Gyrotropy-Induced Symmetry Breaking of Erosion Hot Spots on Antenna Limiters during High-Power Tokamak RF Operation: Mechanisms and Mitigation Strategies. Preprint: APS/123-QED W. Tierens, A. Kumar, J. Lore, G. Urbanczyk, R. Diab, The WEST Team Abstract Ion Cyclotron Range of Frequencies heating (ICRH) and current drive will be essential for sustaining high-performance plasmas in next-generation fusion devices (e.g. ITER, SPARC). ICRH actuators routinely produce localized hot spots on limiters and nearby components, posing serious risks to antenna reliability, material survivability, and overall plasma performance. Remarkably, these hot spots are often strongly asymmetric, even with nominally symmetric plasma conditions and antenna geometries. We show that such asymmetries exist intrinsically in the wave physics rather than solely being due to misalignment or edge plasma variation. Our results strongly suggest that this asymmetry can be compensated for by using either poloidal phasing control (which e.g. the under construction WEST traveling wave antenna can do) or modified limiter shapes, suppressing peak sputtering by a factor ∼ \sim 3 and reducing total erosion by a factor of ∼ \sim 2 compared to state-of-the-art designs. This capability is essential for sustaining high-power, long-duration ICRH operation in reactor-scale devices and other next-generation fusion systems. By distributing heat and particle fluxes more evenly across antenna surfaces, optimized limiter shaping provides a clear pathway to robust, reliable ICRH performance in existing and planned fusion devices. Notice: This manuscript has been authored by UT-Battelle, LLC, under contract DE-AC05-00OR22725 with the US Department of Energy (DOE). The US government retains and the publisher, by accepting the article for publication, acknowledges that the US government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for US government purposes. DOE will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan ( https://www.energy.gov/doe-public-access-plan ). Ion Cyclotron Range of Frequencies Heating (ICRH) is a leading technique for plasma heating and current drive in magnetic confinement fusion, enabling heating the plasma to reactor-relevant conditions Garcia et al. 2024 ; Creely et al. 2020 . Other applications include plasma-based space propulsion systems Takahashi 2022 ; Arefiev and Breizman 2004 and the study of plasma-material interactions Rapp et al. 2019 . In fusion, ICRH contributed 20% of the record 59MJ fusion energy produced during the 2022 deuterium-tritium campaign at JET Maslov et al. 2023 . These results underscore the importance of understanding and controlling RF wave–plasma interactions. In present-day fusion devices, ICRH performance depends critically on minimizing impurity generation and damage to plasma-facing components through optimized antenna design and operation. In particular, careful control of the antenna phasing Bobkov et al. 2019 ; Diab et al. 2025 has proven effective in reducing material sputtering and impurity sources, making ICRH compatible even with the high-Z environment Bobkov et al. 2016 of most planned fusion devices Barabaschi et al. 2025 . However, emerging experimental evidence and modeling indicate that these mitigation strategies may not fully capture certain asymmetries in the wave physics. Figure 1: Sputtering yields during a power scan from 0MW through 1.7MW on the WEST antenna limiters (blue through cyan: left limiter, orange through red: right limiter), computed from the Tungsten ( \Gamma_{W} ) and Deuterium ( \Gamma_{D} ) fluxes observed using visible spectroscopy. Both vertical asymmetry (neither the blue nor the red curves are vertically symmetric) and horizontal asymmetry (the red and blue curves have different height-dependence) are readily apparent. Recent spectroscopic measurements of Tungsten and Deuterium fluxes on the WEST tokamak (figure 1 shows data from shots #61683 through #61689), and similar measurements on AUG Bobkov et al. 2009 and EAST Zhang et al. 2022 show highly asymmetric sputtering patterns. Related asymmetries were also observed on NSTX Swain et al. 2003 . This is consistent with numerical results Tierens et al. 2024 ; Kumar et al. 2025 ; Ragona et al. 2025 . In this Letter we show that an unavoidable asymmetry is inherent in the wave physics, even when other causes such as the antenna geometry, plasma shape, antenna-plasma misalignment, asymmetric excitation of the antenna ports, and the preferential poloidal direction of motion of convective cells, have been considered. This limits the effectiveness of symmetric or flux-surface-aligned limiters. We propose that non-aligned limiters which account for the wave physics may outperform the current state of the art. Figure 2: The symmetric plasma density in front of the symmetric antenna. Left: toroidal cross section, right: poloidal cross section. Numerically, it is possible to disentangle the various sources of asymmetry. We have constructed a fully symmetric WEST-like finite element antenna model based on that used in Tierens et al. 2024 , removing all asymmetric elements, and ensuring the bottom half is the mirror image of the top half. We then consider a perfectly symmetric plasma using the density from figure 2 , and a purely toroidal magnetic field (no tilt angle). We include sheath rectification physics using the dielectric layer sheath model Beers et al. 2021 , which gives us both an RF potential V_{RF} and a DC potential V_{DC} on the limiters. We excite the antenna in dipole, with the same power on the top and bottom ports. Even in this maximally symmetric scenario, numerical results still predict up-down asymmetries in the potential and the sputtering yield, as shown in figure 3 . We computed the sputtering yield from the potential using a WEST-typical sputtering curve as in Tierens et al. 2024 . This up-down asymmetry is robust to small changes in the vertical location of the plasma and small changes in the magnetic field tilt angle. Only by turning the confining magnetic field all the way around (reversing the direction of the currents in the toroidal field coils), which is equivalent to mirroring the situation since B is a pseudovector, do the hot spots end up on the bottom instead. Figure 3: Left: Numerically predicted RF potential at 1MW coupled power, and (right) corresponding sputtering yield in a completely symmetric WEST-like antenna. Despite the symmetry, hot spots preferentially appear at the top. The total (integrated) sputtering yield is 5.2\cdot 10^{17} atoms per second. This remaining asymmetry is inherent in the wave physics itself. Recall the electromagnetic frequency-domain wave equation in Cartesian coordinates \displaystyle\nabla\times\nabla\times\bm{E}-k_{0}^{2}\varepsilon\bm{E}=0 (1) where k_{0}=\omega/c is the vacuum wavenumber, \omega is the wave (angular) frequency, c is the speed of light, and the cold plasma dielectric tensor (in the absence of a magnetic field angle, with the confining magnetic field along z ) is \displaystyle\varepsilon=I-\sum_{s}\left[\begin{matrix}\frac{\omega_{s}^{2}}{\omega^{2}-\Omega^{2}}&\frac{\Omega_{s}}{-i\omega}\frac{\omega_{s}^{2}}{\omega^{2}-\Omega^{2}}&0\\ \frac{\Omega_{s}}{i\omega}\frac{\omega_{s}^{2}}{\omega^{2}-\Omega^{2}}&\frac{\omega_{s}^{2}}{\omega^{2}-\Omega^{2}}&0\\ 0&0&\frac{\omega_{s}^{2}}{\omega^{2}}\end{matrix}\right] (2) where I is the unit matrix, and \omega_{s},\Omega_{s} are the plasma frequency and the cyclotron frequency of particle species s , respectively. The presence of off-diagonal elements in ( 2 ) is referred to as gyrotropy, which vanishes in the highly magnetized limit Bécache et al. 2017 . Assuming E_{z}\approx 0 (fast wave polarisation), \bm{E}=[E_{x}(x),E_{y}(x),0]^{T}\exp(i(n_{x}x+n_{y}y+n_{z}z)\omega/c) , and \varepsilon=\varepsilon(x) , we may derive an approximate expression for E_{y}(x) , proceeding along the lines of Swain et al. 2003 \displaystyle\frac{A}{k_{0}^{2}}\frac{d^{2}E_{y}}{dx^{2}}+\frac{B}{k_{0}}\frac{dE_{y}}{dx}+CE_{y}=0 (3) with \displaystyle A \displaystyle=\frac{n_{z}^{2}-\varepsilon_{11}(x)}{n_{y}^{2}+n_{z}^{2}-\varepsilon_{11}(x)} \displaystyle B \displaystyle=-n_{y}^{2}\frac{\frac{d}{dx}\varepsilon_{11}(x)}{k_{0}(n_{y}^{2}+n_{z}^{2}-\varepsilon_{11}(x))^{2}} \displaystyle C \displaystyle=\frac{-n_{y}/(ik_{0})}{n_{y}^{2}+n_{z}^{2}-\varepsilon_{11}(x)}\left(\frac{d}{dx}\varepsilon_{21}(x)+\frac{\varepsilon_{21}(x)\frac{d}{dx}\varepsilon_{11}(x)}{n_{y}^{2}+n_{z}^{2}-\varepsilon_{11}(x)}\right) \displaystyle-(n_{z}^{2}-\varepsilon_{11}(x))-\frac{\varepsilon_{21}(x)^{2}}{n_{y}^{2}+n_{z}^{2}-\varepsilon_{11}(x)} (4) All coefficients in ( 3 ) are symmetric in n_{y} , with the exception of C , which is symmetric in n_{y} only if the plasma is non-gyrotropic ( \varepsilon_{21}=0 ), or if there are no gradients in the x (radial) direction. Numerical experiments in figure 4 show that only the first option (no gyrotropy) completely restores the symmetry. In fact, the mere presence of a plasma-vacuum interface (the antenna is kept in vacuum in all calculations in this work, hence there is a discontinuous plasma-vacuum interface) suffices to break the vertical symmetry, even if the plasma is otherwise uniform, which can be seen from earlier studies of the scattering problem on this interface Tierens et al. 2021 . Any parasitically excited slow waves, which may play an important role in the sheath excitation physics Colas et al. 2025 ; Myra 2021 , also have this asymmetry. The electric field of the slow wave can be approximated using a potential \phi Bellan 2008 . The electric field is \bm{E}=-\nabla(\phi(x)\exp(i(n_{y}y+n_{z}z)\omega/c)) , and it obeys obeys \nabla\cdot(\varepsilon\bm{E})=0 : \displaystyle k_{0}^{2}\phi(x)\left(n_{y}^{2}\varepsilon_{11}(x)+n_{z}^{2}\varepsilon_{33}(x)\right) \displaystyle=\left(\frac{d}{dx}\varepsilon_{11}(x)\right)\left(\frac{d}{dx}\phi(x)\right) \displaystyle+\varepsilon_{11}(x)\frac{d^{2}}{dx^{2}}\phi(x) \displaystyle+ik_{0}n_{y}\phi(x)\frac{d}{dx}\varepsilon_{12}(x) (5) Once again, all coefficients are symmetric in n_{y} provided that either \varepsilon_{12}(x)=0 (no gyrotropy) or \frac{d}{dx}\varepsilon_{12}(x)=0 (no gradients). The fact that the absence of gyrotropy suffices to fully restore the symmetry can be understood by noting that, in the non-gyrotropic case, if \varepsilon has vertical symmetry \varepsilon(x,y)=\varepsilon(x,-y) , the wave equation ( 1 ) possesses the following symmetry: if \displaystyle\bm{E}=[E_{x}(x,y),E_{y}(x,y),E_{z}(x,y)]^{T}\exp(in_{z}z\omega/c) is a solution of ( 1 ), then so is its vertical mirror image \displaystyle\bm{E}=[E_{x}(x,-y),-E_{y}(x,-y),E_{z}(x,-y)]^{T}\exp(in_{z}z\omega/c) Figure 4: Eq. ( 3 ) suggests two ways of restoring the vertical symmetry absent in figure 3 . Left: by artificially removing the gyrotropy. Right: by removing the gradients (constant density n_{e}=10^{18} m -3 and magnetic field strength B=3.2 T, though the antenna is still in vacuum). Only removing the gyrotropy succeeds in completely restoring the symmetry. Returning to the horizontal symmetry, which is also absent in experiment (figure 1 ), we consider under what circumstances the wave equation ( 1 ) possesses a horizontal mirror symmetry. With the dielectric tensor ( 2 ) a horizontal mirror symmetry exists even in gyrotropic plasma, but this symmetry is broken by the introduction of a nonzero magnetic field tilt angle. Thus, in the experimentally relevant scenario where the plasma is both gyrotropic and the confining magnetic field is not purely toroidal, we should expect neither horizontal nor vertical symmetry. Hot spots like those in figure 3 can be reduced by controlling the phasing and power balance, as is often done to great effect in the toroidal direction Bobkov et al. 2019 . Here, we have to do it in the poloidal direction. This is not possible on current devices, though certain benefits of having this capability were pointed out as early as the 1980s Ram and Bers 1984 , and the future ITER ICRF antenna and the under construction WEST Travelling Wave Antenna will be able to do it. Figure 5 suggests that the peak RF potential can be reduced by \sim 20\% in this way. Figure 5: One way to compensate for the hot spots in figure 3 is to change the poloidal phasing and power balance. The operating regime on WEST is (\pi,0) (circle) on this plot: same amplitude but opposite phase on the top and bottom ports. The optimum (star) has slightly less power on the top port and a phase difference slightly below \pi . The RF potentials in this plot are normalized to the value at (\pi,0) . Another option is to modify the limiter shape. In light of the asymmetries inherent in the wave physics, we consider if an asymmetric limiter might outperform a symmetric one. We have parameterized the limiter shape in our model, so we can retract it further away from the plasma at the location of the hot spots. Figure 6 shows one possible modified limiter shape. Figure 7 shows the predicted sputtering for both proposed modifications (poloidal phasing, modified limiter shape). Optimal poloidal phasing gives a \sim 60% reduction in peak sputtering and a \sim 20% reduction in total sputtering w.r.t. the reference of figure 3 . The modified limiter shape gives a factor \sim 3 reduction in peak sputtering and a factor \sim 2 reduction in total sputtering. Both cases show a more evenly distributed erosion pattern, and hence better material survivability. 24 deg, 0.15m 13 deg, 1m Figure 6: The limiter shape in our model is parameterized using three circular arcs, allowing us to test a modified limiter shape where the top, where the hot spots occur, is further away from the plasma. Note that the antenna box remains entirely in the limiter shadow. The predicted effectiveness of this apparently minor limiter shape change can be understood in terms of two factors: first, the edge density gradient is very steep, so even a small displacement gives rise to a large difference in the incident ion flux. Second, we displace precisely the location of the hot spot, which accounted for most of the sputtering to begin with. This allows us to reduce sputtering while keeping the overall coupled power constant, which is difficult to achieve when one retracts the entire antenna instead (this is possible on WEST). It is now standard in ICRF operation to control the toroidal phasing and power balance, which minimizes the parallel electric field near the limiters and thereby the sheath rectification Bobkov et al. 2019 , and which can also be used to mitigate convective cell-induced asymmetries in the particle flux to the ICRF antenna Diab et al. 2025 . Our results strongly indicate that a substantial further decrease of the RF-specific impurity sputtering, beyond that already achieved by these measures, is achievable by poloidal modifications of the antenna, ranging from relatively simple poloidal phasing and power balance control, to redesign of the limiter shape. This reduction is possible even in scenarios where the plasma is otherwise well-aligned with the antenna. This also achieves a more even distribution of the heat loads on the antenna surfaces, improving material survivability. Figure 7: Predicted sputtering on (left) the standard limiter with the optimal phasing from figure 5 and (right) on the modified asymmetric limiter, with standard phasing. The total (integrated) sputtering yield is 4.2\cdot 10^{17} (left) resp. 2.8\cdot 10^{17} (right) atoms per second. This manuscript has been authored by UT-Battelle, LLC, under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy (DOE). This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research and Office of Fusion Energy Sciences, Scientific Discovery through Advanced Computing (SciDAC) program. *