AVS 55th International Symposium & Exhibition | |
Plasma Science and Technology | Thursday Sessions |
Session PS1-ThA |
Session: | Plasma Diagnostics, Sensors, and Control II |
Presenter: | A. Ranjan, University of Houston |
Authors: | A. Ranjan, University of Houston V.M. Donnelly, University of Houston D.J. Economou, University of Houston |
Correspondent: | Click to Email |
A new diagnostic was developed to measure the energy distribution of a fast (10s to100s eV) neutral beam. Fast neutrals were allowed to collide with slow (thermal) neutrals in a chamber of controlled background pressure (e.g., 10-4 Torr). A fraction of the fast neutrals was ionized as a result of the atom-atom collisions. The ionized species current was measured as a function of energy with a gridded energy analyzer and off-axis channel electron multiplier, housed in a differentially pumped chamber. The energy distribution of the fast neutral beam was determined from the known cross section of the atom-atom ionization collision as a function of energy. The method was applied to measure the energy distribution of a fast neutral beam formed by surface neutralization of ions, extracted through a grid with high aspect ratio holes (neutralization grid). A pulsed-plasma technique was implemented to achieve an ion beam with a tight energy spread. Ion energy was controlled by a DC bias, applied on an electrode in contact with the plasma, during part of the afterglow period. The electron temperature decays rapidly in the afterglow, which yields a nearly uniform space potential, resulting in an ion beam with tight energy spread. The peak of the NED was ~7% lower than that of the parent ion energy distribution (IED), compared to a ~3% expected energy loss, based on specular reflection. The neutral energy distribution (NED) had a larger energy spread as compared to the parent IED. For example, the FWHM of a NED and the corresponding parent IED were 32 eV and 10 eV, respectively. To study the effect of surface roughness of the neutralization grid, results for a metal grid with a "rough" surface (roughness ~ 10s of nm) will be compared with those of a "smooth" (0.15 nm RMS roughness) silicon grid.