Hot spots induced by RF-accelerated electrons in the scrape-off layer on Experimental Advanced Superconducting Tokamak

Preventing impurity emission from hot spots on plasma-facing materials is a critical issue in the maintenance of high-performance plasma on the Experimental Advanced Superconducting Tokamak (EAST). In this study, experimental and theoretical analyses were performed to investigate the mechanism of hot spot formation. In the upper single null magnetic configuration of the EAST, two separatrices were connected to the upper (primary) and lower (secondary) X-points. Experiments on plasma configuration control indicated that the reduction in the gap between the lower (secondary) separatrix and lower hybrid antenna is effective in preventing hot spot formation on the lower divertor, which frequently emits impurities in long-duration discharges. This effectiveness was quantitatively confirmed by magnetic field lines tracking simulation and calorimetric measurement of divertors in the experiment. Two-frequency power modulation of the lower hybrid wave (LHW) was conducted to evaluate power deposition on the scrape-off layer (SOL) during propagation from the LHW antenna to the main plasma. This experiment clarified that LHW-accelerated electrons in the SOL via collision damping deliver their energies to hot spots along the magnetic field line. These findings help alleviate or even eliminate the formation of hot spots and maintain the performance of plasma.


Introduction
Achieving a steady-state operation of magnetic fusion devices is one of the goals of fusion research and has been attempted on several tokamaks, such as JET [1], Tore Supra [2], JT-60U [3], TRIAM-1M [4], KSTAR [5], and QUEST [6], with excellent results. Hot spots are the outcomes of interactions between the plasma and plasma-facing components (PFCs), which frequently hinder the maintenance of high-performance plasmas through unwanted impurity emissions, such as the longest Hmode discharge on the Experimental Advanced Superconducting Tokamak (EAST) [7]. Moreover, hot spots can cause the meltdown of PFCs, which results in serious damage to the machine.
Hot spots induced by various radio frequency (RF) power injections have been observed and investigated in numerous devices. In JET, hot spots were observed during the ion cyclotron resonant frequency (ICRF) injection on the ICRF antenna as well as on the limiters close to the power antenna [8], which was explained by the enhanced heat loads attributed to the acceleration of ions in RF-rectified sheath potentials [9]. A considerable fraction of the heat load on the first wall is supplied by the direct loss of fast electrons, owing to the large deviation from the magnetic surface in the electron cyclotron current drive plasma in QUEST [10]. During argon gas puffing experiment in the EAST, a decrease in the heat flux at the strike point with a corresponding increase at the adjacent hot spot location was observed on the lower outer divertor plate, which was explained by the fact that argon gas puffing can increase the lower hybrid wave (LHW) absorption in the scrape-off layer (SOL) [11]. An enlightening study was performed on high-harmonic fast-wave (HHFW)-heating-dominated plasma in NSTX, which showed that the power of HHFW couples across the entire width of the SOL, and the magnetic field lines can be connected from the SOL in front of the power antenna to the hot spots [12]. These findings indicate that hot spots may be formed by electrons moving along the magnetic field lines that attack the PFCs from the entire SOL.
As one of the effective methods for electron heating and plasma current driving, LHW has been extensively applied to long-pulse operations in numerous tokamaks [13]. Hot spot formation has been frequently observed in many LHW experiments, such as in Tore Supra [14], TdeV [15], JET [8,16], and EAST [11,17]. Hot spots can result in extremely high heat loads (∼10 MW m −2 ) on their PFCs. Regarding the LHW power deposition on the SOL, three candidate mechanisms have been considered.
The first mechanism pertains to electron Landau damping [18], wherein high parallel refractive index (N ∥ ) components of the incident LHW can couple with the plasma close to the LHW antenna. The heated electrons interact with the lower parallel refractive index components of the incident LHW, which can be heated further. Finally, low-energy (typically in the order of 20 eV) electrons in the SOL can be heated to several keV [19]. Fast electrons generated by the parasitic absorption of high N ∥ LHW spectrum have been proven to be the reason for the formation of hot spots on the grill limiter of Tore Supra [20]. A radial deposition width of several millimeters can be quantitatively calculated with self-consistent particle-in-cell simulations.
The second mechanism of LHW power deposition is collision damping [21], which typically occurs in plasma edges at low temperatures. It is a significant part of wave damping because some or even most of the LHW power is prevented from reaching the main plasma region, where Landau damping is expected. The effect of collision absorption on the efficiency of the lower hybrid current drive (LHCD) can be analyzed using experimental data combined with the ray-tracing/Fokker-Planck model, which explains the phenomenon of LHCD experiments on FTU [21].
The third mechanism can be explained through parametric decay instability (PDI) [22], which leads to a spectral broadening of the launched wave, resulting in a more peripheral power deposition profile through the first two mechanisms. By integrating the PDI effect into the ray-tracing/Fokker-Planck code, more accurate simulations of the LHW deposition profile can be obtained, which help interpret the longlasting internal transport barriers sustained by the LHCD on JET [23].
The aforedescribed studies have remarkably enriched the theory of RF-wave absorption in SOL and the mechanism of hot spot formation. These studies stated that two factors, namely, the SOL parameters and RF power deposition in the SOL, play an important role in the hot spot formation in longduration discharges. The SOL parameters affect the absorption of RF power in the SOL, which is an essential energy source for hot spots. Magnetic configuration is an important SOL parameter; however, its effect on hot spot formation has rarely been researched. Preventing the formation of a hot spot is highly crucial for a longer pulse operation of plasma, and its controllability is indispensable to success. In the present study, the effect of the SOL magnetic configuration on hot spot formation was investigated in a series of configuration-control experiments, and the magnetic field line tracking in SOL contributed to an intuitive explanation. The findings of this study helped achieve the longest H-mode plasma [24]. Moreover, the energy source of hot spots and the mechanism of hot spot formation were clarified through two-frequency LHW power modulation experiments. These experiments showed that collision damping of LHW is a major mechanism for LHW power deposition on the SOL, which subsequently leads to hot spot formation. These results were confirmed through power deposition calculations and a joint simulation between raytracing and Fokker-Planck codes.
The remainder of this paper is organized as follows: the experimental apparatus and related diagnostics systems used in the study are described in section 2. The formation of a hot spot and its relationship with the plasma configuration in the SOL are discussed in section 3. Heat load analysis of the divertor and the mechanism of hot spot formation in the LHW power modulation experiment are described in section 4. Finally, the concluding points are summarized in section 5.

Experimental apparatus
Experiments were performed in the EAST, with major and minor radii ranging from 1.7 to 1.9 m and 0.4 to 0.45 m, respectively. The EAST aims to demonstrate stable, longpulse, and high-performance H-mode plasmas with ITER-like configuration and operation schemes; its main chamber was covered with molybdenum, the upper divertor was covered with ITER-grade mono-block tungsten, and the lower divertor was covered with carbon. The EAST has 16 equatorial ports (port A to P in a toroidal direction), as shown in figure 1. There are two sets of LHW systems in the EAST, with frequencies of 2.45 GHz (LHW 1) and 4.6 GHz (LHW 2), located in ports N and E, as shown in figure 1. To cool the PFCs, the EAST uses a pressurized water system, which consists of five water-cooling modules, as shown in figure 2. Calorimetric diagnostics was used to measure the absolute value of heat load on each module with the temperature increment and flux of cooling water [25].
To observe the heat load on the PFCs with improved spatial and time resolutions compared with the calorimetric measurement, a visible-light camera (port D) and an infrared (IR) camera (port G) were used; the installation position and the field of view (FoV) are shown in figure 1. The frame rate of the visible-light camera was in the range of 20-120 Hz, and the spatial resolution was less than 10 mm. The IR camera can provide information regarding the surface temperatures of the viewed PFCs [26]. The spatial resolution in the divertor region was ∼4 mm in large FoV cases, and the frame rate was 115 Hz in the full-frame mode but could be increased up to several kHz by reducing the FoV.

Analysis of hot spot formation in the upper single null (USN) configuration discharges in the lower divertor
During the 2017 campaign, high-performance plasmas were generated in the EAST [27]; however, unwanted hot spots often appeared in the lower divertor. Before the steady-state H-mode plasma discharge #73 999 of duration 101.2 s, several discharges were performed to identify the optimal magnetic configuration. Typical discharge attempts were those from #73 991 to #73 998 (apart from #73 996, because unexpected plasma termination occurred at~3.7 s). These seven discharges were all maintained for ∼9.6 s, with the ion B × ∇B drift toward the lower X-point, and almost contained the same experimental parameters, as shown in figure 3 (used #73 993 and #73 994 as examples). The plasma current of these seven discharges was ∼400 kA (figure 3(a)). The loop voltage was well-controlled to a value that was almost equal to 0 V (figure 3(b)), meaning that almost all the plasma current was driven noninductively. The line-averaged electron density at 6 s was ∼3.0 × 10 19 m −3 (figure 3(c)), as measured using the POlarimeter-INTerferometer diagnostics system (POINT) [28] in the mid-plane. The auxiliary heating power comprised the following parameters: ∼0.43 MW (LHW 1), ∼1.48 MW (LHW 2), ∼0.35 MW (electron cyclotron resonance heating [ECRH]), and ∼0.12 MW (IRCF) heating (figure 3(d)). The injection power values and duration times of all auxiliary heating systems of all seven discharges were the same, except for the heating power of LHW 2 in discharge #73 998, which was ∼0.01 MW higher than that of the other six discharges. The stored energy at 6 s was ∼1.30 × 10 5 J (figure 3(e)). Hot spots appeared at the same location in discharges #73 991, #73 992, and #73 993 on the lower divertor of port C, as observed using the visible-light camera. However, hot spots did not appear at

Analysis of heat loads on the PFCs with and without hot spots
The data from the calorimetric diagnostics can be used to study the relationship between hot spot formation and heat load distribution on the PFCs. The energies eliminated by the cooling water in modules A and E for discharges #73 993 and  [7]). The data of calorimetric diagnostics can be used to estimate the additional heat load on the lower divertor. The hot spots appeared in about 3-8 s of the discharge #73 993, and the energy difference between the discharges #73 993 and #73 994 in the case of module E was 3.82 MJ; thus, the additional deposited power on the lower divertor was approximately equal to 0.76 MW. As the FoV of the camera could not cover the entire lower divertor, the number of hot spots in the entire lower divertor could not be determined, and the heat flux on each hot spot could not be estimated. However, the appearance of hot spots may cause a local excessive heat flux that considerably affects the distribution of heat load on the PFCs.

Main parameters for hot spot formation
The plasma configuration in the SOL plays an important role in determining the magnitude and distribution of the heat load on the first wall because the particle outside the last closed flux surface (LCFS) moves along the magnetic field line and ultimately attack the PFCs; exceptions are the particles and energy transfers across the magnetic field line. The configuration from #73 991 to #73 999 (excluding #73 996) is the typical USN divertor configuration in the EAST, as shown in figure 5. However, a few conceptions related to the plasma configuration are introduced and explained as follows: First, we define the distances dR sep = R sep L − R sep U , where R sep L and R sep U are the lower and upper separatrix radii, respectively, mapped to the outer mid-plane [29]. For the USN, the upper X-point is the primary X-point at the intersection of the upper separatrix; thus, the upper and lower separatrices are the primary and secondary separatrices, respectively. Therefore, the value of dR sep is positive in the case of USN. The primary and secondary strike points denote the intersections of the divertor target plate and the magnetic field lines along the primary and secondary separatrices, respectively. Therefore, in the USN configuration, two primary strike points lie on the upper outer and inner targets, and two secondary strike points lie on the lower outer and inner targets. The gapout is defined as the distance between the primary separatrix and the first wall in the mid-plane on the low-field side. dR mid represents the distance between the secondary separatrix and the first wall in the mid-plane on the low-field side. A schematic of these conceptions is presented in figure 5. The relationship between the gapout, dR sep , and dR mid can be expressed as follows: In the lower single null, the lower and upper X-points are the primary and secondary X-points, respectively, and dR sep is less than zero. For the double null, the lower and upper separatrices are almost coincident, and dR sep is approximately equal to zero. In the EAST magnetic configuration, the secondary X-point is present inside the first wall and alters the distribution of the magnetic field lines near the secondary X-point.
Compared with the plasma configuration (figure 5) and hot spot position (figure 4), we observed that the hot spot location was close to the lower outer strike point; thus, it is necessary to investigate the relationship between the hot spot formation and location of the strike point.
Stable plasma configurations of these seven discharges were established from ∼2.3 s using the Equilibrium Fitting (EFIT) code [30]; at times before 2.3 s, the configuration changes of these seven discharges exhibited the same trends, and hot spots appeared from ∼3 s onwards. Therefore, the temporal evolution (from 3 s to 8 s, at every second) of the positions of the lower outer strike points for #73 991 to #73 998 (except for #73 996) were calculated based on the EFIT reconstruction, as shown in figure 6. The red and blue points in figure 6 represent the heights of the lower outer strike points of these seven discharges in the cylindrical coordinates. Notably, the hot spot and lower outer strike point (from 3 s to 8 s) of all these discharges were on the outer dome of the lower divertor. Figure 6 shows that the distance between the hot spot and lower outer strike point on the divertor had no obvious relationship with the hot spot formation in these discharges, because the position of the lower outer strike point for discharges #73 994 and #73 995 (without hot spots) was almost the same as the discharges with the hot spot.
The time evolutions of the magnitudes of the gapout, dR sep , and dR mid from #73 991 to #73 998 (without #73 996) are shown in figure 7. The discharges with hot spot appearances could not be distinguished from these discharges based on the gapout in figure 7(a) but could be distinguished based on dR sep and dR mid in figures 7(b) and (c), respectively. Hot spots appeared at smaller dR sep and larger dR mid values. The EAST has the ability to control the plasma configuration by adjusting the magnitude of the gapout and dR sep (excluding dR mid ) on a millimeter scale; thus, the difference in the magnitude of dR mid with and without a hot spot in figure 7(c) is not as clear as that of dR sep in figure 7(b). This figure suggests a possible relationship between the hot spot formation and plasma configuration in the SOL. As the electron in the SOL moves along the magnetic field lines, magnetic field line tracking in the SOL was performed to explore the mechanism of hot spot formation in the experiments described above; the results are shown in figure 8. The EFIT code provided the equilibrium magnetic configurations of #73 993 at 6 s required by the tracking. As observed in figure 8, the electrons that moved along the red magnetic field lines (corresponding to the dR mid area) were able to attack the lower divertor, and caused the hot spot formation.
Backward tracking of magnetic field lines from the hot spot area for discharge #73 993 was also performed, as shown in figure 9. The simulation was set to stop automatically when the magnetic line was tracked in front of the LHW antenna in the SOL. Figure 9 shows that the magnetic field line could be backward-tracked to the dR mid area and was finally stopped in front of the LHW 1 and LHW 2 antennas. The end points of the magnetic lines (black dots in figure 9(a)) have broad poloidal and toroidal distributions in front of the LHW antenna. This indicates that electrons in a large geometric space in front of the LHW antenna may deposit the LHW power, and then, the accelerated electrons attack the lower divertor to cause hot spot formation. The onsets of the two tracked magnetic field lines in figure 9(b) were only ∼6 mm apart on the hot spot area on the lower divertor, although the lines differed considerably. Figure 4 shows that the hot spots appearing in these discharges were toroidally localized. This can be attributed to the existence of small bulges in a few lower divertor tiles that lead to a larger contact area with the magnetic field lines; accordingly, more electrons could attack these tiles along the magnetic field lines, resulting in excessive local heat loads on these tiles.

Appearance of the hot spot on the lower divertor of discharge #73 999
Based on the optimization of the configuration of experiments described above, the hot spot on the lower divertor was The magnetic field lines in blue represent tracking in the forward electron direction and the onset between the primary and secondary separatrices (corresponding to the area of dRsep). The magnetic field lines in red represent tracking in the forward electron direction and the onset between the secondary separatrix and the first wall (corresponding to the area of dR mid ). The magnetic field lines in green represent tracking in the backward electron direction and the onset between the primary and secondary separatrices (corresponding to the area of dRsep). The magnetic field lines in pink represent tracking in the backward electron direction and the onset between the secondary separatrix and the first wall (corresponding to the area of dR mid ). alleviated effectively. Therefore, the EAST achieved a longpulse (101.2 s) H-mode discharge, #73 999, in the 2017 campaign. However, the zero drift of the magnetic measurement  system resulted in small, persistent changes of the plasma configuration during discharges, as shown in figure 10. Therefore, redistribution of the heat load occurs between different PFCs, as shown in figure 11(b) in the previous study [7]. The red curve of this figure indicates that, after the flat-top phase, the energy eliminated from module A started to decrease at ∼55 s, suggesting that the heat load on the upper divertor started to decrease at this moment. Subsequently, the reduced power was deposited on the lower divertor and caused hot spot formation, as shown in figure 11 obtained using the visible-light camera. This photograph shows that, after ∼55 s, hot spots at the lower divertor gradually became brighter. The temporal evolutions of the magnitude of the gapout, dR sep , and dR mid for discharge #73 999 are shown in figure 10. As indicated, dR mid increased considerably with time; this proves that dR mid is crucial for hot spot formation, as analyzed in section 3.2.

Hot spot formation in LHW modulation experiments
The analysis described in section 3 indicated that hot spots are generated because electrons in the SOL move along the magnetic field line and ultimately attack the lower divertor. However, the energy source of electrons in the SOL remains unclear. Therefore, an LHW modulation experiment (discharge #74 864) was performed.

Experimental results and analysis
The L-mode discharge #74 864 was USN, with the ion B × ∇B drift directed toward the lower X-point. Some important plasma parameters of discharge #74 864 are shown in figure 12. The plasma current was ∼400 kA ( figure 12(a)). The electron density was well-controlled during the discharge flat-top phase; at 5 s, the line-averaged electron density in the mid-plane was ∼2.23 × 10 19 m −3 ( figure 12(c)). The auxiliary heating power comprised two components, namely ∼0.52 MW attributed to the LHW at 2.45 GHz and ∼0.58 MW attributed to the LHW at 4.6 GHz. The duty ratios of the LHWs at 2.45 GHz and 4.6 GHz were ∼50%, and each heating period lasted ∼2 s ( figure 12(d)). The phases of injections of the LHWs at 2.45 GHz and 4.6 GHz were 325 • and 90 • , respectively, and each power spectrum was simulated using the ALOHA code [31], as shown in figure 13. During the LHW injection, the plasma stored energy increased considerably ( figure 12(e)), thus indicating that the LHW can heat the plasma effectively. In addition, during the LHW injection, the plasma loop voltage decreased ( figure 12(b)), indicating that the LHW can effectively drive the noninductive current in the main plasma through LHCD. However, the average plasma loop voltage during the first LHW heating period at 4.6 GHz was ∼0.2 V, that is, ∼0.06 V lower than that of the first LHW heating period at 2.45 GHz. This finding suggests that the LHWs at 2.45 GHz and 4.6 GHz exhibited different performances during this discharge, such that more LHCD occurred in the main plasma during the LHW injection period at 4.6 GHz.  Figure 14(a) is a photograph acquired using the IR camera in port G; the LHW antenna at 4.6 GHz and its two graphite guard limiters are on the left side of the figure. To research the heat load on the divertors during this discharge, four representative zones were chosen as shown in figure 14(a); where zones 1 and 2 with and without hot spots, respectively, were on the lower outer divertor; zones 3 and 4 were on the upper outer and inner divertor, respectively. The hot spot in zone 1 is dangerous and of interest, and zone 2 without hot spot provides a comparison. In this USN discharge, the upper divertor would stand much higher heat load than the lower divertor, therefore zones 3 and 4 are also selected for the study. Figure 14(b) represents the average temperature evolution, and figures 14(c)-(f ) shows the time derivatives of the temperature of the four zones over time. Figure 14(b) shows that the temperature in zone 1 was strongly related to the injection of the LHW at 2.45 GHz; however, after the second heating  period of the LHW at 2.45 GHz, the temperature in zone 1 continuously decreased, weakly related to the injection of the LHW at 4.6 GHz. This indicates that the hot spot formation in zone 1 was most likely related to the different performances of the two LHWs. This finding is investigated in the next section.
The temperatures in zone 3 and 4 changed simultaneously with the injection of the LHWs at 2.45 GHz and 4.6 GHz, indicating that the heat load of the upper divertor in the USN configuration discharge originated from the core plasma. Because the time derivative of temperature is directly related to heat flux, figures 14(e) and (f ) show that the upper divertor withstand a large heat flux during the LHW injection.
Moreover, the location of the hot spot on the lower divertor in zone 1 also withstand a large heat flux compared with that in zone 2 during an LHW injection at 2.45 GHz, as shown in figures 14(c) and (d).

Analysis of LHW power deposition
Previous studies on several tokamaks have demonstrated that the LHW and LHCD are among the most efficient methods for electron heating and noninductive current driving for the steady-state operation of the device [32]. The LHW transfers energy and momentum to the electrons, primarily based on the Landau damping mechanism. According to the quasi-linear theory, electron Landau damping [33] occurs when ω/k ∥ ⩽ 3v T , where v T = (2T e /m e ) 1/2 ; the corresponding equation can be written as follows: The power density spectra of the LHWs at 2.45 and 4.6 GHz for this discharge are shown in figure 13. Therefore, Landau damping occurs when the electron temperature is high (values of the order of several keV). When the electron temperature is low, the LHW can deposit energy through collision damping. Notably, the initial spectrum of the LHW may be modified by the PDI effect when LH rays cross the SOL before they reach the main plasma.
In the case of discharge #74 864, the hot spot formation in the lower divertor was only related to the injection of the LHW at 2.45 GHz. It is reasonable to consider the PDI effect initially, which can be analyzed using the frequency spectrum broadening of these two LHWs and measured with an RF loop probe installed outside the vacuum vessel of the EAST. This probe was placed far away from the LH antenna and aimed to ensure that the signal was detected after the interaction between the waves and plasma and not directly by the LH antenna. The frequency spectra pertaining to the LHW at 2.45 GHz (blue) and that of the LHW at 4.6 GHz (red) are shown and compared in figure 15. The spectral widths of the two LHWs were almost the same, meaning that the PDI effect is not the primary cause of hot spot formation.
According to the conclusions drawn in section 3, the appearance of a hot spot in the lower divertor is related to the movement of electrons along the magnetic field lines in the SOL in front of the LHW antenna. The electron temperature in the SOL was relatively low; therefore, the difference in the collision effect during the injections of the LHWs at 2.45 and 4.6 GHz should be considered. Collision damping caused by the LHW [34] can be expressed using equation (3): where ν ei ∝ n e /T 3/2 e is the electron-ion collision rate, and ω = 2πf is the LHW circular frequency; for LHW 1, f = 2.45 GHz, and for LHW 2, f = 4.6 GHz. Additionally, ω ce = eB/m e is the electron cyclotron frequency, and ω pe = ( n e e 2 /m e ε 0 ) 1/2 is the electron plasma frequency. Using the typical parameters of the SOL in the EAST implies that n e = 2. pe ω 2 ≈ 9.52 (for 4.6 GHz); the values of N ∥ (near the main probe) and the corresponding N ⊥ were of the same order. Therefore, compared with the second term, the first term in equation (3) is small enough and can be ignored. Thus, Equation (4) shows that the frequency of the LHW considerably affects the collision damping. Assuming that all the other parameters are the same, the collision effect induced by the LHW at 2.45 GHz is ∼6.6 times greater than that of the LHW at 4.6 GHz. The collision primarily occurs in the low electron temperature region (SOL). Therefore, we can assume that, compared with the LHW at 4.6 GHz, the LHW at 2.45 GHz deposits more energy in the SOL through the collision effect and leads to the generation of a hot spot on the lower divertor. The localized hot spot was probably caused by the small bulges in some lower divertor tiles, as explained in section 3. Furthermore, compared with the LHW at 2.45 GHz, the LHW at 4.6 GHz deposited more energy in the main plasma; this resulted in higher LHCD and lower loop voltage, as mentioned in section 4.1. This analysis was verified through the simulation of GENRAY/CQL3D raytracing/Fokker-Planck code package [35,36], using 3.4 s (LHW injection at 2.45 GHz) and 7.4 s (LHW injection at 4.6 GHz) as examples.
The electron densities at these two moments were measured based on the combination of POINT diagnostics (for core plasma) and reflectometry (for edge plasma). The electron temperature data were obtained using a combination of the heterodyne radiometer (X-mode ECE) (for core plasma) and the Thomson scattering system (for main plasma). Moreover, the electron density and temperature outside the LCFS were estimated using the equations listed below [37]: n e (ρ) = n e, LCFS exp T e (ρ) = T e,LCFS exp where λ ne = 0.05 and λ te = 0.02 are experience values of the EAST [38]. The simulation results from GENRAY/CQL3D are listed in table 2. We calculated that the collision absorption caused by the LHW at 2.45 GHz was 35.2 kW higher than that of the LHW at 4.6 GHz. This result tended to be similar to the experimental observation of the surface temperature of the hot spot in zone 1 when the two frequencies of LHW injection, as shown in figure 14(b).

Summary
Hot spots are caused by the excessive localized heat load on the PFCs. Although hot spot formation must be avoided, hot spots often appear on the divertors, main limiter, and LHW guarding limiter during discharges in the EAST. In this study, we investigated the mechanisms underlying the hot spot formation on the lower divertor during the EAST 2017 campaign and found that two factors, namely the magnetic configuration in the SOL and the energy deposition of the LHW in the SOL, played an important role in hot spot formation. By slightly changing the SOL plasma configuration, the hot spot on the lower divertor was effectively alleviated. This helped achieve a record of steady-state H-mode plasma (#73 999) in the EAST for 101.2 s. Experiment analyses and GENRAY/CQL3D simulations showed that the two LHWs had different plasma heating and current driving efficiencies in the main plasma owing to their different frequencies; the LHW at 2.45 GHz likely deposits more energy than the LHW at 4.6 GHz in the SOL through collision damping, which leads to hot spot formation in the LHW modulation experiment.
The findings of this study provide two primary insights. First, reduce the auxiliary heating power (such as LHW) deposition in the SOL. This part of the energy directly contributes to the heat load of related PFCs that are magnetically connected from the SOL. Therefore, efforts to increase the auxiliary heating power deposition in the core plasma for heating and current driving are needed. This will improve the confinement performance of the plasma and reduce the cost-of-energy of the fusion device. Second, the heat load between different PFCs can be redistributed by slightly adjusting the magnetic configuration in SOL. This can be used to alleviate excessive heat load on a particular PFC, thus avoiding problems, such as impurity emission and subsequent plasma degradation.
In the near future, the EAST aims to realize a higher performance plasma operation with higher power injection; therefore, it is important to achieve full FoV detection for PFCs on the EAST to monitor the condition of PFCs, including hot spots. Besides, the mechanism of association between the LHW and plasma should be further investigated to improve the efficiency of plasma heating and current driving through the LHW to achieve lengthier H-mode operations and noninductive current driving. These results have potential applications in exploring advanced operating modes on CFETR in China and ITER.