Thermal population ion cyclotron emission in Wendelstein 7-X: Electrostatic edge eigenmodes in the absence of fast ions
D. Moseev, R. Ochoukov, M. Dreval, I. Kuzmych, P. Aleynikov, K. Aleynikova, S. Bozhenkov, R. O. Dendy, D. Hartmann, J.-P. Kallmeyer, Y. O. Kazakov, L. Krier, H. P. Laqua, S. Marsen, K. G. McClements, J. Ongena, J. W. Oosterbeek, S. Ponomarenko, B. Reman, J. Riemann, M. Salewski, O. Samant, D. Straus, B. S. Schmidt, T. Schröder, B. Schweer, T. Stange, R. C. Wolf, J. Zimmermann,We report ion cyclotron emission (ICE) in Wendelstein 7-X during discharges heated purely by electron cyclotron resonance heating that lack any fast-ion population, a phenomenon we term thermal population ICE (TPICE). TPICE is detected on the unshielded ICRF antenna strap but not on B-dot probes, indicating a predominantly electrostatic nature. A central finding is that ICE is not exclusively a fast-ion phenomenon: the thermal plasma can excite the same edge eigenmodes that fast ions drive. The evidence strongly suggests that the electrostatic component of fast-ion-induced ICE (FI ICE) and TPICE originate from the same instability, supported by the nearly identical spectra of TPICE and electrostatic FI ICE at spatially separated detectors; matching FI ICE peaks from neutral beam injected sources that are injected in opposite directions relative to the magnetic field; and a discharge where a nearly pure helium plasma produces neither TPICE nor helium FI ICE, yet yields clean hydrogen FI ICE upon a hydrogen NBI blip. Kolesnichenko's model, the only theory developed explicitly for TPICE, predicts an electrostatic instability with a strictly positive frequency detuning of 2–5 MHz above the cyclotron frequency, and an electromagnetic instability with a very small detuning of either sign. Only an electrostatic signal is observed, at a frequency closer to the electromagnetic prediction. This disagreement is attributed to the instability likely originating outside the last closed flux surface, where profile data are unavailable. The model is inapplicable to helium discharges, where the signal peaks at 41 MHz rather than at any predicted cyclotron harmonic.