I have been keeping these photo for some time now waiting for the completion of discharge spectra thermal analysis. Since I have been taking forever to do so thanks to the complexity of molecular spectrum; these two photos has been a constant annoyance on my desktop - a reminder of an incomplete work. I have finally decided to release it with some brief information.
Figure 1: Homemade halogen lightbulb transformer driven flyback DC high-voltage source energizing a multimeter probe. UV emission from corona discharge excites blue fluorescence on paper.
The photo above shows a metallic multimeter probe I've energized with positively rectified high-voltage from a vintage flyback transformer driven by a 12 V halogen lamp high-frequency digital "transformer". Under such high electrical tension, the vertex of the sharp tip have the highest charge density with electric field strength significantly stronger than the dielectric breakdown value of air ~3 million V/m [1, 2]. This causes the tip to glow purple from corona discharge followed by simultaneous release of ozone gas (distinctive musty smell) from the breakdown of oxygen molecules.
Right below the metal tip is a piece of paper acting as electric ground, attracting the positive ions spraying out from the metal tip subsequently causing the paper to glow blue from fluorescence (paper fluoresce because "optical brighteners" has been added). This whole set up was taken with an exposure time of 0.5 second to reveal the corona discharge. The multimeter probe was illuminated briefly by a white L.E.D to cast a shadow, indicating the tip is at a distance away from the paper.
Figure 2: Vibration energy level of nitrogen molecule . Major emission lines indicated by red arrow.
Unfortunately before I am able to measure the output voltage accurately, my last attempt to generate high voltage with similar set-up has resulted permanent damage in the secondary winding of the flyback transformer. My guess is that it has an open-circuit potential around + 20 kV estimated simply based on sharp tipped spark gap distance of 25 mm. The damage was most likely contributed by high current draw (far exceeding 1 mA) without a current-limiting resistor in series with the load (I hooked it up directly to a glow discharge tube) causing a meltdown somewhere within its fine wire coils.
The lesson: use a ballast resistor for any sort of load.
Figure 3: Emission spectrum of corona discharge. More than 90% light emitted are UVA radiation.
Figure above is the actual dark compensated, 3-second exposure spectrum of the corona discharge taken with OceanOptics USB4000 spectrometer. Even at 3 second exposure (instrumental limit) the data still appears noisy. It was definitely weak light to be picked up by the human eye but the spectrum certainly features at least four peaks within 300 to 400 nm falling under A-subtype UV radiation (keep in mind that ultraviolet light is invisible).
The four relatively bright emission have wavelengths slightly shorter than peak emission from violet L.E.D but they are less harmful to the eyes compared to emissions from germicidal lamp. All four spectral lines correspond to the second positive system N2 C → B (vibration bands) of atmospheric nitrogen molecules. More specifically for 337 nm line (important because it will be featured in TEA superluminal sources), the vibration level responsible for the transition is N2 (C3Π+u, v'=0) → (B3Π+g, v"=0) of the second positive system .
Figure 4a (top): Two glass vials on paper holding water suspension of highlighter pigments, left and right are blue and yellow respectively. Figure 4b (bottom): Long exposure photography shows yellow highlighter dye fluoresce while blue dye remains dark. Notice the blue fluorescence of paper under UVA irradiation.
Now, without anymore further discussions available, I bring up an experiment from the past dealing with fluorescence of highlighter pigments. Earlier, I have tested the dyes and found out the blue highlighter dye does not fluoresce under violet L.E.D. I thought probably light with wavelength shorter than 390 nm might do the work, so I tested the blue dye with UVA emissions of a corona discharge. Not surprisingly, the blue still shows negative response.
 Tipler, Paul A., "College Physics", Worth, 1987, pp.467
 Rigden, John S., "Macmillan Encyclopedia of Physics", Simon and Schuster, 1996, pp.353
 F. Arqueros, F. Blanco, F. Rosado, "Analysis of the Fluorescence Emission of Atmospheric Nitrogen by Electron Excitation and its Application to Fluorescence Telescopes", New J. Phys., 2009, pp.11
 H. Khatun, A.K. Sharma, P.K. Barhai, "Experimental Study of Low-Pressure Nitrogen Dielectric Barrier Discharge", Braz. J. Phys., Vol.40, No. 4, 2010, pp.450