Search

We have measured the E.coli bactericidal efficiency of DC and AC point-to-plane coronas in air and in argon in the corona modes positive glow, positive streamer, negative Trichel pulse and negative glow. Negative Trichel pulse and positive streamer coronas are the most efficient for production of bactericidal agents in air. Less than 5 minute exposure to the products from 30 μA of these coronas will sterilize the E.coli covered agar surfaces in a 3l volume to a survival ratio of 10−5. Positive glow coronas are two orders of magnitude less efficient. 50 Hz AC coronas seem to mainly work during the negative half cycles. The bactericidal agent(s) produced by the coronas have not all been identified, but ozone O3 with concentration 35–40 ppm is probably the most important one. We show that charged particles and photons can play only a minor role, so the agent(s) must be neutral particles. The main part of them must have a lifetime in the vessel of more than 3 minutes. The much smaller, and volume/distance dependent, efficiency of coronas in humid argon indicates that metastable neutrals or radicals from H2O can play only a minor role. A preliminary test showed negligible effects of the corona treatment on the spores of Bacillus subtilis.

The electronic system of this coincidence counter consists of two 13-stage Venetian blind photomultipliers coupled to a simple DeBenedetti-Richings coincidence circuit, followed by a medium-gain amplifier, pulse selector and scaler. The main feature of the system is its simplicity, chiefly realized by utilizing the quasi-linear transfer properties of the take-the-lesser type coincidence circuit. “Balance-point operation” permits the use of multiplier high voltage supplies and pulse amplifiers not possessing the high stability usually required for scintillation counting. The counter has been operated at 26°C with a tritium counting efficiency of 16% at a background of 100 c/min. The minimum detectable tritium concentration in water is 1.7 × 10−8 curie/liter. However, the time needed for the counter to “quiet down” after a change of sample is prohibitively long, 2–3 days. Of the background, nearly 50% seems to be caused by light interaction between the photomultipliers. Some suggestions for future improvements are given.