Sputtering in a glow discharge ion source-pressure dependence: theory and experiment

Abstract

A simplified theoretical expression has been developed for a glow discharge to show how the average cathode erosion rate (expressed as the number of atoms per ion of the total bombarding flux) varies with primary sputter yield, pressure, 'diffusion length' and sputtered atom 'stopping' cross section. An inverse pressure dependence is predicted which correlates well with experiment in the <1 to 3.5 Torr range. Observed sputtering efficiency variations with the discharge gas show that Xe<Kr<Ar, which is shown to be in good agreement with theoretical differences in the low-energy primary sputter yield expected on the basis of the mass of the bombarding particle flux. There is a change at higher pressures because the rate levels off and the rates for the same cathode in different gases, including N2 and He, tend to converge. It is suggested that this could be due to a change in the mechanism to self-sputtering. Under constant conditions, the erosion rates of different cathode materials still correlate quite well with the differences in their primary sputter yields.