Experimental research of ECW pre-ionization and assisted startup in EAST

Experimental research on the electron cyclotron wave (ECW) pre-ionization and assisted start-up was carried out systematically for the first time in EAST tokamak, which is a superconducting device with ITER-like full metal wall. Breakdown and plasma initiation at low toroidal electric fields (<0.3 V m−1) with ECW pre-ionization and startup assistance has been demonstrated. Also, the parameter domain of breakdown is significantly extended towards higher prefill gas pressure. The effect of ECW injection timing, power, toroidal injection angle on breakdown were also investigated. Injecting ECW earlier leads to an earlier breakdown and a higher plasma current ramp rate. The electron cyclotron heating (ECH) power threshold for breakdown in EAST is approximately 0.4 MW. In the range of ECH power tested in this work, higher ECH power is advantageous for achieving earlier and faster breakdown. Furthermore, the breakdown with radial ECW injection occurs earlier compared with oblique injections (co-current and counter-current). During the ECW-assisted startup, the process of burn-through is prolonged by the higher pre-filled gas pressure even though it enhances the ease of breakdown. In addition, compared to the low hybrid wave assistance, the ECW assistance has an effect in averting the generation of runaway electrons and improving the safety of device during startup. Moreover, the ECW assistance exhibits a high tolerance to the impurity and thus ensures a high ramp rate of plasma current even with a high impurity level.

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
Plasma initiation is an essential process of the tokamak operation.During the inductive startup commonly used in tokamaks, the change in the current in central solenoids (CS) induces a toroidal electric field for the neutral gas breakdown and plasma current drive [1].The initial breakdown is similar to the Townsend avalanche discharge and requires the toroidal electric field to exceed a certain threshold.However, the change rate of the coil current in superconducting tokamaks is limited due to the superconductivity.In addition, the eddy current in the thick vacuum vessel delays the penetration of poloidal magnetic field.As a result, the upper limit of the toroidal electric field in a superconducting tokamak is relatively low.This limitation is an important issue, particularly for future fusion reactors like ITER.For example, the maximum toroidal electric field in ITER will not exceed 0.3 V m −1 [1][2][3].In this case, it is difficult to achieve a pure inductive breakdown, let alone a plasma startup.Thus, an effective pre-ionization and startup assistance are necessary for plasma initiation at low toroidal electric field.
Pre-ionization is the process of generating seed electrons or a pre-plasma before applying the toroidal electric field to enhance the ease of breakdown.In the burn-through phase, only when the heating power exceeds the total energy loss, a continuous discharge is possible.Thus, a complete ionization of deuterium and impurities is necessary to reduce the energy loss induced by the line radiation.On the other hand, the assistance of auxiliary heating is beneficial to enhance the burn-through and current ramp-up.Electron cyclotron wave (ECW) is a widely used method for pre-ionization and startup assistance in a number of tokamaks.It is also a candidate for ITER.
Recently, experimental study on the ECW pre-ionization and assisted startup was carried out systematically in EAST tokamak, which is a superconducting device with ITER-like full metal wall.By comparing with other pre-ionization methods including high frequency glow discharge (HFGD) and low hybrid wave (LHW), it was found that the ECW was capable of pre-ionization to achieve breakdown at low toroidal electric field (<0.3 V m −1 ).Also, the parameter domain of breakdown was significantly extended towards higher prefill gas pressure with ECW pre-ionization.Moreover, the dependence of breakdown on prefill gas pressure and ECW setups including injection timing, power and toroidal angle were investigated via parametric scan.The tolerance to impurity during burn-through phase with ECW assisted startup was also considered.The rest of this paper is organized as follows.Section 2 describes the experimental setup including the heating systems, diagnostic systems, operational process and other related content.Then, detailed experimental results and analysis are provided in section 3. Finally, a summary of this work is given in section 4.

Introduction to EAST device
EAST is the first fully superconducting tokamak with Dshaped poloidal cross-section and advanced divertor configuration.The design parameters of the device are as follows: major radius R = 1.85 m, minor radius a = 0.45 m, plasma current I p ∼1 MA and toroidal field B T = 3.5 T [20].The top view and cross-section of EAST are shown in figure 1, in which the configuration of the heating and diagnostic systems is depicted.The electron cyclotron resonance heating (ECRH) system on EAST adopting the second harmonics X-mode (X2) scheme consists of four 140 GHz gyrotrons [21].It has an output capability of 0.6 MW during the pre-ionization phase, with toroidal directions from 160 • to 200 • .In the pre-ionization experiment, the toroidal field is set to B T = 2.48 T at the position R = 1.85 m.The X2 resonance layer is located at about R = 1.83 m (indicated by the brown dashed lines in figure 1).Most diagnostics designed for plasma with high parameters are unsuitable for measuring the plasma in the initial phase of tokamak discharges.The diagnostics shown in figure 1 are capable of providing measurements during the ECW preionization phase.The extreme ultraviolet spectrometer and filterscope system are used for monitoring the impurity radiation and deuterium Balmer-α (D α ) line emission, respectively.In addition, the fast CCD camera can provide images of EAST cross-section with the field-of-view depicted in figure 1(a) (green cone).

EAST operational process
A brief introduction to the discharge time series in EAST is given in this subsection.Figure 2 shows the start-up of the discharge in EAST, including pre-ionization and assisted start-up.The inductive electric filed is initiated at t =0 ms, and all event sequences requested to EAST discharge are defined relative to this timing.From t = −3000 ms to t = −500 ms, PF coil currents increases from zero to its respective initial magnetization (IM) current state.PF currents are held at the IM state values from t = −500 ms to 0 ms, during which the vacuum vessel eddy currents dissipate and the field null forms for breakdown with a stray field of <2 × 10 −3 T [22][23][24].The prefill gas is also injected and diffuse evenly in this period.Furthermore, the pre-ionization is applied tens milliseconds ahead of the onset of toroidal electric field.From t = 0 ms to 50 ms, the current of PF coils decreases to induce the toroidal electric field.To get a higher electric field required for breakdown, a set of switchable resistors are utilized in the PF circuit [22].The electric field can be increased from 0.15 V m −1 (without resistors) to 0.55 V m −1 (with the resistors).The duration with the resistors is adjustable in the range of 15-45 ms.With a breakdown achieved, the plasma current driven by the electric field starts to ramp up continuously.It needs to be pointed out that in all discharges with ECW pre-ionization, the ECW pulse duration lasts at least 300 ms to cover the entire pre-ionization and assisted startup phase shown in figure 2. Subsequently, the plasma control system enters the feedback control phase at t = 200 ms [24].

Parametric domain of breakdown in EAST
Figure 3 shows examples of breakdown in EAST with different pre-ionization methods including HFGD, LHW and ECW.Here, the breakdown is defined as the clear appearance of D α emission, specifically as the D α emission reaching a certain intensity level (0.003 V of the raw signal).Here channel number 13 of the D α measurement is selected as shown in  [25].The peaking of the D α emission indicates that at least 50% of the deuterium gas is ionized [9].
It can be seen that the breakdown with ECW pre-ionization happens much earlier than with other two methods.Also, the toroidal electric field at the time of breakdown is lower in the discharge with ECW pre-ionization, which implies a reduction of the critical electric field required for breakdown.It needs to be pointed out that the injected LHW power (∼0.2 MW) is much lower than the target (∼0.4MW) due to poor coupling.Further, the maximum electric field can be adjusted by the duration of the resistor utilization as described in previous section.
Moreover, the process of breakdown is also observed in fast camera images.A comparison of images from discharges with HFGD and ECW pre-ionization is given in figure 4. The central column is on the left in each image.The two discharges shown in figure 4 have the same conditions such as toroidal field, field null configuration, prefill gas pressure and other initial plasma parameters conditions except for the pre-ionization methods.The breakdown with HFGD occurs at ∼27 ms on the high field side.In contrast, the breakdown with ECW is observed at ∼5 ms on the X2 resonance layer of ECW (yellow dashed line), which confirms the effect of ECW pre-ionization.Subsequently, both plasma with HFGD and ECW pre-ionization expand rapidly and closed flux surface forms.It is worth noting that the plasma with ECW pre-ionization get darker after breakdown attributed to the reduction of neutral particles due to the ionization, which is consistent with the evolution of D α emission shown in figure 3.
To provide an accurate domain of the breakdown in EAST, the discharges with different pre-ionization methods are analyzed statistically.As shown in figure 5, the ECW is capable of pre-ionization to achieve breakdown at low toroidal electric field (<0.3 V m −1 , the upper limit of ITER).The data with E t = 0 correspond to the discharges with breakdown before the onset of toroidal electric field, which will be explained in detail later.Also, the parameter domain of breakdown is significantly extended towards higher prefill gas pressure with ECW pre-ionization.

Parametric dependence of breakdown time with ECW pre-ionization
As can be seen that in figure 3, there is a time lag between the ECW application and the appearance of D α emission, which is defined as the breakdown time τ BD hereafter.The time-topeak (TTP) of D α emission is also considered.Both breakdown time and TTP of D α emission are used to evaluate the ease of breakdown and the effect of ECW pre-ionization in this work.The parametric dependence of breakdown on prefill gas pressure and ECW setups including injection timing, power and toroidal angle was investigated via parametric scan.Other parameters were kept constant in each parameter scanning.The results are provided in this subsection.
The dependence of breakdown time on the prefill gas pressure across a range of 10 −4 -10 −2 Pa is depicted in figure 6.The pink arrow shows the trend of the breakdown time (τ BD ) as the prefill gas pressure increases.It demonstrates that a higher prefill gas pressure leads to a shorter breakdown time, implying the enhanced ease of breakdown.According to the Townsend avalanche theory, a higher prefill pressure corresponding to higher neutral gas density may lead to more collisions and a promotion of the avalanche process.
Next, the scanning results in ECW setups are introduced.The impact of ECW injection timing on the breakdown is illustrated in figure 7. It can be observed that an earlier ECW injection results in an earlier breakdown and higher plasma current ramp rate.If the ECW is injected early enough, it is even possible to achieve a non-inductive breakdown with no toroidal electric field, i.e.E t = 0 (#127179).This is similar to the preplasma scenarios in DIII-D and J-TEXT [9,19].If the ECW is injected too late (>−10 ms, indicated by the black dashed line), the rise of D α signal cannot be observed, which implies the failure of breakdown (#127181).
In addition, the effect of electron cyclotron heating (ECH) power on the breakdown was investigated.The result of power scan shows that the lowest ECH power for successful breakdown in EAST is about 400 kW.The dependence of the breakdown time τ BD and the TTP of D α emission on the ECH power is illustrated in figure 8. Here, the τ BD and TTP are the mean values of multiple discharges and the error bars represent the  standard deviations.As ECH power increases, both τ BD and TTP reduce, which implies that higher ECH power is beneficial to earlier and faster breakdown.The impact of toroidal injection angles on the breakdown process was studied through the variation of toroidal injection angles under two P ECH conditions.As shown in figure 9, the reduction of τ BD was achieved with the radial injection.Both co-current and counter-current injection angles prolonged the delay in τ BD , highlighting the effect of radial injection for faster neutral gas ionization.Additionally, increasing P ECH exhibited the same trend.However, strong impurity sputtering on the high-field-side wall was observed in the case of radial injection but can be avoided by adopting the toroidally oblique injection in the pre-ionization phase.It can be explained that the ECW is not well-absorbed by the neutral gas before breakdown and the shine through power density on the high-fieldside wall in the vacuum vessel is highest for the case of radial injection.Thus, the toroidally oblique injection is recommended for pre-ionization to control the impurity level and avoid the damage to the wall.

Startup with ECW assistance
Due to the plasma-material interaction such as sputtering and recycling, the vacuum vessel wall is the impurity source of the plasma.Thus, it is necessary to consider the influence of wall materials in the burn-through phase.In a usual case, as the electron temperature is temporally increasing, the line emission of impurities in higher charge states increases and then plasma burns through.Lithium (Li) coating is a routinely used wall treatment in EAST [26].Accordingly, the Li-III line radiation from Li 2+ ions is taken for instance as the characteristic of burn-through in this work since high charge states of Li must go through this process before the plasma completely burns through low-Z impurities.The time lag between the breakdown and the occurrence of Li-III emission (τ Li−III ) indicates the speed of burn-through.A specific change in a parameter may lead to opposite effects in different phases during plasma initiation.Figure 10 displays the results of burn-through obtained from the same prefill pressure scanning experiment shown in figure 6.The pink arrow shows the trend of the time lag between the breakdown and the occurrence of Li-III emission (τ Li−III ) as the prefill gas pressure increases.In contrast to the breakdown phase, the burn-through phase is prolonged as the prefill gas pressure increases.It can be explained that the time required for ionization increases for more neutral particles at a given heating power.As depicted in figure 10, the upper pressure limit for successful start-up of the device depends on the burn-through phase.For higher prefill gas pressure (>10 −2 Pa), breakdown/avalanche may succeed and maintain a trend of decreasing breakdown delay with increasing prefill gas pressure within a certain pressure range.However, the burn-through process may fail, ultimately resulting in startup failure.
In addition to ECW, LHW is also a commonly used assistance for startup especially on the purpose of current drive.The fast electrons of ∼100 induced by LHW through parallel Landau damping under resonance conditions [27], are further accelerated by the toroidal electric field to several hundred keV and then become runaway electrons in the diagnostic range of the detector [28].The runaway electrons may lead to hot spots on the limiter and cause damage to the vacuum wall.For the sake of the device safety, it is necessary to consider this issue during startup.Thus, the intensities of runaway electrons (evaluated by the corresponding bremsstrahlung strength) during plasma startup with ECW and LHW assistance are compared.The toroidal electric field is also considered due to the tight connection with generation of runaway electrons.Figure 11 shows the results of discharges with ECW and LHW assisted startup.Here, the runaway electron intensities and toroidal electric fields are averaged over 0.1-0.4s in each shot.It can be seen that the runaway electron intensity is positively correlated with the toroidal electric field for LHW assisted startup while keeps almost the same in a range of toroidal electric field for ECW assisted startup.Importantly, compared with LHW assisted startup, the level of runaway electron is significantly lower in discharges with ECW assistance.It is because the ECW increases the vertical velocity of electrons through cyclotron resonance rather than the parallel velocity and does not directly drive fast electrons.This phenomenon proves the effect of ECW assistance in averting the generation of runaway electrons and improving the safety of device during startup.
Another factor that needs to be concerned during plasma startup is the impurity.The carbon is considered in this work.Figure 12 shows the influence of impurity on the plasma current ramp rate (dI p /dt) during startup in discharges with different startup schemes.This analysis corresponds to the time window before 0.2 s since the plasma current is adjusted via feedback control after 0.2 s as shown in figure 2. Here, the level of carbon is estimated by the ratio of the intensity of C III emission to the prefill pressure (C III /p).The plasma current ramp rate and C III emission intensity are averaged over the period from breakdown time to 0.2 s.However, the pressure at breakdown time instead of an averaged value is used since the pressure measurement is not accurate after breakdown.Given that the wall condition varies in a day, the data in the same period (20:00-02:00) during the EAST 2023 campaign are selected to ensure the same wall condition.It can be observed that the higher I p ramp rate was obtained with radio-frequency (RF) wave assistance compared with Ohmic startup.It is probably attributed to the lower resistance due to the higher temperature with auxiliary heating.However, a reliable temperature measurement is not available during startup.In addition, it is necessary to point out that the I p ramp rate are also affected by other factors such as loop voltage, self-inductance of plasma, current drive effect of RF wave and so on.Moreover, ECW assistance can ensure a high I p ramp rate even with a high impurity level, which implies a high tolerance to the impurity of ECWassisted startup.

Summary
Motivated by the requirement for plasma initiation at low toroidal electric field in future tokamaks like ITER, experimental research on ECW pre-ionization and assisted startup was carried out systematically for the first time in EAST.By comparing with other pre-ionization methods including HFGD and LHW, it was found that the ECW was capable of preionization to achieve breakdown at low toroidal electric field (<0.3 V m −1 ).Also, the parameter domain of breakdown is significantly extended towards higher prefill gas pressure with ECW pre-ionization.
The dependence of breakdown on ECW setups including injection timing, power and toroidal angle was investigated via parametric scan.Injecting ECW earlier leads to an earlier breakdown and a higher plasma current ramp rate, potentially achieving a non-inductive breakdown with no toroidal electric field.The ECH power threshold for breakdown in EAST is approximately 0.4 MW.In the range of ECH power tested in this work, higher ECH power is advantageous for achieving earlier and faster breakdown.Furthermore, the breakdown with radial ECW injection occurs earlier compared with oblique injections (co-current and counter-current).
Additional analysis is dedicated to the burn-through phase during the ECW-assisted startup.Although a higher pre-filled gas pressure enhances the ease of breakdown, the process of burn-through is prolonged.It can be explained that the time required for ionization increases for more neutral particles at a given heating power.Thus, the upper limit of the prefill gas pressure for successful startup of the device is determined by the burn-through phase.In addition, a comparison to the LHW assistance proves that the ECW assistance has an effect in averting the generation of runaway electrons and improving the safety of device during startup.Moreover, the influence of impurity during startup was also considered.The ECW assistance can ensure a high I p ramp rate even with a high impurity level, which implies a high tolerance to the impurity of ECWassisted startup.More factors that influence the plasma startup, such as the configuration of magnetic field, will be investigated in the future.

Figure 1 .
Figure 1.The top view (a) and cross-section (b) of EAST.The configuration of the heating and diagnostic systems is shown.The brown dashed lines in (a) and (b) indicate the ECW resonance radius R X2 .

Figure 2 .
Figure 2. The flow chart of operation in EAST plasma initiation phase.

Figure 3 .
Figure 3. Examples of the breakdown with HFGD (blue), LHW (green) and ECW (red) pre-ionization.(a) Auxiliary power Paux, (b) toroidal electric field Et, (c) plasma current Ip, (d) Dα emission intensity.The dashed lines indicate the breakdown times of different pre-ionization methods.

Figure 4 .
Figure 4. Fast camera images from discharges with HFGD and ECW pre-ionization.The yellow dashed lines indicate the X2 resonance layer of the ECW in both HFGD case and ECW case for a straight comparison of breakdown positions.

Figure 5 .
Figure 5. Parameter domain of EAST breakdown with different pre-ionization methods.Breakdown at low toroidal electric field Et (<0.3 V m −1 , the upper limit of ITER, indicated by the black dashed line) and significant extension of prefill gas pressure p were achieved with ECW pre-ionization.

Figure 6 .
Figure 6.Dependence of the breakdown time τ BD on prefill pressure.The dashed guide line shows the trend of the breakdown time (τ BD ) as the prefill gas pressure increases.

Figure 7 .
Figure 7.The temporal evolution of several key signals in breakdown phase with different ECW injection times.(a) ECH power P ECH , (b) toroidal electric field Et, (c) plasma current Ip, (d) Dα emission intensity.The black dashed line indicated the latest ECW injection time for the achievement of breakdown.

Figure 8 .
Figure 8.Effect of P ECH on the breakdown time τ BD and TTP of Dα emission.The τ BD and TTP are the mean values of multiple discharges.The error bars represent the standard deviations.

Figure 9 .
Figure 9. Dependence of the breakdown time τ BD on the ECW toroidal injection angle.

Figure 10 .
Figure10.Impact of the prefill gas pressure p on the time lag between the breakdown and the occurrence of Li-III emission (τ Li−III ).The dashed guide line shows the trend of the time lag as the prefill gas pressure increases.

Figure
FigureThe runaway electrons intensity versus toroidal electric field Et,m in discharges with ECW and LHW assisted start-up (averaged over 0.1-0.4s in each shot).

Figure 12 .
Figure 12.The influence of impurity on the plasma current ramp rate (dIp/dt) during startup in discharges with different startup schemes.