Mitigation of the tracer impurity accumulation by EC heating in the LHD

The mitigation of a tracer impurity accumulation in the core region of high-temperature helical plasma was clearly observed by applying electron cyclotron heating (ECH) in the large helical device (LHD). In the LHD, the accumulation of impurities toward the centre of the plasma has been observed in a high-density regime. In this study, for observing clearly the behaviour of impurity ions in the plasma core, the extrinsic ‘tracer’ impurity was injected into that region by means of a tracer-encapsulated solid pellet (TESPEL). The high-density LHD plasma without ECH definitely shows the strong impurity accumulation, and then it causes the reduction in electron and ion temperatures in the core region. When ECH was applied just after the TESPEL injection, the accumulation of the tracer impurity ions was mitigated. Even after ECH was switched-off, the intensities of the line emissions from the highly-ionized tracer impurity were increased very slightly. The micro-turbulence measurement with a 2-dimensional phase contrast imaging diagnostic during ECH does not support the view that the change in the micro-turbulence would enhance the outward flow (an increase in a diffusive flux, a decrease in an inward convective flux and/or a change the direction of the convective flux from inward to outward) of the impurity ions. Moreover, at this moment, there is no conclusive data regarding a radial electric field measured with a charge exchange spectroscopy diagnostic to support the view that the change in the radial electric field would be attributed to the increment in the outward flow of the impurity ions from the core region of the LHD plasma.


Introduction
In a magnetic confinement fusion device, various impurities from low-Z material, e.g. a helium ash, which is a by-product of the fusion reaction, to high-Z material, e.g. a tungsten, which is derived from a possible plasma facing component, will exist inside a high-temperature plasma. When the amount of impurities exceeds an acceptable level for some reason, e.g. possibly due to an accumulation of the impurities towards the plasma core, it can cause radiation losses and plasma dilution resulting in lower fusion power, which leads to a significant fusion reactor performance degradation. Therefore, the impurity accumulation is a potential show-stopper for the realization of the fusion reactor, and it is crucially important to develop an effective scheme for avoiding and/or controlling the impurity accumulation towards the plasma core. In helical The mitigation of a tracer impurity accumulation in the core region of high-temperature helical plasma was clearly observed by applying electron cyclotron heating (ECH) in the large helical device (LHD). In the LHD, the accumulation of impurities toward the centre of the plasma has been observed in a high-density regime. In this study, for observing clearly the behaviour of impurity ions in the plasma core, the extrinsic 'tracer' impurity was injected into that region by means of a tracer-encapsulated solid pellet (TESPEL). The high-density LHD plasma without ECH definitely shows the strong impurity accumulation, and then it causes the reduction in electron and ion temperatures in the core region. When ECH was applied just after the TESPEL injection, the accumulation of the tracer impurity ions was mitigated. Even after ECH was switched-off, the intensities of the line emissions from the highly-ionized tracer impurity were increased very slightly. The micro-turbulence measurement with a 2-dimensional phase contrast imaging diagnostic during ECH does not support the view that the change in the micro-turbulence would enhance the outward flow (an increase in a diffusive flux, a decrease in an inward convective flux and/or a change the direction of the convective flux from inward to outward) of the impurity ions. Moreover, at this moment, there is no conclusive data regarding a radial electric field measured with a charge exchange spectroscopy diagnostic to support the view that the change in the radial electric field would be attributed to the increment in the outward flow of the impurity ions from the core region of the LHD plasma.
Keywords: impurity transport, impurity accumulation, mitigation, ECH, tracer-encapsulated solid pellet (TESPEL), large helical device (LHD) (Some figures may appear in colour only in the online journal) Original content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. plasmas, a global impurity confinement time has been found to generally increase with a line-averaged electron density [1]. Unfortunately, a so-called 'ion temperature screening effect' due to an ion temperature gradient, which is predicted by the neoclassical transport theory and creates a convective flow directed outward [2], is not expected in the helical plasmas unlike in tokamaks [3]. This means that the helical-type fusion reactor based on a scenario with a high-density seems to be unfeasible. However, a favourable confinement mode (highdensity and high-confinement mode, HDH-mode), which has a good particle confinement with a low impurity confinement, has been found in the W7-AS stellarator [4]. And moreover, an extremely hollow impurity profile, which is a so-called 'impurity hole', has been observed in the plasma with a large ion temperature gradient in the LHD [5]. These findings could shed light on the path to the helical-type fusion reactor. Meanwhile, in order to enhance the economic rationality of the possible helical-type fusion reactor, it is highly important to develop several methods to achieve actively the low impurity confinement state. The additional heating, such as electron cyclotron heating (ECH) and ion cyclotron heating (ICH), is one of several such method to control the impurity inside the plasma. Even now, many experiments with the additional heating have been performed to demonstrate and investigate the suppression or mitigation of the core impurity accumulation in tokamaks (among others, see, for example, [6][7][8][9]) and stellarators [1,10]. In both devices, the clear effects of the additional heating on the impurity accumulation have been observed. However, the physical mechanisms of such effects still are not yet cleared.
In this paper, we show observational results on the mitigation of the impurity accumulation by applying ECH in a plasma of the Large Helical Device. This paper is structured as follows. In section 2, the experimental set-up including important diagnostics for this work is explained. In section 3, we show the experimental results. In section 4, we discuss the results obtained to investigate the cause of the mitigation. In section 5, we draw some conclusions.

Experimental set-up
In this study, we investigate high-density discharges in the LHD [11]. The LHD with a heliotron-type magnetic configuration has superconducting l/m = 2/10 helical coils and three pairs of superconducting poloidal coils. Here, l and m are a pole number of the helical coil winding and a toroidal field period, respectively. In this work, the magnetic axis is set at R 3.6 ax = m, and then the resulting averaged minor radius a is ∼0.62 m. The magnetic field at the axis B ax is set at 2.75 T. The plasma was initiated by ECH (injected power: about 1.2 MW) and heated additionally by negative-ion and positiveion based neutral beam injections (NBIs, total injected power: about 7 MW). In the LHD, an impurity hole, which is characterized by the extremely hollow profile of the impurity density, is observed associated with an increase of ion temperature gradient in the plasma with an ion thermal internal transport barrier [5]. Here, as will be shown later in figure 4, the plasmas of interest have no impurity hole, because there is no large ion temperature gradient. No impurity hole is also inferred from the shape of radiated power profile measured with an absolute extreme ultra violet silicon photodiode (AXUVD) array, as will be shown later in figure 3. And moreover, no significant magnetohydrodynamics (MHD) activities are also observed.
In order to inject the impurities into the core plasma of LHD, a tracer-encapsulated solid pellet (TESPEL [12][13][14]) is used. In this work, the TESPEL contained two extrinsic impurities, vanadium (V) and chlorine (Cl). The vanadium is in the core of the TESPEL. And the chlorine is in the shell of the TESPEL in the form of an organic compound, poly-2,6-dichlorostyrene (C 8 H 6 Cl 2 ) n . The light emission from the ablation cloud of the TESPEL injected into the LHD plasma is measured with an optical-fibre-based diagnostic system [15]. The ablation lights from the shell and core are measured separately with the corresponding optical filters; one has 657.2 centre λ = nm and FWHM = 1.2 nm for Hα emission from the shell ablation and the other has 412.3 centre λ = nm and FWHM = 1.1 nm for V I emission from the vanadium ablation. The most important diagnostics in this work is a spectrometer for detecting the behaviour of the impurity ions in the core region of the LHD plasma. When the electron temperature of the LHD plasma of interest is basically above 1 keV, the line emissions from the impurity ions in the core LHD plasma will appear in a vacuum ultraviolet (VUV) domain. A 2 m Schwob-Fraenkel soft x-ray multi-channel spectrometer (SOXMOS) [16,17], which is installed at Port 7-O of LHD, is used, although not calibrated, for measuring the temporal behaviour of the line emissions in the VUV domain. The temporal resolution of the SOXMOS is set at 50 ms. A wide-angle (2π) metal foil bolometer, which is installed at the same port (Port 3-O of LHD) as the TESPEL injector is used for measuring the total radiated power P rad with a 5 ms temporal resolution [18]. Unfiltered 20-channel AXUVD arrays [19], which are installed at the upper side of Port 8-O of LHD, can also measure the plasma radiation, whose energy ranges from 1.1 eV (visible) to ∼6000 eV (x-ray) [20]. The temporal resolution of the AXUVD array is set at 0.1 ms. A YAG Thomson scattering system, which is installed at Port 4-O of LHD, is used for measuring detailed radial profiles of electron density n e and temperature T e along the LHD major radius at a horizontally-elongated cross-section [21,22]. The typical temporal resolution of the LHD YAG Thomson scattering system is 100 ms. A charge exchange spectroscopy diagnostic (CXS), which utilizes the positiveion-based NBI at Port 5-O of LHD, is used for measuring the temporal and spatial evolutions of ion temperature T i , plasma poloidal and toroidal rotation velocities by using a charge exchange reaction between fully-ionized carbon impurity and atomic hydrogen from the neutral beam [23]. Consequently, the CXS can estimate a radial electric field. A 2-dimensional phase contrast imaging diagnostic employing a CO 2 laser [24][25][26], which is installed at Port 8.5-U of LHD, is utilized for estimating the turbulence properties in the LHD plasma. . The TESPEL was injected at the time of 3.941 s. As can be seen from figure 1, just after the TESPEL injection, many emission lines from the highly-ionized vanadium and chlorine appeared clearly. Here, we study the vanadium Be-like emission (V XX 15.936 nm (1s 2 2s 2 1 S-1s 2 2s2p 1 P)) and the vanadium Li-like emission (V XXI 24.04 nm (1s 2 2s 2 S 1/2 -1s 2 2p 2 P 3/2 ) and 29.37 nm (1s 2 2s 2 S 1/2 -1s 2 2p 2 P 1/2 ) (not shown in figure 1)) from the highlyionized vanadium.

Experimental results
In order to investigate precisely the transport of the impurities injected into the core plasma, the impurity injection should be performed with minimal disturbance of the plasma. In this regard, the impurity injection by using the impurity pellet injection technique is not the most suitable. Therefore, it is important to assess the impact of the TESPEL injection on the LHD plasma. In this study, we investigate three LHD discharges: the discharge without additional ECH (#128070), the discharge with additional 0.7 MW, 154 GHz ECH (#128087) and the discharge with additional 1.5 MW, 154 GHz ECH (# 128081). All the discharges have the TESPEL injection at the time of around 3.95 s. The diameter of the TESPEL used is 500 μm for #128070, and 600 μm for #128081 and #128087, respectively. The particle numbers of the impurities injected for these discharges are summarized in table 1. The particle number of the carbon, which exists intrinsically for these LHD discharges is estimated of the order of 10 19 from the measurement with the CXS. Therefore, the particle number of the impurities injected (vanadium, chlorine and carbon) is smaller than that of the intrinsic carbon particle by a factor of one to two orders of magnitude. Figure 2 shows waveforms of LHD discharges that are being studied, including two reference LHD discharges (ref. 1: without the injections of TESPEL and ECH (#128069), ref. 2: without the TESPEL injection, but with additional 1.5 MW, 154 GHz ECH (#128082)). As can be seen from figure 2, there are no significant changes in the electron density and temperature just after the TESPEL injection. As already described in the Introduction, the global impurity confinement time in helical plasmas increases with the line-averaged electron density. We have tried unsuccessfully to make the steady-state line-averaged electron density in the discharges without the additional injection of ECH, and the line-averaged electron density continued to increase even without a gas fueling. This could be attributed to a recycled hydrogen gas from a first wall of the LHD vacuum vessel and the accumulation of impurity toward the plasma centre. In order to establish the steadystate line-averaged electron density, around 4 5 10 19 ∼ × m −3 , we have finally applied a hydrogen gas puff for 120 ms from t = 3.3 s in the discharges with the additional injection of ECH. Then, unfortunately, there are no hydrogen-gas-puffed reference discharges without ECH during the discharge. However, since the values of the line-averaged electron density from t = 4.8 s to t = 6.0 s (about half of the duration of interest) in the discharges without the additional injection of ECH are close to those of the line-averaged electron density in the discharges with ECH, the discharges without ECH (#128069 and #128070) are, although not the best, used as references. When ECH is applied during the discharge, the lineaveraged electron density decreases slightly with time. The effect of ECH on the line-averaged electron density, which is well known as a density pump-out effect [27], is strengthened by the higher-power ECH. In the case without ECH during the discharge (#128070), the electron temperature at the plasma centre decreased gradually, and then it dropped to below half, compared to that around the time of 4.0 s. Such a decrease in the electron temperature is also observed in the case without the TESPEL (#128069), although it is a bit behind the case with the TESPEL. Thus the impurities injected by the TESPEL might hasten the decrease in the electron temperature in the plasma centre. When the total radiated power is compared under the same electron temperature level in the plasma centre, the total radiated power (at t = 6.7 s) in the case without the TESPEL is higher than that (at t = 5.5 s) in the case with the TESPEL. Therefore the contribution to the total radiated power by the impurities injected by the TESPEL is found to be very small. The small contribution to the total radiated power by the impurities injected by the TESPEL is also confirmed by the comparison between the discharges having the additional injection of ECH with (#128081) and without TESPEL (#128082). However, the measurement with the AXUVD array, even though it is the sightline-integrated, indicates a pronounced increase of the local radiated power from the plasma centre in the case with the TESPEL, but without ECH (#128070), as shown in figure 2(g). As can be   s. And V XX (15.936 nm) is also quickly increased, and then it starts to decrease shortly thereafter. However, V XX is increased again, gradually up to the time of t 5.7 ∼ s. These experimental results suggest that the impurities injected is accumulated towards the plasma core. When 0.7 MW, 154 GHz ECH is applied (#128087), the behaviour of impurity injected is certainly changed. Just after the TESPEL injection, V XXI (24.04 nm) is quickly increased, and then gradually decreased. However, the decrease of V XXI seems to be stagnated around t = 5.0 s. And the temporal behaviour of V XX (2 15.936 × nm) shows the similar behaviour as that of V XXI. These behaviours suggest that applying 0.7 MW, 154 GHz ECH certainly affects the impurity accumulation, but it is not sufficient for mitigating that. This is also proven by the temporal behaviour of the line emissions from the vanadium ions after ECH switch-off. After ECH (after the time of t = 6.0 s), both the V Li-like and Be-like emissions start to increase again. When 1.5 MW, 154 GHz ECH is applied (#128081), the behaviour of impurity injected are dramatically changed. V XXI is decreased to 20% of its peak, and then, even after ECH switch-off, it maintains the suppressed level. The temporal evolution of V XX is almost the same as that of V XXI. These behaviours strongly suggest that applying Temporal evolutions of (a) the line-averaged electron density, (b) the electron temperature at the plasma centre, (c) the ion temperature at the plasma centre, (d) the diamagnetic energy, (e) the total radiated power, the sightline-integrated signal intensity of (f) channel 3 (for the edge) and (g) channel 12 (for the centre) in the AXUVD8OU array, and the integrated counts for the (h) V Li-like and (i) Be-like emissions for three LHD discharges (#128070 (black), #128087 (blue) and #128081 (red)). The fundamental heating duration of ECH and NBI is depicted in figure 2(a). The additional 154 GHz ECH (the total injected power is 0.7 MW for #128087 and 1.5 MW for #128081, respectively) was applied from t = 4.0 s to t = 6.0 s, as indicated by a light-brown-colour-hatched area. 1.5 MW, 154 GHz ECH definitely mitigates the impurity accumulation. After ECH, in almost all cases, the electron density increased and the electron and ion temperatures decreased. Taking into account the temporal response of the total radiated power, the revival of the impurity accumulation could contribute to the plasma response. The change in plasma radiation profile would be a better indicator for the appearance of impurity accumulation. As can be recognized easily from figure 3(b), the plasma radiation measured with the AXUVD array shows a very peaked profile in the case without ECH (#128070). And the very peaked radiation profile almost disappears with 1.5 MW ECH, which is close to that before the TESPEL injection. Figure 4 shows the normalised radial profiles of the electron density and temperature, and the ion temperature before and after the TESPEL injection for three LHD plasmas. Here, the deposited regions of the vanadium injected are indicated in figure 4(a). The vanadium injected ablated in a region between r a  Normalised radial profiles of the electron density, the electron temperature and the ion temperature ((a), (c), (e)) before and ((b), (d), (f)) after the TESPEL injection for the three LHD discharges (#128070 (black), #128087 (blue) and #128081 (red)). The deposited region of the vanadium impurity injected by the TESPEL is indicated in figure 4(a). And the normalised radial profile of the absorbed power density of the additional ECH is depicted in figure 4(d).
electron density and temperature, and the ion temperature. The electron temperature profile is changed from peaked to slightly hollow, which is one of the most significant features of the impurity accumulation towards the plasma core. In the case with 1.5 MW ECH, both the electron and ion temperatures increase strongly in comparison to the values before ECH. On the other hand, the overall electron density is decreased slightly. In the case with 0.7 MW ECH, the electron and ion temperatures over the whole region are still higher than that before ECH. However, the electron density inside the r a eff 99 / of 0.4 is higher than that before ECH. This increment of the electron density in the plasma core suggests that the impurities continues to be accumulated even after ECH switch-on. The erosion of the electron temperature could be appeared when the radiation losses exceed the deposited power. However, as already shown in figure 2, the total radiated power is increased a little, when the erosion of the electron temperature due to the impurity accumulation is appeared. This is because the absorbed power in the central region of that plasma is very low, as shown in figure 5. The very low and rather flat absorbed power density is attributed to the high-density discharges. As shown in figure 4, the profiles of the electron density as well as that of the electron temperature are certainly modified by the injection of ECH. And then, the first thing to check is whether a mean charge state of vanadium ion is also changed or not by the changes in the electron density and temperature. It is important to judge whether the temporal variation of the line emissions from the highly ionized vanadium is that of the vanadium ion density or the change in the charge state distribution of vanadium ion. As shown in figure 6(a), the mean charge state ⟨Z⟩ of vanadium ion, which is calculated with the atomic code, FLYCHK [28], depends on the electron temperature, does not on the electron density in the parameter range of interest. Figures 6(b) and (c) show mean charge state profiles just before and after the TESPEL injection, which are calculated by using the data shown in figure 6(a). It should be noted here that the mean charge state profies calculated do not include the transport effect. In the case with 1.5 MW ECH (#128081), the region where the mean charge state is above ⟨Z 19 ⟩ = , the Be-like ion, could exist in the wide region, inside the r a eff 99 / of 0.7 at t = 3.933 s (before the TESPEL injection), and inside the r a eff 99 / of 0.85 at t = 5.700 s (during ECH). Taking into account the fact that the V Li-like and Be-like emissions increase very little after ECH switch-off, the temporal behaviour of the line emissions from the highly-ionized vanadium would reflect that

Discussion
There are several possibilities for the mitigation of the impurity accumulation by applying ECH. One possible cause is the enhancement of the outward flow, i.e. an increase in a diffusive flux, a decrease in an inward convective flux and/ or a change the direction of the convective flux from inward to outward, due to the enhancement of the turbulence by the increase in the electron temperature gradient, such as trapped electron mode [29] and electron temperature gradient mode [30]. Thus, it is important to check the variation in such driving terms, including the gradients of electron density and ion temperature for three LHD discharges of interest. As shown in figure 7, the inverse electron density scale lengths for the plasmas with ECH of 0.7 MW and 1.5 MW is similar to that for the initial state (t = 3.933 s) of the plasma without ECH. The inverse electron temperature scale length with 1.5 MW ECH is slightly larger than the others in the region outside r a 0.6 eff 99 / ∼ . The inverse ion temperature scale length with 1.5 MW ECH is certainly but very slightly larger than the other inverse ion temperature scale length in the region inside r a 0.5. eff 99 / ∼ In either case, the inverse gradient scale lengths with 1.5 MW ECH are found not to be so different from those in the other cases. As can be seen in figure 8, the electron density fluctuations measured with the 2D PCI are found to exist largely outside r a 0.7. / ∼ is considered to be stabilized due to the collisions. Therefore, the expelling of the impurity ions from the plasma core (inside r a 0.7 eff 99 / ∼ ) is not attributed to the change in the micro-turbulences. The maximum amplitude of the density fluctuations measured in the case with 0.7 MW ECH is larger than that in the case with 1.5 MW ECH in the first place.
The other possible cause for the mitigation of the impurity accumulation is the change of the radial electric field from the positive to the negative. In helical plasmas, the radial electric field has a significant impact on the plasma confinement and impurity transport [31,32]. The transition of the radial electric field from the ion root (negative) to the electron root (positive) triggered by enhancing the electron particle flux due to ECH has been already demonstrated in low-density CHS heliotron plasmas [33]. However, the radial electric field evaluated with the CXS during 1.5 MW ECH was applied seems to be almost unchanged compared to that before ECH within the errors, as shown in figures 9(e) and (f). Therefore, at present, there is no conclusive data to rationalize the mitigation of the impurity accumulation due to ECH.
In order to determine the cause for why the impurity accumulation was mitigated by applying ECH, further investigation using the theoretical and simulation studies would be useful. As can be seen in figures 9(c) and (d), there is a large uncertainty in the measurement of the toroidal rotation velocity in the plasma core, inside the r a eff 99 / of around 0.5. The calculation including momentum conservation [34] could provide us a more precise picture on that. Currently, the evaluations of the neoclassical radial electric field taking into account momentum conservation and turbulent flux of the impurities are underway, which will be published elsewhere. From a pragmatic point of view, the optimum conditions (injected power, power absorbed location and ECH duration) of the additionally applied ECH should be also clarified for achieving the mitigation of impurity accumulation in the fusion-reactor relevance conditions.

Summary
We assessed the effect of ECH on the impurity accumulation toward the plasma core in the LHD. When 1.5 MW, 154 GHz ECH was applied just after the injection of the impurity by using the TESPEL, the mitigation of the impurity accumulation toward the core region of the LHD plasma was observed. On the other hand, when 0.7 MW, 154 GHz ECH was applied just after the TESPEL injection, the mitigation of the impurity accumulation was incomplete. During ECH, the change in both the inverse gradient scale lengths of electron density, electron temperature and ion temperature and the properties of micro-turbulence measured with the 2D-PCI does not support the view that the change in the micro-turbulence would enhance the outward flow of the impurity ions. In addition, there is no conclusive data regarding the radial electric field evaluated with the CXS to support the view that the change in the radial electric field would be attributed to increment in the outward flow of the impurity ions from the core region of the LHD plasma.