Dear Colleagues!
I have a high-voltage converter at home, consisting of a horizontal "flyback" transformer from an old tube TV, field-effect power transistors and a square-wave generator with adjustable frequency. Depending on the frequency setting and power supply, this device is capable of producing high frequency voltages (2-100 kHz) of the order ~5-25 kV [determined by the breakdown distance in air between needle electrodes based on the assumption of 1 kV/mm], sufficient to observe corona discharge and interesting arcs, ozone production. (see photos)
When an arc from this converter is passed through baking soda (NaHCO3) powder, the powder melts and microscopic balls or grains form. Some of them fluoresce brightly in ultraviolet rays (365 nm) with a yellow color, suspiciously reminiscent of the color of a monochromatic low-pressure sodium lamp. (Unfortunately, at the moment I do not have a spectroscope and estimate it by naked eye).
(See photo with yellow glow, under microscope x40).
Of course, the same thing happens with soda ash Na2CO3. But also, with sodium sulfate Na2SO4 too.
K2SO4 give brilliant dark-red fluorescence color, or red-violet (suspiciously reminiscent of the color that a gas burner flame is painted with potassium salts). See photo with red glow)))
NaCl, unfortunately, not give any fluorescent effects after arc treatment(((
My hypothesis is that in a high-voltage arc, despite the bipolar current, electrolysis of molten soda or sulfates occurs to some extent. In this case, metallic sodium or potassium is built into the crystal lattice of the salt with the formation of F-centers (https://en.wikipedia.org/wiki/F-center). Their concentration is insufficient for the appearance of color in visible light, but sufficient for the appearance of fluorescence at a characteristic element wavelengths in the UV. (Although with potassium sulfate the yellow color after arc treatment is very noticeable. See photo with yellow grain).
When treated with water, the yellowing of the potassium sulfate microscopic grains and the fluorescent properties of all mentioned compounds immediately disappear, as should be the case with F centers. pH of a solution of K2SO4/Na2SO4 which previously been treated with an arc - alkaline instead of neutral on untreated samples.
In your opinion, what yet can be the reason for these phenomena, the appearance of fluorescence in the UV visually extremely reminiscent of the flame coloring by element's compounds or color of glow discharge at low pressure in metal vapor?