM
z
[mm]
z
[mm]
z
[mm]
ESR gives almost constant LY SUM
19Slide20
Wrappings
Four faced machined strips (aligned)
Teflon and ESR give similar resolution
ESR gives more linear z position estimator
ESR gives more light to improve resolution further
20Slide21
Results summary
21
Crystal
surface
Wrapping/Coating
Rz FWHM
atmid-point (mm)
Energy resolution at mid-point (%)
PolishedBare52
8PolishedTeflon
228PolishedESR37
7
Polished
TiO
2
paint
3.1
15
Depolished
(1 face)
Bare
5.8
18
Depolished
(1 face)
Teflon
4.4
12
Depolished
(1 face)
ESR
4.9
8
Depolished (2 faces)
Teflon
3.4
16
Depolished
(2 faces)
ESR
3.7
11
Machined
strips (aligned)
Bare
3.6
21
Machined strips (aligned)
Teflon
2.6
12
Machined strips (aligned)
ESR
3.2
8
Machined strips (staggered)Teflon3.1
12
Machined strips (staggered)
ESR
3.8
7
A crystal with better R
z
at z=50mm is not necessarily better at other z positionsSlide22
Next steps and outlook
Improve results for 100mm crystalsExtend machined strips to full crystal lengthDepolish
more faces
Shorter 60mm crystalsCould achieve resolution down to 2mmAttenuation per mm would need to be increasedPhotodetectors with higher PDE ?Greater light yield would enable better resolutionCurrent SiPM PDE is about 40%22Slide23
23Slide24
Toy Monte Carlo results
Measured results are consistent with Toy MCEstimated ~3mm as best achievable resolutionCan indicate future possibilities
λ
eff
(mm)Crystal
Length (mm)Max LYRz
at mid-point FWHM (mm)33100
15003.0333
1003000
2.1433601500
2.2720601500
1.85
10
60
1500
1.99
20
60
3000
1.32
7
20
1500
0.70
7
20
3000
0.53
20mm crystal figures match other published studies
Increased LY improves resolution
60 mm crystals would enable improvement
down to 2mm with current light yields
24
Pure Poisson distribution
Noise
Not representative
Slide25
Machined strips width
25Slide26
Depolished (one face)
26