A new tangential gamma-ray spectrometer for fast ion measurements in deuterium and deuterium–tritium plasmas of the Joint European Torus

Among the scientific goals of the forthcoming deuterium–tritium (DT) campaign of the Joint European Torus (JET),¹ there is the need to increase our understanding of alpha particle physics, mostly in terms of transport and the role that alpha particles have in the excitation of fast ion driven instabilities in view of ITER.² From the diagnostics side, a number of enhancements have been made to improve the alpha particle observational capabilities of JET,³ particularly as far as gamma-ray spectrometry is concerned. Here, the main interest is in the detection of the 4.44 MeV line from the α + ⁹Be → n + ¹²C^* reaction⁴ as this emission provides information on the alpha particle energy distribution, as well as its profile.⁵ 

The main instruments developed for this scope are a gamma-ray camera (GC)^6,7 and two sets of gamma-ray spectrometers (GRSs), one vertical (V-GRS) and one tangential (T-GRS). The GC determines the profile of fast ions in general, including α particles in DT, by detecting gamma-ray emission from characteristic nuclear reactions along a set of horizontal and vertical lines of sight deployed in a fan shaped array and with the use of silicon photo-multipliers (SiPMs).

The V-GRS and T-GRS are instead optimized for spectral measurements and have better energy resolution (3.5% compared with 5% at E_γ = 662 keV for the GC), detection efficiency (by a factor of ≈3), and background suppression capabilities compared with the GC. The GRS detectors also provide a linear conversion between channel and gamma-ray energy within few % up to E_γ = 20 MeV. For comparison, the GC has a >30% non-linearity for E_γ > 10 MeV due to the use of SiPMs. V-GRS allows choosing among different spectrometers, which are selected depending on the experimental conditions. The most used are a High Purity Germanium (HpGe)⁸ and a LaBr₃ detector.^9,10

The T-GRS consists instead of a single spectrometer, which was previously based on a 7.5 × 7.5 cm² (diameter × height) Bismuth Germanate (BGO) scintillator.¹¹ In order to meet the needs of α particle measurements in DT, mostly in terms of energy resolution and counting rate capabilities, the detector of the T-GRS has recently been upgraded to a LaBr₃ scintillator and a dedicated neutron attenuator has been built for its line of sight (LOS). The tangential line of sight of the T-GRS allows distinguishing co- and counter-going particles, which is not possible with vertical geometries. This is crucial for the reconstruction of velocity distributions at JET and ITER.^4,12

In this paper, we describe the properties of the upgraded T-GRS. Section II presents the main parameters of the detector and illustrates its complementarity with the HpGe V-GRS. The T-GRS line of sight is discussed in Sec. III. Finally, examples of measurements from recent fast ion experiments in the JET deuterium plasma are provided and compared with observations made by using the HpGe V-GRS.

The T-GRS is based on a 3 × 6 in.² (diameter × length) cylindrical LaBr₃ scintillator developed by Saint-Gobain (see Fig. 1) and replaces a former 7.5 × 7.5 cm² Bismuth Germanate (BGO) spectrometer.¹¹ The upgrade of the BGO was required predominantly to enable measurements at MHz counting rates, such as those expected from the 14 MeV neutron background in DT plasmas¹³ or in runaway electron experiments in D plasmas. Another need was to distinguish lines separated by ≈100 keV in the MeV range, for example, those arising from radio-frequency (RF) heating in hydrogen plasmas with ³He.¹⁴ Both requirements had to be met without compromising the detection efficiency, which had to be comparable with that of the BGO.

The dimensions of the T-GRS are the same as those of the LaBr₃ based V-GRS. A mu-metal wrapping with a thickness of 2 mm provides shielding from the residual magnetic field at the location of the detector, which sits about 23 m from the center of the plasma on Octant 8. The scintillation light resulting from the interaction of gamma-rays with the detector is collected by using a R10233 Hamamatsu photomultiplier tube (PMT) with 3.5 in. diameter.

Each electric pulse from the PMT is digitized when its amplitude exceeds a threshold set by the user. A 400 MHz, 14 bit, Advanced Telecommunications Computing Architecture (ATCA) based digitization system¹⁵ similar to that adopted for the V-GRS LaBr₃ detector is used for the T-GRS. This records individual waveforms and their time stamp, which tells when each gamma-ray interaction occurred. A fitting algorithm with pile-up rejection capabilities¹⁰ is adopted for off-line processing of the waveforms and to build their pulse height spectrum and counting rate in arbitrary time and energy windows for each JET discharge. A control and monitoring (C&M) system⁹ based on a LED source and a bundle of fibers that guides the light to the detector provides a reference signal, which is used to monitor and correct for possible gain shifts that may occur during the discharge. The energy resolution (expressed as the ratio between the full width at half maximum and the centroid of the peak) measured at JET is 3.5%, 2.4%, and 2.2% at the energies of the gamma-ray emission from the ¹³⁷Cs and ⁶⁰Co calibration sources, i.e., E_γ = 662, 1173, and 1333 keV, respectively. Assuming that the 1/$Eγ$ statistical contribution to the energy resolution is dominant, these values extrapolate to better than 1.5% for E_γ > 2.5 MeV, which makes the detection of lines separated by a few hundreds of keV possible in this energy range. For comparison, the former BGO spectrometer had a typical energy resolution around 10% at 662 keV, i.e., about three times worse than T-GRS, but a better full peak (ϵ_P) detection efficiency of 42% at E_γ = 4.44 MeV, which compares with 24% for LaBr₃. The total detection efficiency (ϵ_T) is similar, 87% for BGO and 93% for LaBr₃. In terms of time properties, the decay time of each pulse from a gamma-ray interaction is about 20 ns in the T-GRS, opening up to measurements at counting rates up to a few MHz,¹⁰ and compares with 300 ns for the BGO.

The T-GRS also provides complementary information when the HpGe detector is used as a V-GRS, particularly when the detection of E_γ > 10 MeV gamma-rays born from fusion reactions^16,17 is of interest (see Sec. IV). Figure 2 compares the total and full peak detection efficiencies for the T-GRS and the HpGe detector over a wide energy range between 1 and 15 MeV. We find that ϵ_T is always larger by at least 50% for the T-GRS compared with the HpGe detector, with ϵ_P being up to a factor of 3 higher in the E_γ > 10 MeV range. The T-GRS is therefore particularly useful to detect high energy, fusion born gamma-rays in discharges where the HpGe detector experiences limitations.

Another important feature of the T-GRS is its tangential line of sight (LOS),¹⁸ shown in Fig. 3, and complementary to that of the V-GRS. The LOS has a double pass through the magnetic axis and crosses the plasma center ≈30 cm below the magnetic axis for most discharges. The angle between the line of sight and the toroidal magnetic field in the plasma core is ≈50°.

As LaBr₃ is sensitive to the interaction of 14 MeV neutrons,¹³ a carefully studied attenuator based on LiH has been designed and built for the T-GRS LOS. The design has been driven by Monte Carlo N-Particle (MCNP) simulations, with additional constraints provided by existing mechanical components along the LOS. The LiH attenuator is made of two fixed disks, with a total length of 66 cm, and a movable one, which is 30 cm long. When the full length of the attenuator is used, corresponding to a total length of 96 cm, 14 MeV DT fusion born neutrons are attenuated by a factor of ≈275 according to MCNP calculations,¹⁸ while MeV range gamma-rays are reduced by a factor of ≈2–3 at most, making gamma-ray measurements possible up to neutron rates of some 10¹⁸ n/s. Without LiH, the detector signal would be overwhelmed by 14 MeV neutrons already at rates of a few 10¹⁶ n/s, which correspond to a high performance deuterium plasma at JET. The attenuation provided for 2.5 MeV neutrons from d + d reactions is ≈900. A detailed description of the attenuator design and engineering is given in Ref. 18. The choice of LiH as the attenuator material is motivated also by the fact that neutrons are unlikely to generate additional gamma-rays when they interact in this material. This is predominantly due to the absence of ¹²C, which is abundant, for instance, in the more conventional polyethylene, and which generates 4.44 MeV gamma-rays due to n + ¹²C inelastic scattering.

The availability of a tangential LOS is important for experiments on radio-frequency (RF) heating, especially for off-axis resonances. In this case, fast ion generation occurs away from the plasma core and the resulting gamma-ray emission cannot be measured with the V-GRS. Differences between the spectra observed with the T-GRS and V-GRS can therefore help in understanding the spatial position of the fast ion generation and complement more detailed information coming from the JET gamma-ray camera (GC). This will be useful particularly in DT, when the GC will experience significant limitations at neutron rates >10¹⁷ n/s, mostly because of the insufficient neutron shielding of this instrument.¹⁹ In terms of fast ion velocity-space capability, the LOS of the T-GRS and V-GRS is also complementary. As shown in Ref. 20, the high and low energy tails of the shape of gamma-ray lines measured along the T-GRS LOS is mostly determined by co- and counter-passing energetic ions, while the contribution of trapped fast ions in the same spectral region is larger for lines measured with the V-GRS. The combination of the T-GRS and V-GRS thus makes it possible to separate the dynamics of passing and trapped fast ions by line spectral shape analysis, at least for those scenarios when the fast ion Doppler broadening of the gamma-ray line compares or exceeds the broadening from the instrumental resolution, for example, for the α + ⁹Be → n + ¹²C^* reaction in DT plasmas or, more generally, when the fast ion populations have predominantly MeV range energies. The integrated data analysis of T-GRS and V-GRS could even allow measurements of alpha-particle velocity distribution functions by velocity-space tomography.¹² 

The T-GRS has been used in the most recent deuterium campaign at JET and has provided useful information, particularly in experiments on fast ion generation by RF heating. Figure 4 shows an example of measurements from a recent experiment on the so-called “3 ion scenarios”^14,21 in a plasma of deuterium and ³He at relative concentrations of 50% and 25%, respectively.²² In this experiment, RF heating accelerates a beam of deuterium ions up to the MeV range in a narrow region around the plasma core. The deuterium energy distribution is controlled by the ratio between the neutral beam injection (NBI) and the RF power and modifies the reactivity of the d + ³He → α + p fusion reaction. In this way, fusion born α particles are generated at rates up to ≈2 × 10¹⁶ α/s and can be studied without the difficulties of a full scale DT experiment, particularly those arising from the intense 14 MeV neutron background. As both deuterons and α particles are generated in the MeV range, we expect to observe lines from the d + ⁹Be → n + ¹⁰B^*, d + ⁹Be → p + ¹⁰Be^*, and α + ⁹Be → n + ¹²C^* gamma-ray emission reactions. In the experiment, the HpGe detector was chosen as the V-GRS since the emission from the plasma was sufficiently intense to obtain a good measurement statistics when integrating data for 5 s. A comparison of the V-GRS and T-GRS spectra for JET discharge No. 95680 in the time window t = 47.5–52.5 s is shown in Fig. 4.

Most of the lines seen by the HpGe detector are also observed by the T-GRS. As expected, the HpGe detector has a superior energy resolution, but the statistics is up to three times better for the T-GRS. This is because of the higher T-GRS detection efficiency and the fact that the T-GRS line of sight crosses the plasma center twice at a comparable distance of the detector from the plasma center (≈23 m). Some differences are seen between the two spectra. For instance, the 4.44 MeV emission from the α + ⁹Be → n + ¹²C^* reaction is observed by the HpGe detector, but not by the T-GRS. This is because, in this experiment, α particle generation occurs predominantly in an ≈10 cm circular region around the magnetic axis, as suggested by the GC,²³ and this area is barely intercepted by the LOS of the T-GRS. Another difference is the E_γ = 3586 keV line from the d + ⁹Be → n + ¹⁰B^* reaction, which is hardly seen by the HpGe detector because of its lower detection efficiency. Despite this, using the HpGe detector in the vertical LOS makes clear gamma-ray Doppler broadening measurements possible for many lines, and the HpGe/T-GRS combination is therefore, generally, the preferred option when gamma-ray fluxes and the neutron background allow. The use of LaBr₃ as the V-GRS is available for all other scenarios.

Figure 5 shows the same T-GRS spectrum of Fig. 4 but in logarithmic scale and in an energy range up to E_γ = 20 MeV. We note that, besides the already mentioned lines from the d + ⁹Be → n + ¹⁰B^* and d + ⁹Be → p + ¹⁰Be^* reaction in the range E_γ = 1–5 MeV, there is a clear structure between E_γ = 10 and 20 MeV and that exceeds the background in this region. This is due to 16.4 MeV gamma-rays born from the d + ³He → γ + ⁵Li reaction, which occurs together with d + ³He → α + p, but with a lower probability of ≈10⁻⁵. d + ³He → γ + ⁵Li is of interest to determine the fusion power for aneutronic fusion reactions, such as d + ³He → α + p.^17,24 The broad shape of the 16.4 MeV gamma-ray interaction with the T-GRS is a combination of the intrinsic broadening of the emission at the source and the Compton effect dominated response function of the LaBr₃ scintillator at these energies.¹⁷ 16.4 MeV gamma-rays are barely detected by the HpGe detector due to the low detection efficiency of this instrument in the E_γ > 10 MeV range (see Fig. 2), which reveals the need for the T-GRS also for fusion power measurements, particularly when the HpGe detector is chosen as the V-GRS for gamma-ray Doppler broadening measurements.

A tangential gamma-ray spectrometer (T-GRS) has been developed for the Joint European Torus (JET). The detector is based on the LaBr₃ spectrometer and observes the plasma along a tangential line of sight (LOS) that passes ≈30 cm below the magnetic axis in the poloidal plane. A data acquisition system digitizes individual pulses from the detector and above a threshold set by the user. A pulse height spectrum in arbitrary energy and time windows is built after each discharge. The instrument LOS is equipped with a LiH attenuator, which makes gamma-ray measurements possible in deuterium–tritium plasmas at neutron rates up to a few 10¹⁸ n/s and also provides a neutron attenuation factor up to a factor of ≈900 in deuterium. Measurements from recent deuterium experiments at JET show that the T-GRS can resolve lines from fast ion reactions separated by a few hundreds of keV in the MeV range and can detect the E_γ > 10 MeV emission from fusion reactions. Due to its line of sight, the T-GRS provides complementary information to those obtained from the vertical GRS and the JET gamma-ray cameras, in both the position and velocity space. The instrument is expected to be one of the main diagnostics for studies of confined alpha particles in the forthcoming JET deuterium–tritium campaign.

This work was carried out within the framework of the EUROfusion Consortium and received funding from the Euratom Research and Training Programme 2014–2018 and 2019–2020 under Grant Agreement No. 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission. This scientific work was also supported by the Polish Ministry of Science and Higher Education within the framework of the scientific financial resources in the years 2015–2020 allocated for the realization of the international co-financed project.

The data that support the findings of this study are available from the corresponding author upon reasonable request.