Localized current drive using electron cyclotron resonance heating for NTM stabilization and profile control
Electron cyclotron current drive deposits directed momentum in plasma electrons by launching a millimeter-wave beam at or near the electron cyclotron frequency ωce = eB/me. When the beam is aimed with a toroidal steering angle, it preferentially heats electrons traveling in one toroidal direction through the Fisch–Boozer mechanism: electrons gaining perpendicular energy via cyclotron absorption experience reduced collisionality, producing a net current in the direction of the less-collisional population.1
The defining advantage of ECCD is its extraordinary spatial precision. Electron cyclotron resonance occurs where the local magnetic field satisfies ω = nωce (typically n = 1 or 2), and because the magnetic field varies monotonically across the tokamak cross-section, the deposition layer is a narrow radial shell. A steerable launcher—either a movable mirror or a phased-array antenna—can place the driven current at any desired radial location with a deposition width as small as a few centimeters.2
The normalized current-drive efficiency of ECCD is modest—typically γ ≈ 0.02–0.05 × 1020 A m−2 W−1—roughly a factor of five below LHCD. This lower bulk efficiency is acceptable because ECCD’s role is qualitative rather than quantitative: precise profile tailoring and MHD stabilization, not bulk current sustainment. The current-drive efficiency improves with electron temperature, favoring reactor-grade plasmas.3
High-power gyrotrons generate the millimeter-wave beams, with modern units delivering 1–2 MW continuous wave at frequencies of 110–170 GHz. ITER will deploy an ECCD system totaling 20 MW from 24 gyrotrons at 170 GHz for NTM control and plasma start-up assist. Transmission is via evacuated corrugated waveguides with losses below 10% over distances of 50–100 m.4
Because the beam propagates freely through vacuum and low-density edge plasma without interaction, ECCD is immune to the density-limit accessibility problems that constrain LHCD, making the two methods highly complementary in reactor current-drive portfolios.