(Sub)nanosecond transient plasma for atmospheric plasma processing experiments: application to ozone generation and NO removal

In this paper we use a (sub)nanosecond high-voltage pulse source (2–9 ns pulses with 0.4 ns rise time) to generate streamer plasma in a wire-cylinder reactor and apply it to two atmospheric plasma processing applications: ozone generation and NO removal. We will investigate what pulse parameters result in the highest plasma processing yields.The results show that for ozone generation, secondary-streamer effects appear to have a slight influence on the ozone yield: if the pulse duration increases and/or the voltage increases in such a way that streamers can start to cross the gap in the reactor, the ozone yields decrease. Furthermore, for NO removal, we see a similar effect of pulse duration and applied voltage as for the ozone generation, but the effect of the pulse duration is slightly different: long pulses result in the highest NO-removal yield. However, the NO-removal process is fundamentally different: besides removing NO, the plasma also produces NO and this production is more pronounced in the primary-streamer phase, which is why the pulse polarity has almost no influence on the NO-removal yield (only on the by-product formation). Moreover, the rise time of the pulses has a much more significant effect on ozone generation and NO removal than the pulse duration: a long rise time results in a lower enhanced electric field at the streamer heads, which consequently reduces the production of radicals required for ozone generation and NO removal, and decreases the streamer volume. Consequently, the resulting ozone yields and NO-removal yields are lower. Finally, the main conclusion is that the plasma generated with our nanosecond pulses is very efficient for ozone generation and NO removal, achieving yields as high as 175 g centerdot ${\rm kWh}^{-1}$ for ozone generation and 2.5 mol centerdot ${\rm kWh}^{-1}$ (or 14.9 eV per NO molecule) for NO removal.