Paper PS2-WeM9
Development and Characterization of a Radical Beam Source Based on Surface Waves for Plasma-Surface Reaction Studies

Wednesday, October 22, 2008, 10:40 am, Room 306

Previously, we studied recombination of Cl and O on plasma-conditioned anodized aluminum and stainless steel surfaces. Cl and O atoms formed in chlorine or oxygen plasmas impinged on a cylindrical substrate that was rapidly rotated such that points on the surface were exposed to the plasma and then to a differentially-pumped analysis chamber equipped with either an Auger electron spectrometer or a mass spectrometer. Langmuir Hinshelwood (LH) recombination was observed by monitoring desorption of Cl₂ and O₂ with the mass spectrometer or through a pressure rise. In these previous experiments, however, Eley Rideal (ER) recombination (if it occurs) could not be detected because it would take place instantaneously in the presence of atom flux, and hence would cease as soon as the sample left the plasma. To observe the ER component, as well as to isolate LH recombination in plasmas with multiple radical species (i.e. most plasmas), a separate radical beam source is needed in combination with the plasma and spinning substrate. With this in mind, we investigated a surface-wave chlorine plasma operating at 2.45 GHz and sustained in a 8 mm O.D. quartz tube using a gap-type surfatron wave launcher. With added traces of rare gases, optical emission spectroscopy was used to measure Cl and Cl₂ densities and the electron temperature, T_e, at 50 mTorr as a function of distance from the wave launcher. The Cl(792.4 nm)-to-Xe(828 nm) emission intensity ratio, reflecting the Cl number density, decreased with distance from the launcher, while the Cl₂ (306 nm)-to-Xe emission ratio that is proportional to Cl₂ number density, peaked near the launcher. The Cl₂ percent dissociation obtained from the calibrated Cl₂ -to-Xe emission ratio was very high (97 %) near the launcher, and remained high (89 %) until the end of the plasma column (about 12 cm from the launcher for an absorbed power of 90 W). By selecting Ne, Ar, Kr, and Xe lines excited from the ground state which are characteristic of the high energy portion of the electron energy distribution function (particularly Ne), we found that T_e increased from 5 to 10 eV as the observation point was moved away from the launcher. On the other hand, a nearly constant value of T_e = 3.1 ± 0.6 eV was obtained using Ar, Kr and Xe lines excited to a significant extent through impact with lower energy electrons. Mechanisms for such high energy tails will be discussed.