ICRF production of plasma with hydrogen minority in Uragan-2M stellarator by two-strap antenna

The research of the high temperature plasma for the controlled fusion is essential for the development of magnetic fusion power plant.¹ One of the missions of the EUROfusion consortium's European research roadmap for the realization of fusion energy is the development of stellarator approach to fusion as a possible long-term alternative to tokamaks.² The main focus is posed on optimized stellarator Helical-Axis Advanced Stellarator Helias (HELIAS) line. The first HELIAS machine is the superconducting stellarator device Wendelstein 7-X (W7-X) in Greifswald, Germany.^3,4 The main objective of the research at W7-X is to demonstrate the possibility of obtaining in the stellarator plasma with parameters that could be extrapolated to fusion power plants. Two other medium-sized stellarators, TJ-II in Madrid, Spain,^5,6 and Uragan-2 M (U-2M) in Kharkiv, Ukraine,^7,8 support the W7-X experimental program within the EUROfusion consortium.

The W7-X stellator is equipped with three different plasma heating systems.^9,10 The electron cyclotron resonance heating (ECRH) system at 140 GHz^9–13 is the main system for producing and heating plasma at the second harmonic of the ECR [X2 mode and O2 mode, extraordinary (X) and ordinary (O) modes, respectively] in a magnetic field of 2.5 T.^14,15 The Neutral Beam Injection (NBI) system^9,10,16–18 can be used both together with ECRH and independently to support and heat the plasma produced by ECRH.¹⁸ The Ion Cyclotron Resonance Heating (ICRH) system^9,10,19–22 has several purposes: fast-ion generation, ion heating, wall conditioning, and providing plasma start-up.^19–22 It consists of two generators in the 25–38 MHz frequency range and a two-strap antenna. The ICRH system has recently been installed on W7-X, and the first tests and experiments have been performed during the last campaign, OP2.1.

Experiments on W7-X in a magnetic field of 1.7 T are of interest for the study of fusion-relevant plasmas with high β-values. In this magnetic field, for a fixed ECRH frequency of the system, plasma heating at the 3rd harmonic ECR frequency (X3 mode) is possible.^11,23 However, as calculations²³ show, successful plasma heating in W7-X at X3 mode ECRH requires an initial target plasma with temperature T_e > 700 eV and density N_e > 10¹⁹ m⁻³. Without target plasma at X3 mode, ECRH start-up is not possible that was observed experimentally in Ref. 18. In Large Helical Device (LHD) plasma heating experiments at the X3-mode ECRH, the target plasma was produced by NBI.^24,25 As shown by the modeling performed in Ref. 26, NBI at W7-X keeps its efficiency in reduced magnetic fields down to 0.75 T. The modeling of plasma start-up by NBI in hydrogen, helium, and their mixture for W7-X was performed in Ref. 27. The calculations showed that the delay between NBI start-up and full ionization of the neutral gas is about 5 s. This time is too long and leads to an unacceptably high thermal load on the beam receiving plates.¹⁶ Experiments on NBI start-up on W7-X have shown that no plasma is created for durations up to 0.7 s NBI, and start-up in the combined mode NBI + X3 was unsuccessful in a magnetic field of 1.7 T.¹⁸ According to calculations,²⁷ a target plasma with a density of 10¹⁶–10¹⁷ m⁻³ is required to reduce the start-up time by NBI to less than 0.5 s. Accordingly, experimental studies and theory show that even the generation of such low density target plasma in W7-X at 1.7 T remains an unsolved problem. On the Heliotron J helical device, the target plasma for NBI was created using non-resonant microwave discharge at 2.45 GHz.^28–31 In the Compact Helical System (CHS), the target plasma was created by radio frequency (RF) discharge for NBI.³² The ICRH system at W7-X can be used to create a target plasma.^19–22 Two scenarios are considered. In the first scenario, the plasma is created by ICRH discharge, which parameters are increased by NBI to the above-mentioned level required for start-up using X3-mode ECRH (ICRH + NBI + X3 ECRH). In the second scenario, ICRH is used more efficiently, and it creates plasma with the required parameters for further heating on X3-mode ECRH, without NBI (ICRH + X3 ECRH). In both scenarios, plasma generation by ICRH is the main focus. Studies of plasma creation and heating in the ion cyclotron frequency range (ICRF) in tokamaks and stellarators have a long history.^33–38 They have been carried out in pure hydrogen, deuterium, and helium gases, in the minor regime and in mixtures of these gases. ICRF discharges for wall conditioning in helium have been studied in the tokamaks TEXTOR-94,³⁹ Tore Supra,⁴⁰ HT-7,⁴¹ experimental advanced superconducting tokamak (EAST),⁴² and JET⁴³ and in the stellarator W7-AS.⁴⁴ In the experiments on TEXTOR-94, a maximum plasma density of ≈10¹⁸ m⁻³ was achieved, and increasing the RF power did not significantly change the density.³⁹ ICRF plasma start-up in helium has been investigated on the U-2M stellarator.⁷ A maximum plasma density of ≈10¹⁸ m⁻³ was observed near the fundamental harmonic of the hydrogen ion cyclotron resonance (ICR).⁷ A minor amount of hydrogen in helium was present in these experiments.⁷ Comparison of ICRF discharges in hydrogen and helium at JET showed that a higher plasma density is observed in helium ICRF discharge with hydrogen minority than in pure hydrogen ICRF discharge.⁴³ At the same time, glow in the ICR zone for hydrogen was observed in the helium ICRF discharge.⁴³ 

ICRF discharges in H₂ + He mixtures have also been used for wall conditioning in the tokamaks TEXTOR and ASDEX Upgrade.^45–53 In these experiments, plasma densities in the range of 10¹⁶–5 × 10¹⁷ m⁻³ were observed.^50–53 The plasma of ICRF discharges in H₂ + He mixtures has been studied in Sirius⁵⁴ and H-1NF (H-1) stellarators.^55–59 In H-1NF experiments, densities up to ≈2 × 10¹⁸ m⁻³ and electron temperatures of several tens of eV have been observed with injected RF power up to 80 kW.⁵⁹ The plasma density in ICRF discharges in H₂ + He mixture was slightly higher than that in pure hydrogen.^55,56 Note that plasma production by ICRF discharges in pure gases has also been studied previously in the CHS,⁶⁰ W7-AS,⁶¹ U-3M and U-2M,^62,63 and LHD⁶⁴ stellarators.

In support of the ICRF experiments at W7-X on the U-2M stellarator, studies of ICRF plasma production by discharge and the development of an ICRF start-up scenario that could be realized in a 1.7 T magnetic field at W7-X were initiated.⁷ For these experiments, a two-strap antenna mimicking W7-X antenna was installed on the U-2M.^7,8,65 The first experiments were performed in helium with an minority of hydrogen supplied by the impurities in an uncontrolled way.⁷ Experiments with controlled minority of hydrogen in helium showed a threefold increase in plasma density compared to pure helium.⁶⁶ Further experiments in the H₂ + He mixture showed that a fully ionized plasma with a density of ≈10¹⁹ m⁻³ is observed when the heating frequency is near the fundamental ICR harmonic for hydrogen.⁶⁷ Comparison of the plasma density produced by ICRF discharge in hydrogen, helium, and H₂ + He mixture showed that the plasma density in H₂ +He mixture is higher than in hydrogen and helium.^8,68 Studies of plasma production in He + H₂ + D₂ and D₂ + H₂ mixtures have also been carried out.⁸ A necessary condition for the realization of this scenario is the presence of minor addition of hydrogen and the ICR zone for hydrogen in the plasma column.^7,8,66–68 The size of the U-2M stellarator and the magnitudes of the magnetic fields are significantly smaller than W7-X and LHD. Then, the ICRF plasma production scenario developed at the U-2M was realized at the bigger machine, LHD.⁶⁹ Successful experiments at the LHD showed that the ICRF start-up scenario can be scaled up to large stellarator devices.^68–70 It is possible to produce dense plasmas up to 10¹⁹ m⁻³ in density and up to 2.5 keV in electron temperature.⁷¹ 

The initial results on ICRF plasma production are promising, and more studies on the ICRF start-up scenario are needed to provide a more complete physical picture. When planning these studies, it should be accounted for that the experimental time on large helical devices such as the W7-X and LHD is costly. Medium-sized devices are more flexible in this regard.

The paper is a continuation of earlier research on the ICRF plasma production in a H₂ + He mixture at Uragan-2M. Section II describes the device, the diagnostics, and the experimental conditions. Section III describes the propagation of electromagnetic waves in the given experimental conditions. Section IV discusses the experimental results of ICRF plasma production. Section V summarizes the obtained results.

The experiments were carried out in medium-size stellarator of the torsatron type Uragan-2M (U-2M) at Kharkiv, Ukraine.^7,8,72–74 The U-2M top view sketch is shown in Fig. 1(a). It is a device with a major radius of 1.7 m and a minor radius of the vacuum chamber 0.34 m. The magnetic system consists of two (l = 2) helical coils (windings) with four periods (m = 4) in the toroidal direction; 16 circular toroidal field coils evenly distributed along the torus and eight poloidal field coils. The toroidal magnetic field at the toroidal axis at present is limited by power supply B₀ < 0.6 T (the maximum magnetic field is B₀ = 2.4 T).

The toroidal vacuum vessel with a volume of ≈4 m³ was pumped out by turbo-molecular pumps to a pressure of about 1 × 10⁻⁵Pa. The working gases hydrogen, helium, or their mixtures were supplied into the vacuum chamber by the single-pass gas puff system. The creation of necessary gas H₂ + He mixture composition for the experiments at U-2M was carried out in the gas mixture system⁷⁵ prior to supply to the gas puff system. The percentage of He and H₂ in the mixture was measured using a mass-spectrometer in vacuum chamber U-2M.

A two-strap antenna^7,8,65 mimicking the W7-X ICRH antenna was used to produce the plasma.^19–22 The antenna is installed in the cross section R1 [see Figs. 1(a) and 1(b)]. The antenna was connected to a RF generator system “Kaskad.”⁸ The RF power is calculated using the measured amplitudes of the forward and backward waves in the feeder system.

The time dependence of the average electron density was measured a microwave interferometer in the cross section R [see Fig. 1(a)].⁷⁶ The time-resolved optical emission spectroscopy measures intensities of the spectral lines of hydrogen, helium, carbon, and oxygen in three cross sections P1, Z4, and V [see Fig. 1(a)]. In cross section P1, the chord distributions intensities and the spectral lines are measured using a shot-by-shot technique. Optical diagnostic methods were used to determine the temperature of the electrons and ions.⁷⁷ The ion temperature was measured from the broadening of the spectral lines of helium ions (He II 468.6 nm) and carbon ions (C II 426.7 nm).^77,78 These measurements were taken along the central chord in cross section P1. The average electron temperature was determined from spectral line intensity ratios of He I (471.3 nm) and He I (504.8 nm).^77,79 Measurements were taken along the center chord in cross section Z4. Signals from all systems (magnetic, RF, plasma diagnostics, etc.) were collected by the data acquisition systems.⁸ 

Electromagnetic waves in the ICRF range cannot propagate without plasma in the toroidal chambers of modern tokamaks and stellarators such as JET, W7-X, LHD, and U-2M.⁸⁰ Therefore, at the initial stage of plasma creation by RF discharges, breakdown and ionization of the neutral gas occur due to electrons accelerated in the near electric field of the RF antenna.^80–82 The increase in plasma density opens a possibility for the propagation of waves in the plasma. At the same time, RF power is absorbed by the plasma, which provides further heating of electrons and increase in plasma density.

Expression (6) defines the perpendicular refractive indexes, and sign (+) should be chosen for the slow wave (SW) and sign (−) for the fast wave (FW).

Equations (2) and (3) and Eqs. (5) and (6) were used to calculate the squared perpendicular wave numbers of SW and FW as functions of the plasma density were performed. For the calculations, the parameters corresponding to the U-2M experiments^7,8,66–68 were taken the value of the magnetic field in the center B₀ = 0.35 T, the frequency 4.9 MHz, and the antenna current spectrum for monopole phasing of Ref. 8. The results of calculations for helium, hydrogen, and their mixture are presented in Figs. 2 and 3. Positive values of lg(|k²_⊥|)sign(k²_⊥) correspond to propagating waves. The values of lg(|k²_⊥|)sing(k²_⊥) less than 0 belong to the cutoff regions.

In the case of pure helium [see Fig. 2(a)], a similar situation is observed for wave propagation in the plasma under the condition ω_RF > ω_ci, as had been discussed previously in, for example, Refs. 85, 86, and 88. In this case, there are two critical points for SW [see Fig. 2(a), points 1 and 2], which determine by the critical density values above or below which SW cannot propagate.⁸⁰ A lower density threshold [see Fig. 2(a), point 1] is defined for SW under the condition k_⊥,SW = 0, for which SW propagation is possible (|N_ǁ| > 1) for higher plasma densities. In this case, the value of the critical density can be estimated from the relationship ω_RF² = ω_pe² + ω_pi², where ω_pe and ω_pi are the electron and ion plasma frequencies, respectively.⁸⁰ The upper limit of the plasma density [see Fig. 2(a), point 2] above which SW cannot propagate is determined by the lower hybrid resonance condition, k²_⊥,SW → ∞, which occurs when ω_RF² = ω_LH².⁸⁰ 

For FW, there is also a cutoff point [see Fig. 2(a), point 3] under the condition k_⊥,FW = 0. The value of the critical plasma density above which FW propagation is possible can be estimated from the relation ω_pi² = (N_ǁ² − 1)ω_ci (ω_RF + ω_ci).⁸⁰ In this case, the value of the critical density for FW depends on N_ǁ and accordingly on k_ǁ. These quantities are determined by the antenna spectrum.

In hydrogen [see Fig. 2(c)], ω_RF ≈ ω_ci, and the density dependence of the squared perpendicular wave number is different from that in helium [see Fig. 2(a)]. At low plasma densities, the SW propagates from the density determined by the SW cutoff [see Fig. 2(c), point 1]. As the plasma density increases, when the density above the cutoff point FW [see Fig. 2(c), point 3] is reached, FW begins to propagate. At the same time, SW also propagates. Further increase in the plasma density leads to the FW conversion to SW at the conversion point (Alfvén resonance).^7,87

In the He + H mixture [see Fig. 2(b)], a similar situation is observed as for hydrogen [see Fig. 2(c)]. The main difference is the shift of the cutoff point FW and conversion point to the region of higher density than in hydrogen at the same values of k_ǁ. As can be seen from Fig. 3, with increasing values of k_ǁ, cutoff point FW and conversion point shift to the higher density region. Thus, in the He + H mixture, it is possible to propagate waves in plasma with higher density than in hydrogen under the same initial conditions. The use of He + H mixture makes it possible to produce a plasma with higher density.

The ICRF production scenario was similar to the previous studies in Refs. 7, 8, and 66–68. The H₂ + He gas mixture with the required concentration of hydrogen in the mixture was prepared before the experiments started.⁷⁵ Then, the H₂ + He mixture was continuously injected into the U-2M vacuum chamber up to the pressure required for the experiments. A two-strap antenna^7,8,65 operated in monopole phasing was used to produce the plasma. To control RF power, the RF generator anode voltage U_a is varied stepwise as follows: U_a1 ≈ 0.4 U_a at the start, U_a2 ≈ 0.6 U_a at step 1, and the maximum anode voltage U_a was set at step 2. At an anode voltage U_a ≈ 7 kV, the RF power was ≈100 kW. The RF frequency was equal to the fundamental hydrogen cyclotron harmonic ω_RF ≈ ω_ci (H⁺) ≈ 2ω_ci (He²⁺) ≈ 4ω_ci (He⁺). In a number of experiments described later, the RF frequency was the same, but the magnetic field is twice lower resulting in ω_RF ≈ 2ω_ci (H⁺).

As an example, Fig. 4 shows the typical dynamics of the plasma density and the intensity of the spectral lines of hydrogen and helium in an ICRF discharge. Several stages of plasma production process could be conventionally distinguished. The first stage is the RF breakdown and the formation of a low density plasma N_e < 10¹⁸ m⁻³. This stage is characterized by the beginning of increase in plasma density and intensity of spectral lines H I and He I of hydrogen and helium atoms (see Fig. 4). In the second stage, the plasma density is increased to the maximum value. In this case, the maximum achievable value of the plasma density depends on the initial pressure of the gas mixture [see Figs. 4(a) and 4(b)]. At an initial pressure of 8.2 × 10⁻³Pa, a plasma density of ≈8 × 10¹⁸ m⁻³ is detected [see Fig. 4(a)], and at a pressure of 2 × 10⁻³Pa, the maximum density is ≈4.7 × 10¹⁸ m⁻³ [see Fig. 4(b)] or ≈1.7 times lower. At the stage of increasing plasma density, the intensity of the He II spectral line at lower pressure [see Fig. 4(e)] is ≈4.8 times higher than at 8.2 × 10⁻³Pa [see Fig. 4(b)]. This difference can be explained by the higher electron temperature at lower pressure, which will be discussed in more detail in Subsection IV B. The intensity of the spectral line depends on the excited state density of atoms (ions) in the plasma and the spontaneous emission probability.⁸⁹ Assuming that the excitation of atoms (ions) occurs from the ground state, the excited state density of atoms (ions) depends on the ground state density, the electron density, and the excitation rate coefficient, which depends on the electron temperature. The ionization and excitation processes have a threshold energy, below which these processes do not occur. The ionization energy for the helium atom is ≈24.58 eV, and the upper level excitation energy is ≈51 eV. Excitation of this level results in emission of the 468.6 nm line of the helium ion.⁹⁰ Accordingly, an increase in the electron temperature should lead to an increase in the helium ion density and the probability of ion excitation, resulting in an increase in the intensity of the spectral line. Furthermore, with increasing plasma density, the intensity of the He II line decreases at both pressures [see Figs. 4(b) and 4(e)]. This is apparently due to the decrease in electron temperature and the corresponding decrease in the probability of ion excitation. After the RF pulse is turned off for 25 ms, the plasma begins to fade. The intensity of the spectral lines decreases rapidly.

In the experiments on plasma production in the He + H₂ mixture by ICRF discharge, the plasma density depended on the hydrogen concentration in the mixture, the initial pressure of the mixture, the injected RF power, and the magnetic field value at a constant RF frequency. In these experiments, as well as in the previous ones,^7,8,66–68 plasma with a density greater than 10¹⁸ m⁻³ was produced when the RF frequency was close to the hydrogen cyclotron harmonic frequency. To illustrate these results, Fig. 5 shows the dependence of the average plasma density on the magnetic field at a constant value of the RF frequency. It can be seen that the maximum plasma density is observed in the range of ω_RF ≈ ω_ci (H⁺). Increasing or decreasing the value of the magnetic field leads to a significant decrease in the plasma density by an order of magnitude (see Fig. 5). Further increase or decrease in the magnetic field value led to the impossibility to produce plasma even with low density higher than ≈10¹⁷ m⁻³. The production of more dense plasma of ≈10¹⁸ m⁻³ is observed in the range of ω_RF ≈ 2ω_ci (H⁺). However, the maximum plasma density is ≈2.8 times lower than in the case of plasma production at ω_RF ≈ ω_ci (H⁺). As in the case of plasma production at ω_RF ≈ ω_ci (H⁺), the plasma density decreases with increasing or decreasing magnetic field for the case of ω_RF ≈ 2ω_ci (H⁺) (see Fig. 5).

Another specific of the ICRF plasma production by discharge in He + H₂ mixture is the dependence of the plasma density not only on the pressure but also on the hydrogen concentration.^8,66–68 Figure 6 summarizes the data on the average plasma density obtained in this series of experiments and earlier in Refs. 8, 67, and 68. The injected RF power of ∼100 kW was practically the same in all these experiments. As can be seen from Fig. 6, in the range from 4% to 75% of hydrogen concentration in the mixture, the plasma density is several times higher than in pure hydrogen and helium. This may be due to the fact that in the plasma of the He + H₂ mixture, there is an effective mechanism of energy transfer from the RF wave to the plasma electrons. These mechanisms have been discussed in Subsection IV C and earlier in Ref. 7. As a result, there is a significant increase in the density and degree of ionization of the plasma. Note that in the case of helium plasma, a small addition of light ions, in this case an initial concentration of 4% hydrogen [see Fig. 6(g)], significantly changes the maximum achievable value of the plasma density as well as the pressure dependence of the density compared to pure helium [see Fig. 6(h)].

At the same time, in the hydrogen plasma, the addition of heavy ions, with an initial concentration of 25% helium [see Fig. 6(b)], also increases the maximum achievable plasma density, but the pressure dependence of the density does not change significantly [see Fig. 6(a)]. The plasma density can vary significantly with the pressure of the He + H₂ mixture. For example, at a hydrogen concentration of 26%, the plasma density varies with pressure up to ∼5.3 times [see Fig. 6(d). For comparison, in hydrogen and helium, the maximum plasma density varies weakly with pressure by a factor of less than ∼2 [see Figs. 6(a) and 6(h)], as found in Refs. 8 and 68. The pressure range where it is possible to generate plasma with density higher than 10¹⁸ m⁻³ is wider in the He + H₂ mixture than in pure hydrogen and helium. Moreover, the pressure range where the maximum density is reached is observed at hydrogen concentration from 4% to 49% [see Figs. 6(c)–6(g)].

An increase in the injected RF power, which, in this case, is proportional to the anode voltage at the RF generator, leads to an increase in the plasma density, as can be seen from Fig. 7. The dependence has a linear character, which has already been observed in Refs. 8 and 66. A similar dependence has been observed in LHD experiments.⁶⁸ Experiments on U-2M and LHD have shown that RF power density at the level or more than 60 kW/m³ is necessary to produce a plasma with relatively high density ≈10¹⁹ m⁻³.⁶⁸ In the case of plasma production in helium, a significant increase in RF power did not significantly change the plasma density.^7,66 A similar situation when producing ICRF plasma in helium was observed earlier in the TEXTOR-94 tokamak.³⁹ 

Figure 8 shows the dependence of the average electron temperature for helium plasma and He + H₂ mixture on the initial pressure value. The electron temperature was estimated from spectroscopic measurements of the helium line intensity ratio along the central chord. They characterize the temperature of the electrons at the edge of the plasma. It can be seen that in the case of pure helium [see Fig. 8(a)] and the helium–hydrogen mixture [see Fig. 8(b)], the dependence has a general trend.

Electron temperature increases as gas pressure decreases. Accordingly, an increase in pressure results in an increase in the frequency of collisions of electrons with neutral atoms and molecules, which leads to an increase in the loss of electron energy mainly in inelastic collisions. Including to electron energy losses at excitation of rotationally vibrational levels of the molecule. With decreasing pressure and increasing T_e, the intensity of the spectral line of the helium ion increases (see Fig. 4). Although the temperature of electrons increases with decreasing pressure, it remains rather low. Note that the maximum T_e is observed at the initial stage of He + H₂ plasma creation up to ≈40 eV. Further in time, the temperature decreases, and at 20 ms, it is up to ≈15 eV in helium [see Fig. 8(a)] and up to ≈22 eV in the He + H₂ mixture [see Fig. 8(b)]. In previous experiments on the production of ICRF plasmas in the He + H₂ mixture at U-2M, low values of T_e ∼ 10–15 eV were also observed.⁶⁷ Note that non-high values of the electron temperature up to ∼10–40 eV have also been observed in the production of RF plasmas in the He + H₂ mixture on the H-1NF stellarator.^59,91 Although the temperature of the electrons was low in the experiments on ICRF plasma production in He + H₂ mixture on U-2M as well as in the first experiments on LHD,^68–70 plasma heating up to high temperatures is possible with high power, as shown in the last experiments on LHD.⁷¹ 

The ion temperature is also observed to be low in these experiments. Figure 9 shows the time dependence of the temperature of helium and carbon ions. An increase in T_i with time is observed. The maximum temperature for He⁺ ions was T_i ≈ 18 eV, and for C⁺ ions, it was T_i ≈ 23 eV. Despite the fact that the initial hydrogen concentration in these experiments was different, the values of T_i were close. Note that in experiments with He + H₂ at U-2M, fluxes of charge exchange neutrals with energies in the hundreds of electron volt to a few kilo electron volts are observed.⁹² 

The chord measurements in a shot-by-shot regime for the optical lines excited atoms H I and He I and excited ions He II, C II, and C III were carried. A poloidal scan of integral emission optical lines was made using 41 chords with uniformly distributed impact parameters, P, the vertical positions of the viewing line over the toroidal axis (see Fig. 10). The operating regime of the RF generator was as follows: 10 ms (start), 12 ms (step-1), 14 ms (step-2), and 25 ms (shutdown).

Figure 11 shows the chordal distribution of the intensity of the spectral lines emitted by the RF discharge in helium gas mixture. As can be seen from Fig. 11, the emission source has the same dimensions as the plasma column (|P| < 15 cm, see Fig. 10), and the intensity distribution has a certain symmetry. The hydrogen spectral line appears first. The maximum intensity is reached at ∼15 ms and is observed in the region P ≈ (7.5) − (−10) cm [see Fig. 11(a)]. Then, with a small delay with respect to hydrogen, the intensities of the spectral lines of the excited He I atom and the He II ion appear successively. The maximum intensity for He I is observed in the region P ≈ (7.5) − (−7.5) cm [see Fig. 11(b)] and for the He II ion in the region P ≈ (5) − (−7.5) cm [see Fig. 11(c)]. The observed sequence of appearance of the spectral lines for hydrogen and helium is due to the different excitation energies of the energy levels in hydrogen, helium, and helium atoms, as well as the excitation and ionization cross sections, which are discussed in Refs. 7 and 66. The excitation energy of the upper level for H_β 486.1 nm is ≈12.7 eV, and for the He I 447.1 nm and He II 468.6 nm lines, the energies are ≈23.7 eV and ≈51 eV, respectively.⁹⁰ The intensity of these spectral lines decreases with time, as can be seen in Fig. 4. The observed picture of the chordal distribution of the intensity of the spectral lines is similar to that one observed previously in Ref. 67 for plasma in a mixture of helium and hydrogen at pressure p ≈ 1 × 10⁻²Pa.

Note that for the RF discharge in pure helium, the chordal distributions for He I and He II⁷ differ significantly from the results obtained in this paper and in Ref. 67. The chordal distributions of line intensities in helium plasma were significantly non-symmetric.⁷ 

The chord distributions for carbon (see Fig. 12) differ significantly from those ones for hydrogen and helium. The appearance of the carbon lines is delayed with respect to the appearance of the hydrogen and helium lines. The maximum C II intensity is reached at ∼17.5 ms and is observed in the region P ≈ (10) − (−12.5) cm [see Fig. 12(a)]. At ∼20 ms, the maximum intensity for C III is reached in the region P ≈ (7.5) − (−10) cm [see Fig. 12(b)].

The maximum intensity for C II and C III is observed before the end of the ICRF discharge. In this case, carbon enters the plasma due to the interaction of the plasma with the surface of the vacuum chamber and release of the impurities. Accordingly, the appearance of excited carbon ions will occur with some delay after the beginning of plasma creation. The time of which will be determined by the time of flight from the wall to the plasma and subsequent ionization. The ionization energies for the carbon atom (C I) and the C⁺ ion (C II) are ≈11.26 eV and ≈24.38 eV, respectively.⁹⁰ The upper level excitation energies for C II 426.7 nm and C III 229.6 nm are ≈20.9 eV and ≈18 eV, respectively. Since the electron temperature is low, and also at the initial stage of plasma formation, the density is not high, the impurity mean free path is longer than the plasma size, and carbon enters the whole plasma volume. As a result, the radiation losses increase, and the electron temperature decreases. A similar picture was observed at LHD in experiments on ICRF plasma production.^68–70 

Experiments on ICRF plasma production in the He + H₂ mixture with initial hydrogen concentration up to 75% at U-2M have indicated that: (i) the plasma density achieved from the gas mixture is several times higher than from pure hydrogen and helium gases in the investigated range of hydrogen concentration, (ii) plasma with a density higher than 10¹⁸ m⁻³ was produced when the RF frequency was close to the fundamental hydrogen cyclotron harmonic ω_RF ≈ ω_ci (H⁺), (iii) the plasma density depends on the hydrogen concentration, and the densest plasma is observed at an initial concentration of 14% hydrogen in the mixture, (iv) the pressure range where the maximum density is reached is observed at hydrogen concentration from 4% to 49%, (v) the electron temperature depends on the initial pressure in the mixture and was low to ≈5–40 eV, (vi) the maximum ion temperature was up to ≈18–23 eV, (viii) the chordal distribution of spectral line intensities showed the presence of carbon ions in the whole plasma column.

In experiments on producing ICRF plasmas in the He + H₂ mixture at U-2M^8,66–68 and LHD,^68–70 a plasma with high density ≈10¹⁹ m⁻³ and low temperature was achieved. The presence of impurities, low RF power, and high radiation losses prevented the plasma from being heated to high temperatures, although plasma with such parameters can be used as a target plasma for NBI. Recent experiments on LHD have shown the possibility of producing a dense and hot plasma in a hydrogen–helium mixture by ICRF alone.⁷¹ These results show further promising scenario of ICRF plasma production.

The experiments on U-2M and LHD indicate that the studied ICRF plasma production scenario appears to be scalable and could be used in large and small stellarator machines. This is, in particular, important for the future experiments ICRF production of target plasma in W-7X in conditions where ECRH start-up is not possible.

This work has been carried out within the framework of the EUROfusion Consortium, funded by the European Union via the Euratom Research and Training Programme (Grant Agreement No. 101052200—EUROfusion). Views and opinions expressed are, however, those of the author(s) only and do not necessarily reflect those of the European Union or the European Commission. Neither the European Union nor the European Commission can be held responsible for them.

The views and opinions expressed herein do not necessarily reflect those of the ITER Organization.

We are pleased to acknowledge the assistance of the Uragan-2M Team.

The authors have no conflicts to disclose.

Yurii Kovtun: Conceptualization (equal); Formal analysis (equal); Investigation (lead); Methodology (equal); Validation (equal); Visualization (lead); Writing – original draft (lead); Writing – review & editing (equal). Vladimir Moiseenko: Conceptualization (equal); Investigation (equal); Methodology (equal); Supervision (lead); Writing – review & editing (equal). Oleksiy Lozin: Formal analysis (equal); Investigation (equal). Mykhailo Kozulya: Data curation (equal); Formal analysis (equal); Investigation (equal). Rostislav Pavlichenko: Data curation (equal); Formal analysis (equal); Investigation (equal). Anatoliy Shapoval: Data curation (equal); Formal analysis (equal); Investigation (equal). Vladislav Bondarenko: Data curation (equal); Formal analysis (equal); Investigation (equal). Demian Baron: Data curation (equal); Formal analysis (equal); Investigation (equal). Sergiy Maznichenko: Investigation (supporting). Valerii Korovin: Investigation (equal). Yevhen Siusko: Investigation (equal). Vladislav Romanov: Investigation (supporting). Yurii Martseniuk: Investigation (equal); Writing – review & editing (supporting). Alexandr Krasiuk: Investigation (supporting). Viktor Listopad: Investigation (supporting). Igor Garkusha: Investigation (equal); Project administration (equal); Supervision (equal). Arturo Alonso: Investigation (equal). Andreas Dinklage: Investigation (equal); Project administration (equal). Dirk A. Hartmann: Investigation (equal). Yevgen Kazakov: Investigation (equal). Heinrich Laqua: Investigation (equal). Jef Ongena: Investigation (equal). Torsten Stange: Investigation (equal). Tom Wauters: Investigation (equal).

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