Negative ion density fronts

Abstract

Negative ions tend to stratify in electronegative plasmas with hot electrons (electron temperature T e much larger than ion temperature T i , T e >>T i ). The boundary separating a plasma containing negative ions, and a plasma, without negative ions, is usually thin, so that the negative ion density falls rapidly to zero -forming a negative ion density front. We review theoretical, experimental and numerical results giving the spatio-temporal evolution of negative ion density fronts during plasma ignition, the steady state, and extinction (afterglow). During plasma ignition, negative ion fronts are the result of the break of smooth plasma density profiles during nonlinear convection. In a steadystate plasma, the fronts are boundary layers with steepening of ion density profiles due to nonlinear convection also. But during plasma extinction, the ion fronts are of a completely different nature.Negative ions diffuse freely in the plasma core (no convection), whereas the negative ion front propagates towards the chamber walls with a nearly constant velocity. The concept of fronts turns out to be very effective in analysis of plasma density profile evolution in strongly non-isothermal plasmas.

Periodic formation and propagation of double layers in the expanding chamber of an inductive discharge operating in Ar∕SF6 mixtures

“…It is well established that double layers, or negative-ion fronts, can form in low-pressure electronegative discharges. [23][24][25] A necessary condition for double-layer formation is that the ion sound speed exceeds the local ion acoustic speed. In electronegative plasmas, the ion acoustic speed is significantly lower than that in electropositive plasmas, 26 which may explain why the DL forms above a minimum SF 6 fraction.…”

Radio-frequency discharges in oxygen: I. Particle-based modelling

“…The meta-stable O 2 (a 1 ∆ g ) requires special attention because it also appears as an educt in the entrance channel for associative detachment (17). Its concentration is therefore important, and we should actually build-up the O 2 (a 1 ∆ g ) distribution function, that is, we should also simulate O 2 (a 1 ∆ g ) particles.…”

“…This is not surprising because the occurrence of negative ions leads to abrupt changes in the ion density (density fronts [17]) which in most cases force the discharge to stratify into a central quasi-neutral ion-ion plasma and a peripheral electro-positive electronion plasma [10,11,12,13,14,18,19]. The transition between the two is rather subtle.…”

“…It takes place through direct detachment (16), resulting in an atomic oxygen, an oxygen molecule and an electron, and through associative detachment (17), which leads to an O 3 molecule and an electron. The latter is rather surprising because there is no evidence for it in beam experiments (where only direct detachment is observed [58,59]).…”

“…Binary collisions with one charged particle in the exit channel, such as dissociative attachment (15) and associative detachment (17), are treated within the modified binary collision model described in the previous subsection. Impact ionization (18) and impact detachment (19) are modelled as follows: First, an inelastic binary collision is performed, in which the parent electron looses the ionization (detachment) energy.…”