P. Berlemont et al., NEGATIVE-ION VOLUME PRODUCTION-MODEL - STATE OF THE EXPERIMENTAL CONFIRMATION, Review of scientific instruments, 64(10), 1993, pp. 2721-2728
We report the results of simulating two multicusp volume negative ion
sources in which plasma parameters, hydrogen vibrational spectra, and
in one case the negative ion density, have been measured. We find gene
rally good quantitative and qualitative agreement between experiment a
nd theory for the basic discharge parameters. Good agreement is also f
ound for atomic density, although the wall recombination coefficient g
amma(H) is 0.05 in one source and 0.5 in the other. The calculated den
sities of vibrationally excited states with upsilon'' greater-than-or-
equal-to 5 are higher than the measured densities, but the calculated
negative ion densities are lower than available measurements. Possible
causes for these discrepancies are suggested. Fundamental modificatio
ns to the model may be necessary, and we have discussed two possibilit
ies: a suprathermal rotational spectrum, and a spatial gradient in vib
rational excitation (due to elastic collisions which can localize the
effect of vibrational wall cooling near the wall).