ICRH-related impurity source and control across experiments in H, D, T plasmas at JET-ILW

The experimental and theoretical analysis were focused on experiments conducted to assess the effect of plasma isotopes, protium (H), deuterium (D), and tritium (T) on ion cyclotron resonance heating (ICRH) related plasma wall interactions. Comparison of L-mode discharges with N = 1 3He and N = 1 H minority ICRH heating scenarios were done for different isotopes. For the selected pulses, the behaviour of high-Z, mid-Z and low-Z intrinsic impurity and radiated power behaviour was investigated based on data from VUV, visible spectroscopy, and bolometry diagnostic at Joint European Torus. It was found that for N = 1 3He scenario during radiofrequency antennas operation, core W, Ni content, Be source and the radiated power are higher for π/2 in comparison to dipole antenna phasing. Lowest core Ni, W content and radiated power is clearly observed for H plasmas in comparison to D and T, where for this ICRH scenario behaviour was similar. However, lower Be photon flux is observed for T in comparison to D plasmas. Be sputtering by He particles is responsible for such an effect. Additionally, several computer simulations were conducted using the COREDIV code. The difference in the electron temperature was due to the difference in the isotope masses. Increased temperature in the central plasma in the case of T plasmas leads to higher radiation in the central plasma in comparison to H plasmas. As a result, the power across separatrix is lower and the temperature on the divertor plate decreases with the increase of the isotope mass. At these temperatures on the divertor plate, W is not sputtered by the main plasma ions H, D and T and by He. For the N = 1 H ICRH scenario clear difference between D and T plasma was observed with higher metallic impurity content for T plasma in comparison to D. Impurity content in the plasmas is found to be sensitive to the power balance between the antenna straps. Its minimum is observed for the maximum of P cen/P tot.

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.

Introduction
Ion cyclotron resonance heating (ICRH) is one of the main heating methods in the Joint European Torus (JET) tokamak, used to control impurities [1], for bulk heating of fuel ions [2,3], and increase the energy of the plasmas.However, the interaction between the ICRH waves and the plasma can also lead to the transfer of energy and particles to the wall, i.e. plasma-wall interaction (PWI) [4,5].Plasmas are usually contaminated by intrinsic impurities.Elements with a high atomic number, like tungsten (W) can have a significant influence on fusion performance.This is of particular importance for tokamak devices with a metallic wall.One of the major challenges in achieving controlled fusion reactions is maintaining the right conditions for the fuel, usually isotopes of hydrogen, to undergo fusion and release energy.High Z impurities can affect this delicate balance.JET-ILW is well suited for extrapolation to ITER, due to its size, the full metal wall (W divertor and beryllium (Be) first wall [6]) and its unique capabilities to operate with tritium (T) and deuterium-tritium (D-T) fuel.At JET-ILW, the presence of W, Be and nickel (Ni), in the plasma is common [7,8].Controlling and minimizing the presence of high Z impurities is crucial.Extensive experimental work has been carried out to elucidate the origin of the isotope effect.Experimental campaigns in deuterium (D) (2019-2020), protium (H) (2020 with radiofrequency (RF)), T (2020-2021) and the second D-T experimental campaign (DTE2) (2021-2022) [9], addressed the impact of isotope effect on the PWI, core, scrape-off layer (SOL) and edge plasma physics to answer essential questions for magnetic fusion development that will affect ITER operation [10].In particular, for the constant ELM frequency, the confinement showed a clear dependence on the isotope mixture.The pedestal density increase with effective mass was found to play a major role.The pedestal density isotope dependence was found to be similar to the pedestal pressure, pedestal particle confinement, and stored energy isotope dependence, while both the central electron and ion temperature show a reverse trend [11].As it was reported in [12] for all isotopes the core confinement time increased with pedestal pressure due to electromagnetic turbulence stabilisation and profile stiffness.The changes in the pedestal density can also impact other aspects, for example, the plasma rotation which in turn affects the transport in the plasma core.In the ohmic plasmas, the global energy confinement was highest in T and lowest in H [13].In D-T and in T plasmas the transition from L-mode to H-mode requires less input power than in D, due to the reduction in the density value at which the power threshold is minimum [14].In the dimensionless isotope mass scaling experiment, higher scaled energy confinement time is found in T in comparison to D plasmas [15].In the Hybrid scenario, good access to Hmode with sustained high fusion power and controlled impurity radiation was obtained after re-optimisation of gas and power waveform.In the hybrid high performance plasmas, current ramp-up and q-profile were impacted by isotope mass [16] and the pulse with higher isotope mass had improved confinement [17].However, the change in isotope from D to T led to a delay in the occurrence of type I ELMs after the Hmode entry resulting in very high edge radiation.The change of the isotopic mix from D to D-T in JET operations showed an increased incidence of neoclassical tearing modes (NTMs), in the Baseline scenario because of the higher β N .NTM's are responsible for a decrease in plasma performances and leading in some cases to disruptions.In contrast, the Hybrid scenario with similar levels of β N , resulting in a comparable magnetohydrodynamic activity [18].The sawtooth instabilities period tends to increase with high isotope mass [19].For the same core plasma density, T plasmas were more strongly detached than H, D, D-T [20].The impact of isotope mass on Be and W gross erosion was also investigated.The results showed the increase in Be physical sputtering yield and total W source from H to T [21,22].Higher power loads on the Be limiter, caused by Neutral Beam re-ionisation, and much cooler divertor target surfaces were observed in T plasmas than in their D [23].Both phenomena are driven by SOL physics.The T plasmas had higher separatrix density and collisionality and broader SOL profiles which is beneficial for the power exhaust in the divertor.Isotopic effects were already observed in the first D-T campaign performed at the Tokamak Fusion Test Reactor in 1994 [24] and during the first D-T experimental campaign (DTE1) at JET executed in 1997 [25][26][27][28], however experiments at JET were performed in carbon wall (JET-C).Most PWI studies (and practically all on RF sheaths) in the literature were conducted in D plasmas, while fusion reactors will use other fuel mixtures.From the previous study, there is evidence that the RF-induced PWI is different between H and D plasmas but a strict comparison of H, D, and T plasmas for quantitative conclusions are missing.Therefore, the purpose of this paper is comparison of isotopes effect on ICRH related impurity sources and core plasma contamination.Analysis are performed for two ICRH heating schemes, the fundamental helium-3 (N = 1 3 He) and the fundamental hydrogen (N = 1 H) minority scenarios, respectively.In these scenarios, the ICRH waves have frequencies close to the frequency at which minority ions, gyrate around magnetic field lines.The notation N = 1 refers to the fundamental (first harmonic) of the ion cyclotron frequency that is being used to resonate with and heat the ions in the plasma.Previous experiments at JET with 3 He minority heating [29] and numerical prediction for ITER [10,30] showed that plasmas with a 3 He concentration below 2% gave the best ion heating and fusion performance [10,29].Such heating scheme and 2nd harmonic of T are the reference ICRH scenario for ITER full-field (B T = 5.3 T) D-T plasmas [10,31].The demonstration of ITER D-T ICRH scenarios was one of the objectives of the JET-DTE2 program.Dedicated experiments in DTE2 performed in T-rich plasmas with high D-NBI power highlighted the key impact of the fundamental D minority ICRH scenario which significantly increased the net fusion reactivity [2,3].Thereby, this ICRH scheme was a key ingredient to achieve a world record for the most fusion energy produced in a single fusion shot, by generating 59 MJ of heat from fusion reactions.That is more than twice the previous record of 22 MJ, also set at JET in 1997.The isotope effects are likely present in several aspects of the experiments (RF physics, transport, PWI) and need to be disentangled, if possible.A robust method is tested to mitigate this RF-related PWI.In section 2, the isotope impact on high-, mid-and low-Z impurity and plasma radiation for the fundamental ICR frequency of 3 He minority ICRH scenario is given.In section 3 the isotope effect on intrinsic impurities is analysed for the H minority ICRH scenario with power splitting between JET ICRH antenna straps.Both experimental and numerical modelling results are discussed.Finally, conclusions are given in section 4.

The fundamental ICR frequency of 3 He minority ICRH scenario in H, D and T
The experimental and theoretical analyses were focused on identical L-mode discharges performed in H, D and T plasmas at B T = 3.35 T, I p = 2.15 MA with both the inner and outer strike points on the horizontal W-coated divertor target plates (see figure 1(c)).Main species gas was adjusted to match, as far is possible, the SOL density in the pulses.The T stored in uranium beds was injected into the torus through dedicated tritium injection modules [32].Helium (He) gas puff was used to apply the N = 1 3 He minority ICRH scenario at resonance frequency f = 32.5 MHz.The 3 He concentration was ∼2%.The A2 ICRH antenna, comprising four antenna modules designated as A, B, C, and D, was employed in the experimental setup.A schematic top view depicting the spatial arrangement of modules A, B, C, and D is illustrated in figure 1(a).Antenna A + B and C + D were powered separately as an antenna pair via a 3 dB hybrid coupler network, and via an external conjugate-T network, respectively [33].As it is presented in    1.Since H shots suffered from bad antenna B phase control, which affects power coupling, heating efficiency and PWI, only the data in the ohmic phase and during C + D antenna pair operation were analysed.The experimental data has been averaged over the half-second period for each discharge in the ohmic phase and at 9 s and 10 s (see figure 2).Clear isotope effect on the heating performance and plasma confinement was observed.For the similar coupled ICRH power, the diamagnetic plasma energy (see figure 3) was 25% lower in D plasma and 50% lower in H plasma in comparison to T plasmas.An increase in the ICRH heating power from 0 to 2 MW resulted in an increase in diamagnetic energy by 0.2 MJ, 0.3 MJ and 0.4 MJ in H, D and T plasmas, respectively.The central electron temperature was highest for T plasmas and lowest for H plasmas.The same trend is observed and reported in the dedicated confinement experiments [11,13].

Isotopic dependency of the plasma radiation, core high-, and mid-Z impurities
The behaviour of high-Z (W) and mid-Z (Ni) impurities and of the radiated power was investigated based on data from the JET VUV SPRED survey spectrometer [34], with the line-ofsight along the vessel mid-plane (see figure 1(b), green horizontal line) and the bolometry diagnostic [35].As previously reported in [8], the source of Ni in JET tokamak is Inconel or Ni-based alloy components from which the JET vessel is constructed.Ni is also associated with the operation of JET A2 ICRH antennas.As can be seen in figure 1(a), the main W source, marked in red, is the W divertor and in the main chamber NBI shine-through protection plates.Measurement of the Ni concentrations (c_Ni) in the plasmas were determined by the method described in detail in [36].However, the quantitative measurement and comparison of W concentration was not possible due to disconnections of the SXR and XUV diagnostics, to protect them from damage when working with T. Such diagnostics were routinely used in previous studies [7,8,37,38], before the first T injection.The VUV SPRED spectrometer records, within the wavelength range of 147-213 Å, the W feature characterized by multiple overlapping unresolved transition arrays of W-ions ranging from W 14+ to W 35+ .However, W measurement integrated in this wavelength range includes interfering lines due to other elements.Since these can be intense, an alternative measurement technique, denoted in this paper as I W was proposed in [39] to mitigate such contamination.This alternative method involves computing the average of integrations centred on wavelengths of 176 Å and 201 Å, with a background subtraction of the integrated area about a wavelength of 142 Å.For each wavelength, an integration range technique was used, in which Simpson's rule was applied to the area defined by a certain number of pixels on either side of the wavelength centre.With 2048 pixels in the diagnostic detector, the integration range covered ±5 pixels.This methodology was deemed the optimal compromise between minimizing blending and using a sufficient proportion of the line profile to ensure reliable measurements.In the ohmic phase of the analysed discharges, the rise of radiated power (see figure 4(a) for P ICRH = 0) and Ni, W impurity content, due to the rise of isotope mass from H to T is observed, which suggests that isotope effects do exist without RF heating.Such effect is also observed during RF heating and can be seen in tomographic reconstructions of bolometry data, showing highest the radiated power density in T plasmas (figure 5).Observed higher plasma radiated  power proximal to the divertor is attributed to the emission of low-Z and mid-Z intrinsic impurities within the JET reactor.These impurities exhibit the highest cooling rate at lower temperatures.Conversely, fewer elements demonstrate significant radiative properties within the plasma core.The isotope effect on PWI is consistent with other JET experiments [22].Therefore, to distinguish the RF-induced effects from isotopic effects, ∆cNi and ∆I W are calculated for each discharge as a difference in the time-averaged value of the Ni concentration and W intensity during ICRH and that before the ICRH power is applied and presented in figures 4(b) and (c) for analysed pulses.Lowest values of ∆cNi and ∆I W is found in H discharges, but no significant difference between D and T plasmas is observed.It was found that during ICRH Antenna operation the radiated power as well as Ni and W content in the central plasma is higher for π/2 antenna phasing compared to dipole antenna phasing (see figure 4).As it was also reported in [5], impurity content decreases with the parallel wave number k || , which is higher in the case of dipole antenna phasing.

Isotopic dependency of Be source from the limiter
The change of impurity content in the plasma during ICRH can be attributed to an increase of impurity sources at the walls.JET is the only fusion device operating with Be limiters that protect the inner and outer first wall.This allowed studying Be sources in the main chamber and from ICRH antennas equipped with Be screen bars.The isotope effect, H versus D, on the Be effective sputtering yield, has been measured in Ohmic limiter plasmas in [21].The Be erosion yield of D was about twice as high as that of H over the entire range of plasma densities studied.Also, the Z eff of the D studied limiter plasmas was systematically larger compared to H ones.According to previous studies, the RF sheath rectification plays a significant role in the Be sputtering at the limiters near to or connected to active ICRF antenna along the magnetic field lines [40][41][42].Charge exchange recombination spectroscopy (CXRS) is used on JET to study RF sheath-enhanced plasma-surface interactions in ICRF heated discharges [43].The Be influx was diagnosed at four different outer poloidal limiter locations, marked in figure 1(a) as 7D, 8D, 8B and 1B, which are close and far from the active antennas pair C + D. Observation spots are nearly centred on the limiter mid-plane.Figure 6 shows the effect of hydrogen isotopes on the local BeII spectral line emission at 467 nm for the observation location 7D and 1B marked in figure 1(a).Changing H by D, the RF-induced Be source was increasing with the isotope mass.Higher Be photon flux is observed for π/2 antenna phasing in comparison to dipole phasing in D, T and H plasmas.This trend is observed for all analysed lines of sight.As it is presented in [44], Be sputtering yield by He particles is stronger than by D and T. This effect explains the observed greater BeII intensity for D in comparison to T plasmas for which the measured 3 He minority concentration during the ICRH phase was 3 He[X] ∼ 2% and 3 He[X] ∼ 0.8%, respectively.The experimental value of 3 He[X] was calculated based on the Optical Penning Gauge technique, used for the measurement of the fuel isotopic composition and He concentration from a highresolution spectrometer [45].It can be seen in figure 6(a) that the difference between D and T points is notably greater than that observed in figure 6(b).This divergence can be attributed to the heightened influence of 3 He on Be production at position 7D, which resides in closer proximity to the active antennas C and D, as opposed to position 1B, situated at a greater distance from the active antennas.This observation implies that the local RF electric field may contribute significantly to the acceleration of 3 He particles towards the reactor wall.To highlight the larger Be sputtering in T plasmas in comparison to D and H, the BeII brightness presented in figures 6(a) and (b) was normalized to 3 He[X] (see figures 6(c) and (d)).Additional analysis performed on another data set showed that for higher 3 He[X], higher BeII intensity is observed at different limiter locations.However, the 3 He concentration was not found to affect the Ni and W content in the plasmas.
Be sputtering can be sensitive to the electrical settings of the antenna if the sheath potentials explore a domain where sputtering yields Y eff varies with the accelerating voltages driven by ICRF fields.To highlight some expected differences between H/D/T and to illustrate the effect of sheath rectification on the W and the Be sputtering, theoretical curves of gross erosion Y eff versus accelerating voltage were calculated based on formulas presented in [46].
As can be seen in figure 7, W sputtering increases with the isotope mass of the majority species, but no isotope effect on W production is expected if the sheath voltage does not exceed 150 V in T, 250 V in D and 500 V in H.At lower sheath voltage, the sputtering is likely dominated by light impurities.This is illustrated by adding to the plot 2% of 3 He 2+ ions on the plasma mix.Unlike W, lower values of accelerating voltages are needed for Be sputtering.The Y eff for the Be, presented in figure 8, is rather flat for accelerating voltages >50 V. Thus, Be should be influenced by the sheath rectification effect.Also, the isotope effect of the majority species is expected.However, in D and T plasmas weak effect of the light impurities is expected for voltages >50 V.

Modelling with the COREDIV code
`the COREDIV code [47][48][49], which couples self-consistently the plasma core with the SOL, and the impurities with the main plasma.In this study, four impurities are considered, Be, Ni, W and He.The Be, Ni, and He fluxes are given as code inputs while the W flux is self-consistently computed as dependent on the fluxes of Be, Ni, He and on self-sputtering.The main inputs of the code are the plasma volume averaged electron density, the input power, the plasma thermal energy, and the main ion and impurity transport coefficients.The discharges with π/2 antenna phasing, listed in table 1, were simulated.Two time points at t = 9 s and t = 10 s, characterized by different heating powers were selected for the analysis.Given the consistency in conclusions, the subsequent discussion is founded upon the outcomes derived from the second time point.The numerical results for t = 10 s are compared with the experimental profiles of electron temperature and density from high resolution Thomson scattering diagnostic (see figure 9), source temperature at the LFS strike point on the divertor plate (T e PLATE ), W, Ni, Be, He impurity concentrations, effective charge state (Z EFF ), total radiated power (R TOTAL ), core radiated power (R CORE ) and SOL radiated power (R SOL ).A comparison of the main plasma paramaters s provided in table 2. A good agreement between experimental and simulated kinetic profiles was achieved (see figure 9).Similar n e profiles were obtained in H, D and T plasmas.The observed difference in T e was due to the difference in the isotope mass.Increased T e in the central plasma in the case of T leads to higher plasma radiation in the central plasma in comparison to H and D plasmas.As a result, the power across the separatrix is lower and T e PLATE decreases with the increase of the isotope mass.The simulations are consistent with the measurements that show the reduction of the electron temperature at the LFS strike point.T e PLATE measured by the Langmuir probe was 7.19 eV, 5.9 eV and 3.9 eV for H, D and T plasmas, respectively.It must be noted that the probes measurements in such temperature range are moderately accurate.Accordingly, the largest temperature measured by the protection camera looked at the horizontal W-coated outer target plate (see figure 1 In these electron temperature conditions on the divertor plates, W is not sputtered by the main plasma ions H, D and T nor by the He minority.It turned out that in the examined plasmas W is sputtered mainly by Be and Ni.Simulations showed that the Ni radiation in the central plasma is 0.34 MW for H, 0.62 MW for D and 0.77 MW for T plasma and this is about 50% of the core radiated power.At lower heating power in T plasmas, Be does not affect W production.Additional simulations for T plasmas with the change of Be content were addressed to assess its influence on plasma parameters.For this purpose, simulations with two different Be streams C Be = 0.7% and C Be = 2.17%, were performed.It was observed that the Be affects the Z EFF , increasing its value from 1.3 to 1.48.However, for the prevailing temperatures on the divertor plate, it does not affect W production.The higher value of Be content had a very weak effect on plasma radiation.

Power splitting between straps in H minority ICRH scenario in D and T plasmas
The influence of ICRH on the PWI was also studied running identical D (#94998) and T plasmas (#100187) with B T = 3 T, I p = 1.8 MA, in low triangularity L-mode plasma with the inner and outer strike-line on the vertical W-coated and on the semi-horizontal bulk-W target plates (tile 5, see figure 1(c)).The fundamental hydrogen minority N = 1 H ICRH scenario at resonance frequency f = 42.5 MHz was utilized for the experiment.Discharges were run with A2 antenna D only and dipole antenna phasing.As it can be seen in figure 10(a), in D plasmas 1 MW of ICRH power was launched, however in the T case only 0.5 MW.Higher power would saturate Be-line  spectroscopy in the T plasmas.However, in such conditions similar density and temperature profiles were obtained, therefore the comparison of discharges in a quantitative way was possible.Then our data were normalized to ICRH power to get rid of the power effect.In all discharges, a scan of the power partition between the inner/outer straps of one of A2 antenna D was performed aiming at minimising the RF-sheath rectification effect.The fraction of power from the inner straps to total power P cen /P tot is illustrated in figure 10(a).Enhanced RF-PWI are observed in T plasmas.Bulk radiated presented in figure 10(b) power is higher in T plasmas in comparison to D plasmas.The production of neutral Be atoms depends on the isotope, T plasmas showing higher Be neutral particle fluxes than D plasmas (see figure 10(c)).Indeed, the BeI line intensity is found to be sensitive to the power balance between the antenna straps for a given dipole antenna phasing, as the metallic Ni (see figure 10(d)) and W (see figure 10(e)) impurity content in the plasma is affected.The minimum in the radiated power and impurity content is observed for the maximum of P cen /P tot .The optimal RF setting is isotope independent.Reduction of the RF-driven PWI for P cen /P tot ∼ 0.8 is consistent with the A2 antenna near field calculation and the ICRH antenna optimisation concept presented in [50].Different hypotheses are under consideration, to explain the difference between D and T plasmas in this experiment.For example, an increased Be sputtering yield by T, or invoking slow wave properties in the SOL and larger RF-sheath rectification with T. At JET-ILW, W impurities come mainly from W divertor baffles, however, it must be noted that they can also be sputtered from PFCs located in the main chamber, such as the NBI shine-through protection plates.Also, high-Z impurities migrate throughout the vessel and can be deposited in the main chamber limiters.In the present study, the effect of RF power balance on impurity behaviour is mostly observed in the limiter region and not observed in the divertor W source.This can be seen in figure 8(f ) showing neutral WI emission at 400.9 nm from the outer divertor derived from the combination of the signals detected by the divertor spectroscopy and the W filterscope diagnostics to separate the contribution of the plasma continuum.This is an identical approach to the one described in [51,52].Additionally, the WI signal is normalized to ion flux to the divertor plate.The divertor W source is lower for T plasmas in comparison to D. This is explained by the lower power across the separatrix and lower divertor plate temperature with the increased isotope mass obtained from COREDIV simulations and the experimental data.The electron temperature near the divertor target was measured by both, Langmuir probes and the lineintegrated plasma emission spectroscopy using the Balmer photo-recombination continuum feature [53].For the conditions achieved in the experiment, the temperature along the divertor plate with the separatrix is in the range of 2-8 eV in T plasmas, in which case W is not sputtered by hydrogen isotopes [54].This suggests that W sputtering in the divertor region is mainly due to Be and Ni impurities.A reliable probe or spectroscopic measurement for D plasmas was not possible.However, the protection camera looked at the semi-horizontal bulk-W target plates (see figure 1(c), Tile 5), measured maximum temperature of 660 • C for D plasmas and 520 • C for T plasmas.Indicating a higher load with 1 MW in D than with 0.5 MW in T.

Conclusions
The increase of plasmas radiated power with isotope mass was measured and the increase of W, Ni impurity content in the plasmas is confirmed.Such behaviour is related to the increase in Be physical sputtering yield from H to T. In addition to normal plasma effects, there are those associated with ICRH.Results from recent experiments demonstrate that the ICRHspecific impurity release can differ significantly depending on the ICRH scenario and minority content as well as hydrogen isotopes constituting the plasma, with T-plasmas showing higher impurity content compared to D and H plasmas. Application of the N = 1 3 He minority ICRH scenario in ITER will require measurements and control of helium concentration because as it was found for JET-ILW it can play a major role in Be sputtering.The effect of ICRH on impurity behaviour is mostly observed in the limiter region, highlighting the important role of the RF-sheath rectification effect and induced potential driven by the existence near the antenna of an RF electrical field component E || , parallel to the magnetic field.This potential leads to the acceleration of plasma particles towards the first wall which causes impurity sputtering.Due to the lower power across the separatrix and lower divertor plate temperature, obtained from COREDIV simulations and the experimental data, the W source from the divertor is lower with the increased isotope mass in this experiment.JET A2 Antenna RF modelling using the TOPICA [55] and ERMES [56] code is ongoing in support of this experiment.Both codes predict a slightly higher electric field in front of the antenna in T plasmas as compared to D. Some more work is required to understand code results in H plasmas. Also, in ERMES strong toroidal asymmetries in the field pattern in front of the antenna are observed in the case of D and T which still need to be understood/explained.A detailed report will be the subject of future publications.ICRH-specific PWI such as those observed at JET tokamak must be considered in the development of heating scenarios in ITER.RF settings were found on A2 antennas that can reduce the RF-related impurity production near the antenna for 2 isotopes at least.Simulations in [52] suggest that similar settings exist on the ITER antenna.These do not reduce excessively the maximum coupled power.

Figure 1 .
Figure 1.(a) Toroidal cross section of JET showing ICRH auxiliary heating system.Also showing Be (green) and W (red) plasma facing components.CXRS spectroscopy poloidal limiter locations 7D, 8D, 8B, 1B.(b) Poloidal cross section with the plasma configurations in N = 1 3 He (blue) and N = 1 H (magenta) minority ICRH scenario with the lines of sight of JET VUV spectrometer (green horizontal line) and ICRH antenna (orange) at JET tokamak.(c) Magnetic configuration in the divertor with the outer strike point on tile 5 (in magenta) and tile 6 (in blue).

Figure 2 .
Figure 2. Time evolution of the launched ICRH power in #95645 by using two antennae in pair, with marked in red time intervals for calculations of average values plasma parameters.

Figure 3 .
Figure 3. (a) Diamagnetic energy as a function of ICRH power in H, D, T plasmas for discharges with dipole or π/2 ICRH antenna phasing.

Figure 4 .
Figure 4. (a) The bulk radiated power, (b) ∆cNi and (c) ∆I W (Data with subtracted impurity level in the ohmic phase) as a function of ICRH power for dipole or π/2 antenna phasing in H, D, T plasmas.

Figure 5 .
Figure 5. Tomographic reconstruction of the radiated power density at t = 10 s for the H pulse (left), D pulse (centre) and for the T pulse (right) in π/2 antenna phasing.

Figure 6 .
Figure 6.BeII at 467 nm line emission, measured at outer poloidal limiters locations (a) 7D and (b) 1B, BeII/He[X] measured at locations (c) 7D and (d) 1B as a function of ICRH power for dipole or π/2 antenna phasing in H, D, T plasmas.

Figure 7 .
Figure 7. Theoretical curves for W gross erosion by H, D and T isotopes (dashed lined) and the mix of majority plasma ions with 2% 3 He 2+ as a function of accelerating voltage.

Figure 8 .
Figure 8. Theoretical curves for Be gross erosion sputtering yields by H, D and T isotopes (dashed lined) and the mix of majority plasma ions with 2% 3 He 2+ + 2% 9 Li 4+ as a function of accelerating voltage.

Figure 9 .
Figure 9. Experimental electron density (ne) and temperature (Te) profiles, for pulses #98370, #95645 and #100129 at the considered time t = 10 s and the corresponding COREDIV computed profiles.

Figure 10 .
Figure 10.Comparison of (a) launched ICRH power and fraction of power from the inner straps, (b) bulk radiated power, BeI at 457 nm line emission measured at outer poloidal limiters locations 7D, (d) Ni concentration and (e) W quasicontinuum intensity and (f ) WI line emission at 400.9 nm from the divertor (normalised to 1 MW of ICRH power), for the time evolution of D pulse (#94998) and T pulse (#100187).

Table 1 .
List of analysed pulses with N = 1 3 He ICRH scenario.

Table 2 .
Main plasma parameters with experiment and simulations for t = 10 s.