The properties of the filamentary microdischarges found in dielectric barrier discharges depend on the manner of charging of the dielectric. The charging of the dielectric removes voltage from the gap thereby reducing E/N and producing a transition from an avalanching discharge to a recombination or attachment dominated discharge. In this article, we report on a computational investigation of these processes using a one-dimensional plasma chemistry model. We find that the expansion and ultimate stalling of the microdischarge is largely determined by charging of the dielectric at larger radii than the core of the microdischarge. The lowering of E/N in the core of the microdischarge in attaching gases can quickly consume electrons. This transition produces a discharge consisting of an expanding shell having a high electron density and an inner core dominated by negative ions. In extreme cases where the gas mixture contains thermal electron attaching gases, the core of the microdischarge is essentially a negative ion-positive ion plasma. Using square wave voltage pulses, the residual charge on the dielectric after the microdischarge, which contributes to the gap voltage on the next voltage pulse, is largely determined by the attachment rate in the core of the microdischarge. Rapid attachment reduces the plasma conductivity and leaves residual charge on the dielectric.

Dielectric barrier discharges (DBDs) are pulsed atmospheric pressure devices in which the plasma forms as an array of microdischarges having diameters expanding from 10 to 100’s μm and area densities of 10 to 100’s cm−2. The microdischarges are <10’s ns in duration and are terminated by charging of the dielectric barrier which removes voltage from the gap. If microdischarges are spaced sufficiently close together they may interact during their expansion. In this article, we discuss results from a two-dimensional plasma hydrodynamics model for microdischarge development in DBDs with the goal of investigating the interaction between closely spaced microdischarges. We find that the efficiency of ionization is only moderately affected by microdischarges which expand into physical contact. The residual charge left on the dielectric following a current pulse can, however, significantly impact the spatial extent of the subsequent microdischarges. During expansion the underlying dielectric charges to progressively larger radii as the microdischarge expands. This leads to voltage collapse in the center of the microdischarge prior to the outer radius. In attaching gas mixtures larger rates of attachment relative to ionization at the lower values of the electric field/number density produce cores which are highly electronegative, surrounded by shells of higher electron density.

Perfluorinated compounds ͑PFCs͒, gases which have large global warming potentials, are widely used in plasma processing for etching and chamber cleaning. Due to underutilization of the feedstock gases or by-product generation, the effluents from plasma tools using these gases typically have large mole fractions of PFCs. The use of plasma burn-boxes located downstream of the plasma chamber has been proposed as a method for abating PFC emissions with the goals of reducing the cost of PFC abatement and avoiding the NO x formation usually found with thermal treatment methods. Results from the two-dimensional Hybrid Plasma Equipment Model have been used to investigate the scaling of plasma abatement of PFCs using plasma burn-boxes. An inductively coupled plasma ͑ICP͒ etching chamber is modeled to determine the utilization of the feedstock gases and the generation of by-products. The effluent from the etching chamber is then passed through a plasma burn-box excited by a second ICP source. O 2 , H 2 , and H 2 O are examined as additive gases in the burn-box. We find that C 2 F 6 ͑or CF 4 ͒ consumption in the etching reactor increases with increasing ICP power deposition at constant C 2 F 6 ͑or CF 4 ͒ mole fraction, and decreasing C 2 F 6 ͑or CF 4 ͒ mole fraction or total gas flow rate at constant power. The efficiency of removal of C 2 F 6 ͑eV/molecule͒, however, is strongly dependent only on the C 2 F 6 mole fraction and total gas flow rate. All PFCs in the effluent can generally be abated in the burn-box at high power deposition with a sufficiently large flow of additive gases. In general CF 4 generation occurs during abatement of C 2 F 6 using O 2 as an additive. CF 4 is not, however, substantially produced when using H 2 or H 2 O as additives. The efficiency of PFC abatement decreases with increasing power and decreasing additive mole fraction.