Energetic wave-trapped electrons of ϳ20 eV, moving at the 13.56 MHz helicon wave phase velocity, are directly measured with a 3 ns response-time retarding-potential analyzer. The rf axial wavelength is measured with B-dot probes scanned axially. Changes in the axial magnetic field or the rf power cause order of magnitude changes in the energetic electron current. The current correlates with either a transition from partially propagating to nearly pure standing waves or quantized jumps in the number of half wavelengths between the two azimuthal antenna straps. [S0031-9007(97)03548-5] PACS numbers: 52.50. Dg, 52.40.Db, 52.70.Ds, 52.70.Nc Pulses of argon line radiation have been reported [1] at the heating frequency in a helicon plasma source: the pulses originated in the near field of the antenna and propagated at the helicon wave phase velocity, consistent with trapping and accelerating pulses of electrons to 30 eV in the radio-frequency (rf) fields under the antenna. We report on strong variations in the current of energetic, wavetrapped electron pulses with changes in the magnitude of the axial magnetic field or rf power. The electrons are directly measured with a retarding-potential analyzer, having sufficient time response to resolve 15 ± of phase angle of the 13.56 MHz rf wave. We have previously shown that the energetic electrons are wave-trapped within a 75 ± FWHM phase angle each rf cycle [2]. This supports the picture that electrons "surf" on the helicon wave [1]; although, for our parameters, a given electron is expected to remain trapped for only a mean-free path of ϳ4 cm (limited by electron-neutral collisions) which is less than the helicon wavelength of ϳ20 cm. Orbit code calculations are used to assess this effect.While Langmuir probes are a more common method of measuring the electron energy distribution function, the proper technique in rf fields continues to be controversial [3][4][5]. Retarding-potential gridded-energy analyzers (GEA) avoid these uncertainties; however, only electrons with axial energy exceeding the sheath potential drop can be detected. Sheath potential oscillations at the heating frequency are small. The primary diagnostic for this paper is a 1.9 cm diameter GEA [2,6] with grid voltages swept to measure electron and ion distribution functions. The distance between the last grounded entrance grid and the collector is 0.20 cm, providing 1 ns time response for 20 eV electrons, limited to 3 ns (corresponding to 15 ± of phase) by the 150 MHz analog bandwidth recorder. For 5 eV argon ions, the response time of 400 ns averages the ion distribution over 6 wave periods, so plasma potential measurements are not phase resolved. The GEA and other probes are downstream of the antenna region where hot electrons are produced, so should not perturb these data.Helicon sources are attractive for plasma processing and a number of other applications because they provide high plasma density in low magnetic fields, with a high ion-ization efficiency, and a low, ϳ3 eV, electron temperature [7-9]....

Gas desorption and electron emission coefficients were measured for 1 MeV potassium ions incident on stainless steel at grazing angles (between 80 and 88 from normal incidence) using a new gaselectron source diagnostic (GESD). Issues addressed in design and commissioning of the GESD include effects from backscattering of ions at the surface, space-charge limited emission current, and reproducibility of desorption measurements. We find that electron emission coefficients e scale as 1= cos up to angles of 86 , where e 90. Nearer grazing incidence, e is reduced below the 1= cos scaling by nuclear scattering of ions through large angles, reaching e 135 at 88. Electrons were emitted with a measured temperature of 30 eV. Gas desorption coefficients 0 were much larger, of order 0 10 4. They also varied with angle, but much more slowly than 1= cos. From this we conclude that the desorption was not entirely from adsorbed layers of gas on the surface. Two mitigation techniques were investigated: rough surfaces reduced electron emission by a factor of 10 and gas desorption by a factor of 2; a mild bake to 220 had no effect on electron emission, but decreased gas desorption by 15% near grazing incidence. We propose that gas desorption is due to electronic sputtering.

Making axisymmetric mirrors magnetohydrodynamically (MHD) stable opens up exciting opportunities for using mirror devices as neutron sources, fusion-fission hybrids, and pure-fusion reactors. This is also of interest from a general physics standpoint (as it seemingly contradicts well-established criteria of curvature-driven instabilities). The axial symmetry allows for much simpler and more reliable designs of mirror-based fusion facilities than the well-known quadrupole mirror configurations. In this tutorial, after a summary of classical results, several techniques for achieving MHD stabilization of the axisymmetric mirrors are considered, in particular: (1) employing the favorable field-line curvature in the end tanks; (2) using the line-tying effect; (3) controlling the radial potential distribution; (4) imposing a divertor configuration on the solenoidal magnetic field; and (5) affecting the plasma dynamics by the ponderomotive force. Some illuminative theoretical approaches for understanding axisymmetric mirror stability are described. The applicability of the various stabilization techniques to axisymmetric mirrors as neutron sources, hybrids, and pure-fusion reactors are discussed; and the constraints on the plasma parameters are formulated.

We report results of plasma confinement experiments with an auxiliary warm-plasma component flowing along magnetic field lines to suppress ion-cyclotron instabilities. The reduced plasma losses, with the lower fluctuation amplitude, permits neutral-beam buildup of a 13-keV deuterium plasma to densities as high as 4 x 10 13 cm" 3 corresponding to peak beta values of 0.4. Variation of the beam energy demonstrates that longer confinement times are achieved at higher ion energies.

A tandem-mirror plasma has been sustained and heated by rf alone without the need for neutral beams. End plug density and energy are maintained by ion cyclotron-resonance heating which traps and heats a fraction of the central-cell loss stream. The central-cell plasma is maintained by gas fueling and rf heating. Magnetohydrodynamic stability limits the ratio of the central-cell to plug plasma pressure, and the central-cell electron temperature must be kept high enough for ionization. A quasi steady state is achieved that lasts much longer than the decay times of the plugs and central cell.PACS numbers: 52.55. Mg, 52.50.Gj, 52.55.Ke The tandem-mirror approach to the development of a fusion reactor promises to have many advantages over simple mirrors as well as other magnetic confinement devices. 1 In a tandem mirror, a central-cell plasma in a solenoid is bounded by "plug" plasmas in mirror cells. The TMX experiment 2 demonstrated that electrostatic confinement of central-cell ions by the plug plasmas significantly reduced end losses. Calculations indicate that the overall Q (power gain) of a tandem-mirror fusion reactor can be quite high (~5-10) if the volume of the central cell greatly exceeds that of the plugs. 3 In previous experiments, plug plasmas were fueled and heated by energetic neutral beams. However, technological constraints on neutral-beam heating when scaled to a reactor could be eased or eliminated by supplementing or replacing neutral-beam power with rf power. We show in the Phaedrus experiment that a tandem-mirror plasma can be sustained by rf alone, without the application of neutral beams.The operation of a tandem mirror can be thought of as taking place in two stages, a transient stage followed by a quasi-state stage. In the transient stage, plasma is injected into the machine by stream guns. In fact, earlier experiments with Phaedrus 4 and Gamma 6 5 were restricted to the transient stage. Such plasmas side-stepped many of the issues of tandem-mirror physics because plasma characteristics were dominated by the external plasma between the plugs and the stream guns. The high-density external plasma line-tied the plasma to the stream guns, significantly reducing both magnetohydrodynamic (MHD) instabilities and microinstabilities and fixing the electron temperature. When the stream guns were turned off in Phaedrus, the plasma decayed away in approximately 150 jus in the plugs and 400 JUS in the central cell. Quasi-steady-state operation after stream guns are turned off requires sources of particles and energy for both the plug and central-cell plasmas and also adds constraints that must be satisfied in order to maintain microstability and MHD stability. In TMX, fueling and heating were provided by neutral beams in the plugs and gas puffing in the central cell. In the Phaedrus experiment, fueling is provided only by gas puffing in the central cell while rf provides heating. In both experiments, stability was provided by high plug energy density and the presence of the central-cell loss flow.The ...