AVS 61st International Symposium & Exhibition | |
Plasma Science and Technology | Friday Sessions |
Session PS2-FrM |
Session: | Plasma Surface Interactions II |
Presenter: | Floran Peeters, Eindhoven University of Technology, Netherlands |
Authors: | F.J.J. Peeters, Eindhoven University of Technology, Netherlands R.F. Rumphorst, Eindhoven University of Technology, Netherlands M.C.M. van de Sanden, FOM institute DIFFER, Netherlands |
Correspondent: | Click to Email |
Dielectric Barrier Discharges (DBD’s) are used on a large industrial scale and have been studied for more than a century, with increasing interest in recent years in the areas of materials processing, plasma medicine and solar fuels. DBD’s in filamentary mode consist of many small, transient microdischarges with diameters of ~ 0.1 mm and durations on the order of several 10’s of nanoseconds, distributed over the dielectric surface. We study the collective behaviour of many such filaments in air by using a fast analog circuit capable of measuring the conductively transferred charge per filament. By using a miniature planar DBD with a 7 mm2 electrode area, we determine charge/filament distributions without significant overlap between filaments in time, even at high filament number densities of up to 200 filaments/cm2/period. Contrary to previous work, we find that the charge/filament distributions are log-normal in nature. Furthermore, the distributions are independent of filament number density for a given DBD geometry. With conventional charge-voltage (Q-V) measurements, Lissajous figures are obtained for the miniature DBD, where the slope during a discharge period has a clear ‘staircase’ shape. Analysis of these Lissajous figures, for 8 different DBD geometries, reveals that filaments do not occur randomly within a discharge period, as is often assumed, but affect each other’s moment of ignition. Using both measurement techniques, we infer that multi-filament discharge dynamics are regulated by residual conductivity of the gas near the surface of the dielectric, resulting in step-wise ignition of filaments as a function of applied voltage. The log-normal charge/filament distributions, on the other hand, develop from the locally trapped charges on the dielectric. We suggest that both mobile charges from residual conductivity and immobile trapped charges need to be considered in models of DBD’s, especially when converting data from the abundant single-filament models in DBD literature to real devices.