Mg. Rusbridge et al., Observations of the interaction of a plasma stream with neutral gas: evidence of plasma loss through molecular-activated recombination, PLASMA PHYS, 42(5), 2000, pp. 579-602
We describe an experiment, the UMIST Linear System (ULS), in which a hydrog
en plasma stream, guided by a longitudinal magnetic field, is injected thro
ugh a diaphragm containing an orifice into a separately-pumped target chamb
er in which the neutral hydrogen pressure can be raised to a maximum of 8 m
Torr. The stream is about 6 mm in diameter, has an electron temperature of
up to 15 eV and an ion flux of 3 x 10(18) s(-1); it is supersonic with Mach
number up to M approximate to 3, We have studied both the passage of the s
tream through the orifice and the interaction of the supersonic plasma with
neutral hydrogen in the target chamber. We find that transmission is incom
plete even when die orifice diameter is five times that of the plasma; we a
ttribute this to the presence of ion trajectories which extend well outside
the visible plasma and are intercepted by the diaphragm. In the target cha
mber, the stream does not broaden, but the ion flux decreases approximately
exponentially with distance, with a scale length of the order of the mean
free path for momentum transfer in ion-neutral collisions. and much less th
an that expected for other processes, such as charge exchange or electron-i
on recombination. Elastic collisions alone cannot decrease the flux, but wo
uld lead to a large accumulation of slow ions in thermal equilibrium with t
he neutral gas, which must be limited by some other loss process: collision
al diffusion and electron-ion recombination are too slow, leading to a dens
ity approaching 10(20) m(-3). The observed density is of the order of 10(18
) m(-3), requiring a process with a rate of 10-100 times faster. Calculated
rates for molecular-activated recombination (MAR) of the slow ions are of
this order, and the predicted density agrees with our observations to order
of magnitude.