In this work, we describe the solution developed by the gamma ray camera upgrade enhancement project to improve the spectroscopic properties of the existing JET γ-ray camera. Aim of the project is to enable gamma-ray spectroscopy in JET deuterium-tritium plasmas. A dedicated pilot spectrometer based on a LaBr3 crystal coupled to a silicon photo-multiplier has been developed. A proper pole zero cancellation network able to shorten the output signal to a length of 120 ns has been implemented allowing for spectroscopy at MHz count rates. The system has been characterized in the laboratory and shows an energy resolution of 5.5% at Eγ = 0.662 MeV, which extrapolates favorably in the energy range of interest for gamma-ray emission from fast ions in fusion plasmas.
I. INTRODUCTION
Gamma-ray spectroscopy is a plasma diagnostic technique which can investigate the behaviour of fast ions in high temperature fusion plasmas, as demonstrated at JET.1–6 In particular, it plays a key role in the study of alpha particle confinement, which is crucial for plasma self-heating in a high power discharge. Gamma-ray emission in thermonuclear plasmas is mainly due to reactions between fast particles and fuel ions or impurities. Of particular relevance is the detection of 4.44 MeV gamma-rays from the 9Be(α, nγ)12C reaction, as it gives information on alpha particles in deuterium-tritium plasmas. At JET, a horizontal and a vertical neutron/γ camera7 provide information on the radial profile of the neutron/γ emission source by collimated measurements along 19 channels. The gamma-ray camera upgrade project aims to improve the spectroscopic properties of the existing γ-ray camera of JET in terms of energy resolution and high counting rate capability, in order to operate in the deuterium-tritium (DT) campaign. This requires a rather significant improvement of the existing CsI detector performance, in order to reach an energy resolution of 5% (Energy Resolution = FWHM/En) at 1.1 MeV and count rate capability in excess of 500 kHz. An important constraint is the limited available space which, for example, makes it impossible to use photo-multiplier tubes. The use of fast, high light yield inorganic scintillators such as LaBr38 together with Silicon Photo-Multipliers (SiPMs) can represent a good alternative to the existing CsI and photodiodes, given their high photon detection efficiency, high internal gain, insensitivity to magnetic field, and an extremely compact size. SiPMs, which have experienced great improvements in the last years, still show voltage-temperature dependence9 and a limited linearity but both issues can be corrected for.9 In this work, we describe the solution developed to meet the requirements. This combines the good energy resolution results achieved by Grodzicka et al.10 and the promising counting rate capability reached by Nocente et al.11 The pilot detector is based on a LaBr3 scintillator crystal (25.4 × 16.9 mm2) coupled to a silicon photo-multiplier (12 × 12 mm2, see Fig. 1). The final design with the electronic readout circuit of the pilot spectrometer will be presented together with the laboratory measurements at low count rate with radiation sources and at high count rate with a LED pulser.
LaBr3 scintillator crystal and silicon photo-multiplier with its readout circuit board.
LaBr3 scintillator crystal and silicon photo-multiplier with its readout circuit board.
II. SILICON PHOTO-MULTIPLIER AND ELECTRONIC READOUT CIRCUIT
Silicon photo-multiplier detectors, also known as Multi-Pixel Photon Counters (MPPCs), are relatively novel solid state photo-sensors. They are made up of multiple Avalanche Photo-Diode (APD) pixels connected in parallel and operating in Geiger mode which provide an internal high gain of the order of 106, depending on the operational condition. For each APD cell, the Geiger mode is activated when a reversed bias above the electrical breakdown voltage (Vbd) is applied. The chosen SiPMs are 12 × 12 mm2 in size, made by 16 channels with 3464 pixels each, and they are manufactured by Hamamatsu, model S12642-0404PB-50. Large area/high pixelization MPPCs are needed in order to match the scintillator size and for photon counting in the multi MeV energy range. The characteristic I-V curve of the device shows a steep increase of the current at the breakdown voltage, which is roughly 65 V. In order to achieve high counting rate capability, a short output signal is necessary to minimize the fraction of pile up events. For this purpose, the dedicated electronic circuit suggested in Nocente’s article11 has been developed by implementing a CR differentiator circuit on the MPPC readout board of Fig. 1. Fig. 2 shows an example of a signal from a standard 137Cs radioactive source after the CR differentiator. A signal width of 120 ns has been achieved without a significant loss of the amplitude (about 60 mV at the 662 keV full peak from 137Cs). This is an improvement with respect to what was done in Ref. 11 since the MPPC signal output can be fed directly in the ADC without the need for further amplification.
Output signal from the LaBr3 crystal coupled to the SiPM with the pole zero cancellation network.
Output signal from the LaBr3 crystal coupled to the SiPM with the pole zero cancellation network.
III. LABORATORY TEST AT LOW COUNTING RATE
A. Measured spectra and energy resolution
Laboratory measurements with standard radioactive sources have been performed in order to characterize the MPPC response and its dedicated electronic readout circuit. The MPPC was coupled with a LaBr3 scintillator crystal (25.4 × 16.9 mm2) and powered with a bias voltage of 65.2 V–67.5 V provided by a TTi EX752M voltage supply. Optical grease and an aluminum foil were used to improve light collection from the scintillator. The output signal from the detector was fed into a waveform digitizer CAEN module DT5730 (14 bit, 500 Msps) equipped with CAEN software able to perform on-line measurements of the pulse area. In order to characterize the pilot spectrometer, several measurements have been performed at different bias voltage, revealing an improvement in the energy resolution by increasing the applied voltage up to 67.5 V. Example of calibrated pulse height spectrum with 137Cs (Eγ = 662 keV) and 60Co (Eγ = 1173 and 1333 keV) radioactive sources at Vbias = 67.2 V (roughly 2.3 V over Vbd) is shown in Fig. 3. The peaks of the spectrum were fitted by a Gaussian function on a background described by polynomials. An energy resolution (Energy Resolution = FWHM/Energy) of 5.5% was obtained at 662 keV which improves to 3.7% for 1333 keV gamma rays. The trend of the energy resolution (see Fig. 4) is well fitted by the curve12 f(E) = (a/√(E) + b/E) which extrapolates favorably in the energy range of interest (<2.5% in the range 3–5 MeV) for the observation of 4.44 MeV gamma-rays from the 9Be(α, nγ)12C reaction at kHz counting rates. The limited linearity of the SiPM has been investigated at the National Centre for Nuclear Research (NCBJ, Poland) mainly by measuring the 4.44 MeV gamma rays emitted by a PuBe source and that resulted in three peaks in the pulse height spectrum: a full energy peak (4.438 MeV), a single escape peak (3.927 MeV), and a double escape peak (3.416 MeV). Among the other sources we used are 137Cs and 65Zn (Eγ = 1.116 MeV). Measurements have been performed in a climate chamber at stable temperature to prevent peak shifts due to temperature changes. Results (see Fig. 5) indicate that there is a nonlinear relation between the channel position and the gamma-ray energy, but this is of the order of 20% at 4.5 MeV and can be easily corrected for.
Pulse height spectrum recorded with 137Cs and 60Co sources. Nonlinear calibration has been applied to generate the x-axis.
Pulse height spectrum recorded with 137Cs and 60Co sources. Nonlinear calibration has been applied to generate the x-axis.
Measured energy resolution as function of the energy. Error bars are of the same magnitude of the black dots.
Measured energy resolution as function of the energy. Error bars are of the same magnitude of the black dots.
B. Comparison with a standard photo-multiplier tube
A comparison in terms of energy resolution with a conventional Photo-Multiplier Tube (PMT), manufactured by Hamamatsu (model R9420-100-10), was done. In order to reduce the effect of the geometric efficiency, an aluminum mask with a hole of the same size as the MPPC has been used to cover the PMT surface coupled to the LaBr3 crystal. In this way, the only small difference in the collecting area between the PMT and the MPPC was due to the geometrical Fill Factor (FF) of the MPPC itself, which is the ratio of the effective photosensitive area to total area.13 Results show that energy resolution values of the MPPC operating in the optimal condition are comparable within few fractions of % to those with a PMT having the same collecting area (see Table I). These values are also close to those expected from the statistical fluctuation of the number of photoelectrons generated at each energy and that can be estimated at each energy (see the column labelled as “Calculated”).14
Linearity of the MPPC. Linear trend has been extrapolated by fitting first two points corresponding to energies 0.511 MeV and 0.662 MeV, respectively. Error bars are of the same magnitude of the black dots.
Linearity of the MPPC. Linear trend has been extrapolated by fitting first two points corresponding to energies 0.511 MeV and 0.662 MeV, respectively. Error bars are of the same magnitude of the black dots.
Comparison between the energy resolution of the PMT covered with the aluminum mask and the one of the MPPC. The calculated energy resolution is the one expected from the statistical fluctuation of the number of photoelectrons generated.
Peak energy (keV) . | Measured energy resolution (%) . | Calculated (%) . |
---|---|---|
PMT with mask (collecting area 12 × 12 mm2) | ||
662 | 4.7 | 4.1 |
1173 | 3.8 | 3.1 |
1333 | 3.4 | 2.9 |
MPPC 12 × 12 mm2 | ||
662 | 5.5 | 5.0 |
1173 | 4.0 | 3.8 |
1333 | 3.7 | 3.6 |
Peak energy (keV) . | Measured energy resolution (%) . | Calculated (%) . |
---|---|---|
PMT with mask (collecting area 12 × 12 mm2) | ||
662 | 4.7 | 4.1 |
1173 | 3.8 | 3.1 |
1333 | 3.4 | 2.9 |
MPPC 12 × 12 mm2 | ||
662 | 5.5 | 5.0 |
1173 | 4.0 | 3.8 |
1333 | 3.7 | 3.6 |
JET-like measurement with a perturbation window of 10 s at high counting rate.
IV. LABORATORY TESTS AT HIGH COUNTING RATE WITH A LED PULSE
A. Measurements with LED pulse
In order to simulate the high rate environment expected at JET, a mock up of high counting rate measurements15 has been performed in the laboratory by using a blue LED powered by Keysight (model 81150A) pulse generator. With this setup, we were able to illuminate the MPPC with blue light at a chosen intensity and at increasing counting rates up to about 1 MHz. A 137Cs source was then placed close to the LaBr3 crystal and used to monitor the effect of the LED light (perturbation) on the position and energy resolution of the corresponding full peak at 662 keV from 137Cs. We have observed that the mean position of the 137Cs progressively drifted as the counting rate from the blue LED was increased. This effect was mainly caused by the voltage drop across a resistor placed between the power supply and the MPPC and can be easily reduced by decreasing the value of this resistor, without any appreciable effect on other detector parameters. The peak shift was then observed to increase at higher MPPC bias voltages and at larger intensities of LED perturbation source. Both parameters control the current that flows in the device and that is increased at higher intensities and rates, resulting in an augmented drop on the input resistor. Several measurements allowed investigating and reducing the shift of the peak position in the pulse height spectrum due to the high counting rate. By selecting a sufficiently low resistance (10 Ω) as input, we were able to reduce the shift to 5% at the 137Cs peak, when the LED was operated at 500 kHz and with an equivalent energy of 2 MeV. Part of the contribution to this shift may also come from local heating of the MPPC induced by the larger signal current, as we did not perform our tests in a climatic chamber. This second contribution can however be reduced by adjusting the MPPC gain with a temperature feedback sensor.9
Dedicated tests at counting rates up to few MHz were also successfully performed by using reactions emitting gamma-rays at nuclear accelerators and in a condition that closely mirrored the radiation load expected from deuterium-tritium plasma at full power at JET. The experimental setup and results of these measurements are addressed in Nocente’s article.16
B. JET-like measurements
In order to investigate the behavior of the pilot spectrometer during a JET discharge, a 60 s measurement with two LED pulses has been performed. A LED pulse was used as reference pulse with fix amplitude (corresponding to 3 MeV gamma rays) and fixed repetition rate (10 kHz). The second one, corresponding to a gamma ray of 662 keV, worked as perturbation source in a temporal window of 10 s in which its repetition rate was ramped up to 500 kHz in order to simulate the experimental conditions of a JET shot. The recorded spectrum has been analyzed by sampling the measurement in temporal windows before, during, and after the high rate flux. Results, shown in Fig. 6, highlight a modest shift of 2.4% obtained, thanks to the previous evaluations on the input resistor. We can also assert that this shift can be judged acceptable.
V. LaBr3 INTRINSIC RADIOACTIVITY AS GAIN MONITOR SYSTEM
Due to spatial and cabling constraints, a gain monitor system based on a LED fiber optics cannot be installed on each detector of the gamma ray camera at JET. However, the intrinsic radioactivity of the LaBr3 crystal can be in principle used as a gain monitor to verify the detector stability.12 Fig. 7 shows a spectrum of the detector intrinsic radioactivity integrated over 20 min, which is the average time in between two JET discharges. A clear peak at 1.47 MeV is observed and can be used to monitor peak shifts of few % from one discharge to the next one. Here we also note that the natural radioactivity of LaBr3 does not interfere with peaks expected from fast ions in the plasma (E > 3 MeV), as it does not extend above 2.6 MeV.
Intrinsic activity spectrum of a 25.4 × 16.9 mm2 LaBr3 crystal collected in 20 min of measurement. The x-axis has been generated with the nonlinear calibration.
Intrinsic activity spectrum of a 25.4 × 16.9 mm2 LaBr3 crystal collected in 20 min of measurement. The x-axis has been generated with the nonlinear calibration.
VI. CONCLUSIONS
A dedicated pilot spectrometer based on a LaBr3 scintillator crystal (25.4 × 16.9 mm2) coupled to a Silicon Photo-Multiplier (SiPM) (12 × 12 mm2) has been developed in order to meet the requirements needed for the gamma camera upgrade project. A read-out electronic circuit has been built by implementing a proper CR differentiator able to shorten the output signal width down to 120 ns, allowing for gamma-ray spectroscopy at MHz counting rates and good energy resolution. Laboratory measurements with standard radioactive sources show an energy resolution of about 5.5% at 662 keV at the bias voltage of 67.2 V which extrapolates to <2.5% in the range of interest for plasma diagnostics (3-5 MeV). Mock up measurements at high count rate with LEDs revealed a peak shift in the measured spectra, which has however been minimized by choosing a sufficiently small input resistor for the readout board. The intrinsic radioactivity of the LaBr3 crystal can be used to monitor gain changes in between shots. The overall performance of the system is significantly better than that of CsI detectors of the present gamma ray camera and will allow for improved measurements of the γ-ray emission profile in fast ion experiments of the next JET campaigns and, later on, in high performance deuterium-tritium plasmas.
Acknowledgments
This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 under Grant Agreement No. 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.