Observations of xenon spectra on the EAST x-ray crystal spectrometer for high-temperature plasma diagnostics

The Xe44+ 2.7203 Å line, which has been proposed as one of the diagnostic lines for the x-ray imaging crystal spectrometer on ITER, is observed on the EAST tokamak together with its several satellite lines. The observations are made under high electron temperature (T e ) conditions (core T e > 5 keV). Most of the observed xenon lines are identified by comparing the experiment results with the atomic simulation results. The first ion temperature measurements made by the xenon spectra on EAST are also reported in this article. These xenon spectra observations contribute to the justification for using xenon as the diagnostic impurity in x-ray crystal spectrometers in future reactor-scale high-temperature plasmas.


Introduction
X-ray crystal spectroscopy diagnostics are widely applied on major magnetically confined plasma devices around the world [1][2][3][4][5][6][7]. As one of the plasma diagnostics methods, x-ray crystal * Authors to whom any correspondence should be addressed.
Original Content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. spectroscopy mainly provides ion temperature (T i ) and plasma toroidal rotation velocity (V t ) profiles. X-ray crystal spectrometers (XCS) detect the transition lines of heavy impurity ions (like Ar 16+ ions) and obtain ion temperatures from Doppler broadening of the line spectra and rotation velocities from the wavelength shifts due to the Doppler effect. Ar 16+ ions are presently used as the seeding impurity for XCS diagnostics on EAST [1], but for future reactor-scale tokamak devices, impurity ions with a higher atomic number Z are expected. Xenon (Z = 54) is considered to be suitable to serve as the seeding impurity element for XCS diagnostics on ITER [8][9][10] and CFETR [11]. It has the merit of being able to cover the whole temperature range from the core to the edge of future fusion plasmas using a single element [8,11]. Among all the charge states of xenon, the neon-like state (Xe 44+ ) has a relatively high concentration and has high line transition emissivity within the temperature range of 3-15 keV [8,12]. The Xe 44+ 2.7203 Å , 2p 5 3d (3/2, 5/2) o 1 → 2p 6 1 S 0 transition line, which is also named the Xe 44+ 3D line, is nominated as one of the major diagnostic lines for the ion temperature measurements on ITER [13] and CFETR [11] due to its high intensity.
There have been observations of the Xe 44+ 3D line in other tokamak experiments [14]. In recent years, the Alcator C-Mod team reported an observation of the Xe 44+ 3D line and the satellite lines [15]. To better understand the Xe 44+ 3D line and its satellite lines, comprehensive observation experiments are also made on the poloidal XCS (PXCS) of the EAST tokamak. The latest observations are made during the 2022 campaign, in which the Xe 44+ 3D line and its satellite lines are clearly observed and most of the major transition lines are identified. The first T i results measured by the Xe 44+ 3D line on EAST are reported in this article. This article will introduce the observations in detail. The second part of this article will introduce the system setup, the heating conditions, and the simulation assumptions of these experiments. The third part will present the results of the observations, followed by a discussion and conclusion section.

Experiment setup
The EAST tokamak has implemented two high-resolution XCS systems [16,17]: the toroidal XCS and the PXCS. The PXCS has been upgraded for the Xe 44+ spectra measurement [12]. The PXCS is presently adopting a double-crystal configuration, which means that two slices of Bragg diffraction crystals are assembled side-by-side on a single silicon glass substrate with a spherically bent surface. This configuration enables simultaneous measurements of two different spectra on a single detector if the Bragg angles of the two spectra are close to each other. This method has already been adopted in the previous experiments on the EAST tokamak [1,18]. The current double-crystal assembly of the PXCS is designated to observe the Ar 16+ W line spectra and the Xe 44+ 3D line spectra. The parameters of the two crystals are presented in table 1. However, we filter out the spectra diffracted by the argon crystal and only collect those diffracted by the xenon crystal in the series of observations reported in this article. This is achieved by setting the threshold of the detector, which is installed at the focal plane to detect the patterns of the spectra. The energy of the photons diffracted by the xenon crystal is within the 4525-4580 eV range, while the energy range of photons diffracted by the argon crystal is in the 3080-3150 eV range. Therefore, the photons diffracted by the argon crystal can be filtered out by setting the detector threshold at 4 keV. This modification reduces the noise from the continuous spectra and eliminates the interference from the argon and the tungsten lines, which may appear in the wavelength range of the argon spectra. This modification greatly improves the signal-to-noise ratio of the xenon spectra and enables the observation of the  weaker xenon lines. The detector is a Pilatus3 900 K hybrid CMOS pixelated x-ray detector. The detector has nine modules vertically arranged, and each module has 487 pixels in the horizontal dimension and 195 pixels in the vertical dimension. The pixel size is 172 × 172 µm 2 .
To observe the xenon spectra, the electron temperature (T e ) of the discharge is an important factor. Figure 1 presents the relationship between the impurity line intensities and the T e , which is simulated by the FLYCHK code [19] and the flexible atomic code (FAC) [20]. The simulation results imply that the Xe 44+ 3D line emerges at 3 keV T e and reaches its maximum intensity at 7 keV. The EAST tokamak has already achieved long-pulse high T e discharges (T e = 12 keV in Shot No. (SN) 98 958) in previous campaigns. Therefore, the EAST tokamak is capable of conducting xenon-spectrum observation experiments. During the 2022 campaign, valid xenon spectra observations are made in as many as 18 shots. The basic plasma parameters and the heating conditions of the representative high T e shot (SN 113 711) are shown in table 2 and figure 2. In this series of shots, higher electron cyclotron resonance heating (ECRH) injection powers [21], higher lower hybrid wave (LHW) injection powers [22], and lower electron densities compared to the normal shots are applied. The neutral beam injection (NBI) beams are not active in these shots. This configuration can efficiently achieve a T e as high as 8 keV.  The xenon gas is injected through the equatorial gas valve to the plasmas. For most of the time at EAST, a small trace of the mixture of argon and xenon gases at a partial pressure ratio of argon:xenon = 5:1 is injected at the beginning of each shot to provide the impurities needed for XCS measurements. However, in SN 113 706-113 717 neither mixed gas nor pure xenon gas are injected because the remaining gas from the previous discharges is already sufficient. This also ensures that the discharges reach the high T e range in a stable manner.
To analyze the newly observed xenon lines, the FAC and the FLYCHK codes are used to simulate the patterns of the xenon spectra at a certain T e value. The energy levels (EN tables), the radiative transition rates between the levels (TR tables), and the collisional excitation cross sections (CE tables) of xenon ions with different ionization states (e.g. the Xe 43+ ion, the Xe 44+ ion, etc) are calculated using the FAC code. The concentration distributions of the ionization states under different electron temperatures are calculated using the FLYCHK code under steady-state conditions. Then, the FAC code initializes the ion models with the EN tables, the TR tables, the CE tables, and the concentration of the ionization states. After an iteration process, the ion models converge and provide the results of the energy level populations and the transition intensities between the energy levels. The simulations assume the electron velocity distributions to be Maxwellian and the ionization state distributions to be in equilibrium.

Results and analyses
A comparison of the x-ray spectrum observed in SN 113 711 and that of the simulation is presented in figure 3. The Xe 44+ 3D line (marked as line 7), the Xe 44+ 3 F line (line 5) and the Xe 43+ 3D line (line 1 from that marked line 1-3) are clearly identified. The Xe 44+ 3D line is chosen to be the wavelength reference for the observed spectrum because its theoretical wavelength has already been precisely measured in many other tokamak devices [14,15]. The identified lines in the spectrum are listed in table 3. It is worth noting that the identified lines are observed not only in the SN 113 711 alone, but also in other shots that are not presented in this article.
The line-integrated measured (LIM) T i profiles measured by the Ar 16+ W line and the Xe 44+ 3D line are shown in figure 4. The spectral lines used for the LIM T i fittings are the directly observed integrals over the sightlines. The lines are fitted by the Voigt function. The natural line widths, the crystal rocking curve widths, and the Johann error widths of the two different lines are calculated and corrected. The natural line shapes are treated as Lorentzian line shapes by definition, while the rocking curve line shapes and the Johann error light spots are approximately treated as Gaussian line shapes. The core LIM T i results measured by the Xe 44+ 3D line are significantly higher than those measured by the Ar 16+ W line. This can be explained by the fact that the Ar 16+ ions are likely to be ionized under high T e conditions (T e > 2 keV) and the intensity of the Ar 16+ W line drops as a consequence, as illustrated in figure 1. As the spectra directly observed by the XCS are the integrals over the sightlines, the emission at each part of the plasma contributes to the final observations. As the Ar 16+ ions only exist in the outer plasma regions where the T e is low, the LIM T i at the equatorial plane measured by the Ar 16+ ions does not exactly correspond to the core T i but to the T i in the outer plasma regions at the equatorial plane, which is lower than the real core T i . The Xe 44+ ions, however, only exist in the core region of the plasma where the T e is high enough. Therefore, the LIM T i measured by the Xe 44+ 3D line should be closer to the real core T i of the plasmas.
The discussions above can be further verified by the local emissivity distribution results obtained using an emissivity tomographic inversion method [23][24][25][26]. A comparison of the T e profiles and the Ar 16+ W line emissivity profiles is shown in figure 5. The SN 113 698 is an ordinary low T e discharge (core T e ≈ 3 keV), where the Xe 44+ 3D line is not observed. Both emissivity distributions in the two discharges have a hollow shape, while the profile of the higher T e discharge SN 113 711 has an obviously larger and deeper hollow. Conversely, the Xe 44+ 3D line can only be observed in the center channel in high T e discharges (core T e ≈ 6 keV). This discovery confirms that the 'core' LIM T i results measured   by the Ar 16+ W line are expected to be the T i in the outer regions of the plasmas when the core T e is much higher than 2 keV.
Additionally, the regions where the T e is about 2 keV have the highest emissivity in each of the two discharges according to Figure 5, which is in accord with the simulation results in Figure 5. Comparison of (a) the Te profiles obtained from the Thomson scattering diagnostics [27], and (b) the Ar 16+ W line local emissivity profiles obtained from the XCS observations and processed using the tomographic inversion method at t = 6.0 s. figure 1. Therefore, the reliability of the atomic simulations carried out in this study is also verified.

Discussion
This article introduces the observation experiments of the Xe 44+ 3D line and its satellite lines. By filtering out the spectra diffracted by the argon crystal and by achieving high-T e discharges, we have obtained the clear patterns of the xenon spectra. Most of the major xenon lines have been identified with the aid of the atomic simulation method based on the FAC and the FLYCHK tools. The first T i measurements made by the Xe 44+ 3D line on EAST have been conducted. The core LIM T i measurement discrepancies between the Xe 44+ 3D line and the Ar 16+ W line have been explained by the hollow profiles of the Ar 16+ W line local emissivities. The experiments and the simulations introduced in this article greatly enhance our understanding of the Xe 44+ 3D line and its satellite lines. The use of the Xe 44+ 3D line in the XCS T i measurements compensates for the lack of accuracy in the Ar 16+ W line measurements in high T e discharges on EAST. The practice of using the Xe 44+ 3D line in the T i measurements can also serve as references for the design of XCSs on future reactor-scale tokamaks like ITER and CFETR. Additionally, as there are several spectral lines from different ionization states in the xenon spectra, e.g. line 7 and line 3 in table 3, the measurement of the T e profiles using the xenon spectra is also potentially feasible.
However, there are still lines that have not been explicitly identified at the moment, like line 4 in figure 3 and table 3. Line 4 can be very strong in certain cases, usually when a larger amount of pure xenon gas is injected at the beginning of the discharge. However, its existence has not yet been explained by steady-state atomic simulations. Further studies on the unidentified xenon lines can also be valuable in understanding the heavy impurity transport processes. These studies will be carried out in future works.

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