Somewhat different applications for photon counting pixel detectors 11th ICATPP Conference Villa Olmo, Como, October 9 2009 Thilo Michel , G. Anton, J. Durst, P. Takoukam, E. Guni, M. Böhnel, U. Gebert, T. Rügheimer, M. Firsching
Jan 02, 2016
Somewhat different applications for
photon counting pixel detectors
11th ICATPP Conference
Villa Olmo, Como, October 9 2009
Thilo Michel, G. Anton, J. Durst, P. Takoukam, E. Guni, M. Böhnel, U. Gebert,T. Rügheimer, M. Firsching
Outline
• Material resolved X-ray imaging
• Radiation monitoring
• X-ray polarimetry
• Hybrid Photon Detector (HPD)
• Summary
The Medipix2/Timepix detectors: hybrid photon counting pixel detectors
ASIC/Sensor: – Development: International Collaboration
with seat at CERN– Bump-bonded with Pb/ Sn– 65536 pixels– Pixel pitch: 55 µm– Size of the matrix: 14 mm (approx. 2 cm2)– 0.25 µm CMOS technology
Sensor:– Materials: Si, GaAs, CdTe– Bias voltage: e.g. 150 V (300 µm Si)
14 mm
E
Timepix: counting or Time-Over-Threshold or Time-To-Shutter in each pixel individually
Time-over-Threshold Time-To-Shutter
Reactionin sensor
Outputpreamp.
Discrimi-nator
Counter
ToT = f(Edepos) = 10 nsec*Nclockpulses
t
t
t
t
TtS = 10 nsec*Nclockpulses
t
t
t
t
DAQ
t
Start Stop
DAQ
t
Start Stop
Reactionin sensor
Outputpreamp.
Discrimi-nator
Counter
Charge sharing affects the energy resolution
Simulation of the response matrix: deposited energy versus primary energy
Simulated set of response functions
Principle of material resolved imaging
X-Ray transmission througha compound object
mass attenuation coefficient areal density
index for basis materialsDetector
source
Aim: determine the densities of (base) materials in the object
Material reconstruction without explicit spectrum reconstruction
Number of counts in energy
depositioninterval
Sum overenergy
depositionsin
interval
Sum over
primary energies
in spectrum
Sum over
materials
Vary to get bestagreement of Mb
and measurement in a 2-fit
Test in a small animal scanner
• Measurements performed with a small
animal scanner of the University of
Canterbury fulfilling ethics
requirements
• Threshold adjustment with flatfield
images at the K-edge of iodine
• Measurement and reconstruction of
the X-ray tube spectrum with Medipix2
• Data acquisition: 4 images with
different thresholds (12, 17, 33 und 42
keV) per detector position (3) and
projection (360)
Details
Clinical contrast agentwith iodine
Material reconstructed projections
Photon counting image Non-iodine (water) image Iodine image
Material reconstruction in CT: Brightness = density of base materials (g/cm3)
Photon counting images @ 4 thresholds
water /non iodine
iodine
Brightness = density of materials
Outline
• Material resolved X-ray imaging
• Radiation monitoring
• X-ray polarimetry
• Hybrid Photon Detector (HPD)
• Summary
Methods for the reconstruction of incident X-ray spectra at very high flux
Response spectrum of polychromatic irradiation
N
jjjii ESEEREM
1
)(),'()'(
SRM
Matrix-Inversion-Method
MRRRS TT
1)(
Best estimate for impinging spectrum is given by:
S
Spectrum-Stripping-Method
Subtract the monoenergetic response functions successively:
NN
NN R
MS
1,1
,111
NN
NNNNN R
SRMS
Medipix2 response spectrum
Impinging spectrum
Energy [keV]
Nu
mb
er o
f co
un
ts o
r p
ho
ton
s
R(Ei|, Ej) : the probability that a photon of
energy Ej causes an energy deposition of Ei|
…
Threshold scan of the spectrum of an Am-241 photon source
Energy [keV]
Inte
nsity
[a.u
.]
Derivative of the threshold scan
Energy in keV
Inte
nsity
[a.u
.]
Reconstructed spectrum of the Am-241 photon sourceusing the (iterative) spectrum stripping method
• Reconstruction bin width 0.5 keV
• Energy resolution (RMS) better than 1 keV
• 0.25 Counts per pixel per second at the lowest energy threshold (5keV)
• 8 days acquisition time (Medipix operates very stable)
Nu
mb
er o
f p
ho
ton
s p
er 0
.5 k
eV
14.6keV
18.9keV
22.0keV
26.3keV
59.5keV
Energy in keV
Measurement of an X-ray tube spectrum at high fluxN
um
ber
of
ph
oto
ns
per
2 k
eV
Energy in keV
Literature(scaled)
Measured
Method for dosimetry with pixel detectors
Monochromatic irradiation with Ej
Calibrate with jmax different photon energies Apply pseudoinverseof the „counts matrix“
iENEH jiijp )()(
max
1max
)(1
)(i
ijiijp EN
iEH
maxmaxmaxmax
max
max
...
)(...)(
.........
)(...)(
)(
...
)( 1
1
1111
ijij
i
jp
p
ENEN
ENEN
EH
EH
Known doses
Number of counts in
each energy deposition
bin
Unknowncalibration
factors from counts to
dose
Measure/simulate the number of counts in energy deposition intervals
pTT HNNN
1
Best estimation (maximum-likelihood) for calibration factors
Energy deposition [keV]
Co
un
ts
N1(60) N2(60) N3(60)
Experimental test of the method with the Medipix2
Measurement details
• Threshold adjustment registers detuned: 8 thresholds simultaneously (15 – 71 keV)
• Calibration with 9 different X-ray spectra (filtered/unfiltered)
• Air-Kerma free-in-air measured with reference dosemeter
• Test with different X-ray spectra
Result of dose (kerma) reconstruction for X-ray spectra
.
.
.
N1(RQR 40 keV)
N8(RQR 40 keV) Good reconstruction of air kerma free-in-air
Rel
ativ
e er
ror
of
reco
n. K
erm
a [%
]
Tube voltage [kV]
Measurement of the temporal structure of pulsed radiation fields with the Timepix
Measurements performed in collaboration with the PTB
• Portable 150 kVp X-ray generator
• Timepix operated in Time-To-Shutter mode with 90 MHz clock frequency
• Approximately 12 m away from tube
• Analysis of single hits
Time after X-ray flash in ns
Nu
mb
er o
f co
nts
in a
.u.
Outline
• Material resolved X-ray imaging
• Radiation monitoring
• X-ray polarimetry
• Hybrid Photon Detector (HPD)
• Summary
Measurement of the degree of linear X-ray polarization using photoelectric effect
Polarized irradiation
• This example: The probability to trigger adjacent pixels in one column (Nc) is higher than to trigger two adjacent pixels in one row (NR)
• Measuring quantity: the asymmetry
PANN
NNA
RC
RC
%100
E
NC
NR
Unpolarized irradiation
• Differential cross section for K-shell photoeffect:
• Only the first microns carry the information about the original direction of emission
• Diffusion „disturbs“ the polarization signal
Total path length:R(60keV) =
23 µm
Experimental setup
Lead/tungsten
collimator
Target (PMMA)
Production of linearly polarized photons by 90° Compton scattering
Timepix on rotation
device
X-ray tube(100 kVp) +
Fe filter
P = 98.6 %
The measured asymmetry is modulated with the angle of the columns to the plane of polarization
Apol = (0.96 +- 0.02) %
Aapp = (0.19 +- 0.01) %
apppol AAA )2cos()(
Apol = (0.888 +- 0.042) %
• Simulation (ROSI):
Results
Counting mode
• Measurement:
T. Michel, J. Durst (2008): “Evaluation of a hybrid photon counting pixel detector for X-ray polarimetry'', Nucl. Instr. and Meth. A 594: 188-195
Measurement of the degree of polarization in dependence on energy deposition in double hit events
2
)90()0(
AAApol
2
)90()0(
AAAapp
• Subtraction of target and no-target measurements
• Polarization asymmetry
• Apparative asymmetry
• Simulation and measure-ment in agreement
• Apol = 0.2 % @ 29 keV• Apol = 3.4 % @ 78 keV
Time-over-threshold mode
Principle of polarimetry using Compton scattering
Comptonelectron(1st hit)
Comptonphoton(2nd hit)
Impinging photon
Timepix-ASIC
E
Scattered photon
From topCross section
Cross Section of Compton scattering
Measurement result after search of clusters, coincidences, distance cuts, random subtraction, acceptance correction
• Modulation factor µ:
• Modulation curve:
BAM )(cos)( 02
PII
IIµ
%100minmax
minmax
BAI max
BI min
)%4.161.68( Measuredµ
Comptonelectron(1st hit)
Comptonphoton(2nd hit)
Comptonelectron(1st hit)
Comptonphoton(2nd hit)
E
Nu
mb
er o
f ev
ents
Scattering angle °]
T. Michel, J. Durst, J. Jakubek (2008): “X-ray polarimetry by means of Compton scattering in the sensor of a hybrid photon counting X-ray detector'', Nucl. Instr. and Meth. A, accepted manuscript
Simulation resultN
um
ber
of
even
ts
Scattering angle °]
• Simulation: ROSI (EGS4-based)
• Incoming radiation completely polarized
• Modulation factor:
• Degree of polarization of detected events:P = 97.1%
• Result:
)%2.57.64(%100 Simulationµ
)%0.58.62( Simulationµ
Comptonelectron(1st hit)
Comptonphoton(2nd hit)
Comptonelectron(1st hit)
Comptonphoton(2nd hit)
E
Outline
• Material resolved X-ray imaging
• Radiation monitoring
• X-ray polarimetry
• Hybrid Photon Detector (HPD)
• Summary
Measurements with an HPD test set-up at CERN
High voltage discharge lamp Vacuum vessel with
Deflection mirror (position adjustable) CsI photocathode Accelerating electric field (max. 25 kV) Timepix chipboard (mounted upside down) Vacuum: about 10-5 mbar
Experimental set-up at CERN: Timing:
Precision of the timing information oflight signals (time resolution): 3 ns Due to the statistical light flashes of the lamp Trigger the end of the acquisition 3µs after the lamp flash
In collaboration with:
Jacques Séguinot
Christian Joram
André Braem
In collaboration with:
Jacques Séguinot
Christian Joram
André Braem
Results of the timing measurements
Time resolution (singles):single = 10.4 ± 0.1 ns
All hits
Only singlehits
20 keV
An ion feedback time stampN
umbe
r of
hits
/ 1
0 ns
Time stamp in 10 ns
Fotography of the ion feedback
Investigating the imaging properties of the test set-up (proximity-focusing electron optics)
Position resolution as a function of acceleration voltage
d = 41mmEi = 0.45eV
Energy in in keV
Wid
th o
f Poin
t-S
pre
ad-F
uct
ion
in µ
m
Outline
• Material resolved X-ray imaging
• Radiation monitoring
• X-ray polarimetry
• Hybrid Photon Detector (HPD)
• Summary
The Medipix/Timepix can be used in a lot of applications, because they can measure ....
N(E) (x,y)
N(E) and N(t)
P(x,y)
t(x,y)
Acknowledgement
ECAP:G. Anton, J. Durst, P. Takoukam-Talla, E. Guni, A. Loehr, M. Böhnel, U. Gebert, T. Rügheimer, P. Bartl, B. Kreisler, F. Bayer, P. Sievers, F. Lück, I. Münster, A. Ritter, T. Weber, W. Haas,M. Firsching (now at Fraunhofer)Collaborators:J. Jakubek (CTU Prague), S. Pospisil (CTU Prague), A. Butler (U Canterbury), J. Seguinot (CERN), C. Joram (CERN), A. Braehm (CERN), M. Campbell (CERN), P. Ambrosi (PTB), U. Ankerhold (PTB), O. Hupe (PTB)
…and the whole Medipix collaboration