Transcript TH0(E)

Characterization of the C-MOS Cd-Te Imager Pixirad
for energy discriminated X-ray imaging
D. Pacella1, A. Romano1, G. Claps1, F. Causa1, L. Gabellieri1
1
Association EURATOM-ENEA, C.R. Frascati, via E. Fermi, 45 - 00044 Frascati, Rome, Italy
email: [email protected]
INTRODUCTION
 The goal of this work is the characterization of the PIXIRAD Imaging Counter to assess it as candidate for high definition energy resolved X-ray imaging for slow control in magnetic fusion plasma experiments
 In magnetic plasma experiments, the energy resolved X-ray imaging is appealing because it allows to investigate different regions of the plasma, which emit in different X-ray spectral bands depending on electron temperature and
impurity content
 X-ray energy resolved imaging could allow better performances in radiography and tomography in many domains: bio-science, medical imaging, material science
DESCRIPTION OF PIXIRAD











SETUP
The PIXIRAD Imaging Counter is an INFN-Pisa Spin-off
It works in photon counting
Pulse discrimination with two thresholds
Active area: 30.7×24.8 mm2
Pixels matrix: 512×476, each one of 55×55 μm2
CdTe sensor: 650 μm thick
Energy range: 2-100 keV
Count rate > 30 GHz
Frame rate ~ 100 frame/s
Noise free
Acquisition: 2 color reading (2 thresholds, 2 counters) or counting
in one counter while reading the other one
Fig. 1
 Two absolutely calibrated X-ray sources
(Moxtek 50 kV Bullet and Oxford
Instruments SB-80-1M) were utilized
 SXR line transitions were generated by
fluorescence on different samples (Ca, Fe,
Cu, Br, Mo, Ag, I, Ta) and by means of
BaCs radioactive source
PIXIRAD
SAMPLE
X-ray source
 The spectra were measured and
optimized with an SDD (XGLab) up to 30
keV or with a CdTe spectroscopic detector
(Ampetek X-123CdTe )
Fig. 2
Fluorescence lines:
 Ca 3.7 keV
 Fe 6.4 keV
 Cu 8.0 keV
 Br 11.9 keV
 Mo 17.4 keV
 Ag 22.1 keV
 I
28.6 keV
 Ta 57.5 keV
Radioactive source:
 BaCs 81 keV
SMOOTH ENERGY BAND DISCRIMINATION
ENERGY RESOLUTION
 As an example, pulse amplitude
distributions are simulated as
gaussians
 The threshold is defined as the number of electrons collected on the pixel (and
then transformed in a pulse); it goes from 50 to 5000 electrons, while the
dynamic range arrives up to about 10000
 Dotted line represents the threshold
at 8 keV
 Our scans start from 350 electrons in order to cut completely the noise off (Fig. 3)
 Different lines are cut off:
3.7 keV
10.5 %
6 keV
27.6 %
8 keV
50.8 %
12 kev
89.6 %
17 keV
99.9 %
 The scan of Ta starts from a 2800 threshold because the low energy side of the
spectrum could not be cut off
 Scans are normalized to the value corresponding to the minimum threshold (350)
Fig. 3
Fig. 5
 Energy discrimination in bands is possible, in a “smooth way”
 Since the pulse amplitude distribution is pretty wide, there is no
correspondence one to one between threshold and energy
 With TH0(E) we can cut all the energies below E off, but the counts corresponding to
energies higher than E are reduced, less and less as E increases
 Th0(E) is defined as the threshold where the pulse amplitude distribution,
corresponding to an energy E, is completely cut off
 With TH50(E) we can cut the counts, corresponding to the energy E of 50%
 Th0(E) is defined as the threshold where the pulse amplitude distribution,
corresponding to an energy E, is cut of 50%

 Discrimination of Fe (6.4 keV) and Cu (8 keV)
lines (red curve) is smoother than discrimination
of Cu (8 keV) and Mo (17.3 keV) lines (blue
curve)
TH0(E)= -2.77×103 + 5.23×103 log (E)

TH50(E)= -0.64E2 +
1.14×102E+1.03×102
 Thresholds vs Energy are linear up to about 30 keV, revealing a good energy resolution
Fig. 4
 Beyond this value non linearity and then saturation are due to the cluster size that becomes larger than one because the energy is
no more released in one single pixel
Fig. 6
ENERGY DISCRIMINATED IMAGING
 The energy resolution is coupled to the imaging
properties
Fig. 7
Moxtek
X-ray source
 Images at different energy bands can be retrieved
thanks to the energy discrimination
MF spectrum
35 kV-100 μA
Moxtek spectrum
15 kV-100 μA
Microfocus
X-ray source
PIXIRAD
Fig. 8
MF 35 kV – 100 μA and Moxtek 15 kV – 100 μA
THR = 350
Fig. 9
Polycapillary lens
 An image of a Microfocus X-ray tube,
in the range 12 -35 keV (see
spectrum in Fig. 8) is produced on
the detector by means of a
polycapillary lens (Fig.10). On the
right of the Fig. 10 is the intensity
along the line shown in the X-ray
picture
MF 35 kV – 100 μA, THR = 350
Fig. 11
MF 35 kV – 100 µA and Moxtek 15 kV - 100µA
THR = 2600 corresponding to TH0 (12 keV)
Fig. 10  Counts across the modulations go
from 34 to 7 (contrast =5)
CONCLUSIONS
 Energy resolution is good up to 30 keV and then progressively degrades due to the cluster size higher than one
 Different thresholds have been defined to describe the effects in term of energy discrimination
 A “smooth energy discrimination” in bands has been demonstrated
 Energy discriminated imaging has been shown, revealing as an example “hidden” structures buried under a background
 This technique could improve significantly the imaging capabilities for radiography and tomography, in many fields, from
bio-sciences to imaging and material science
 When an X-ray uniform background
(100 times more intense), produced
by another tube at 15 kV (see
spectrum in the Fig. 9), is
superimposed to the previous one, it
covers
completely
the
image
produced by the lens
 Once we put the threshold at 2600 ,
corresponding to an energy of 12 keV
(TH50), the background is completely
cut off and the image of the lens
appear again (Fig. 12), only slightly
less intense (counts from 20 to 4) but
with the same contrast (=5)
Fig. 12
 This is possible thanks to the energy
discrimination of the detector
REFERENCES
[1] http://pixirad.pi.infn.it
[2] D. Pacella, et al., Self Consistent Calibration of Detector and Sources for Hard and Soft X-Ray
Diagnostics, Modern Instrumentation (2014).
[3] D. Pacella, et al., Polycapillary optics for soft X-ray imaging and tomography, Il Nuovo Cimento,
Vol. 34 C, N. 4, 513-520 (Luglio-Agosto 2011).
[4] A. Romano, et al., Characterization of a 2D soft X-ray tomography camera with discrimination in
energy band, Rev. Sci. Instrum, Vol. 81, 10E523 (2010).
16th International Workshop on Radiation Imaging Detectors - Trieste, Italy 22-26 June 2014