Prime pagine RA2010FUS:Copia di Layout 1 - ENEA - Fusione

024
progress report
2010
New diagnostics for soft x–ray imaging and tomography. Magnetic fusion plasmas (MFP) are extended sources
of x–rays and these emissions could reveal a lot of information about the processes occurring inside the
plasmas.
The aim of this project is the development of a 2–D detector with independent energy discrimination
capability for each pixel. Plasma diagnostics based on soft x–ray (SXR) tomography and/or imaging for
magnetic confinement fusion plasmas could be greatly enhanced if different energy bands (which are
representative of different plasma zones and their impurity content) could be selected dynamically. A gas
detector with 2–D pixel read–out is being proposed for such a diagnostics.
Moreover, the constraints posed by toroidal devices (highly radiative background, extremely high
radiofrequency powers, high magnetic fields, optical limitations and so on) are very severe and strongly limit
the possibility to install x–ray detectors directly into, or close to the machine. Therefore, it becomes mandatory,
in particular for future burning plasma experiments, to study the possibility of transporting the SXR radiation
far from the machine. Polycapillary lenses appear promising for these purposes and suitable to be used for
x–ray imaging and tomography in MFP. All these activities have been carried out in collaboration with the
National Institute of Nuclear Physics (INFN) – Frascati National Laboratory (LNF) and CEA laboratory of
Cadarache, under the auspices of EFDA (Work Program 2010 (WP2010)).
1) Realization and characterization of a triple GEM gas detector. A triple gas electron multiplier (GEM) detector
has been designed and built by the LNF–INFN laboratory of Frascati and subsequently installed in the SXR
laboratory of CEA for this preliminary characterization [1.13]. It is based on a gas detector with triple GEM
as amplifying structure and with a two–dimensional read–out, coupled to the integrated front–end electronics.
The energy discrimination in bands, at the level of individual pixel, has been studied. The front-end
electronics of the GEM detector, working in photon counting mode with a selectable threshold for pulse
discrimination, is optimized for high rates.
The energy resolution of the detector has been accurately studied in laboratory with continuous SXR spectra
produced by an electronic tube (continuous spectra with a Moxtek 40 kV Bullet) and line emissions produced
by fluorescence (K, Fe, Mo), in the range 3–17 keV (fig. 1.28). For the measurements presented, the detector
has been filled with a gas mixture at atmospheric pressure with
10 2 counts/s
6
4
2
K
Fe
Mo
0
0 10 20
Energy (keV)
Figure 1.28 – SXR fluorescence of K, Fe and Mo
samples measured with the Si–PIN detector
Counts (Arb. units)
0.6
0.4
K
3.3 keV
6.39 keV
0
0 100 200 300 400
Energy (mV)
Figure 1.29 – Reconstructed spectra for
samples of K and Fe sources
Fe
Ar (70%) and CO 2
(30%). In order to assess the intrinsic
energy resolution of the detector, the distribution of the pulse
amplitude of the signals collected on a single pixel has been
indirectly derived, for different K α
line radiations (K, Fe, Mo).
The integral of all the counts whose amplitude is greater than
the threshold has been measured instead of the effective
distribution of the counts as function of the peak amplitude.
The distribution of the pulse amplitude has been indirectly
derived by means of scans of the threshold and by fitting it
with a guess function with three free parameters: position of
maximum and half width at half maximum of the Gaussian
peak and decay length of the tail (fig. 1.29). The distribution of
the pulse amplitude, in the range 3–17 keV, was found nearby
a Gaussian. The best agreement is found for a total detector
gain of 500 and an energy resolution of about 30% and a tail
at higher energy. The pulse broadening is entirely due to the
poor energy resolution of the detector. Scans in detector gain
have been also performed to assess the capability of selecting
different energy ranges. Combining together two samples, Fe
and Mo, fluorescence spectra with two lines have been
generated, to investigate the sensitivity of the threshold scan to
a more complex spectrum.
The reconstruction, by means of a threshold scan, of the
incident spectrum has been demonstrated in case of simple
spectra, single or double line emissions or bell shaped features.
In case of more complex spectra, with different features, the
only scan in threshold might be not sufficient. In this case a