Open Access

Raphael Rataj, Matthias Werneburg, Harald Below, Juergen F Kolb

• AbstractPulsed streamer discharges submerged in water have demonstrated potential in a number of applications. Especially the generation of discharges by short high-voltage pulses in the nanosecond range has been found to offer advantages with respect to efficacies and efficiencies. The exploited plasma chemistry generally relies on the initial production of short-lived species, e.g. hydroxyl radicals. Since the diagnostic of these transient species is not readily possible, a quantification of hydrogen peroxide provides an adequate assessment of underlying reactions. These conceivably depend on the characteristics of the high-voltage pulses, such as pulse duration, pulse amplitude, as well as pulse steepness.A novel electrochemical flow-injection system was used to relate these parameters to hydrogen peroxide concentrations. Accordingly, the accumulated hydrogen peroxide production for streamer discharges ignited in deionized water was investigated for pulse durations of 100 ns and 300 ns, pulse amplitudes between 54 kV and 64 kV, and pulse rise times from 16 ns to 31 ns. An independent control of the individual pulse parameters was enabled by providing the high-voltage pulses with a Blumlein line. Applied voltage, discharge current, optical light emission and time-integrated images were recorded for each individual discharge to determine dissipated energy, inception statistic, discharge expansion and the lifetime of a discharge.Pulse steepness did not affect the hydrogen peroxide production rate, but an increase in amplitude of 10 kV for 100 ns pulses nearly doubled the rate to (0.19 ± 0.01) mol l−1 s−1, which was overall the highest determined rate. The energy efficiency did not change with pulse amplitude, but was sensitive to pulse duration. Notably, production rate and efficiency doubled when the pulse duration decreased from 300 ns to 100 ns, resulting in the best peroxide production efficiency of (9.2 ± 0.9) g kWh−1. The detailed analysis revealed that the hydrogen peroxide production rate could be described by the energy dissipation in a representative single streamer. The production efficiency was affected by the corresponding discharge volume, which was comprised by the collective volume of all filaments. Hence, dissipating more energy in a filament resulted in an increased production rate, while increasing the relative volume of the discharge compared to its propagation time increased the energy efficiency.