Dan VandenAkker & Ross Norman
Summer Research 2006
Download the printer friendly Word Document version of this page.
All the experiments on this page involved a radio-frequency oscillator circuit adapted from "Alkali Metal Vapor Spectral Lamps," W.E. Bell, A.L. Bloom, & J. Lynch, Rev. Sci. Instr. Vol. 32, pg 688 (1961). The diagram used to construct the cell is below.
Diagram of RF Circuit
Initial Experiment: create plasma in an enclosed tube of krypton gas.
This experiment involved the plasma discharge cell circuit discussed above, a 400V power supply, a bath of Liquid Nitrogen (LN2), and a glass tube of krypton gas. The power supply was connected to the plasma discharge cell circuit. The coils from the circuit were wrapped around the tube of krypton. The power supply was turned on, the lights were turned down, and the tube of krypton with the coils still wrapped around it was cooled in the LN2 bath. The LN2 bath caused the temperature inside the tube of krypton to drop significantly. Since temperature and pressure are directly related, the pressure also decreased dramatically. Once the pressure was low enough, the electric field generated in the coils caused the ignition of the plasma discharge. As the tube was slowly removed from the LN2 bath and the pressure slowly increased, the large cloud of plasma that had formed between the coils turned into a large ball-like mass of plasma centered between the coils. Larger coil separations would permit multiple large masses all equally spaced between the coils. As the pressure increased, the plasma balls tended to decrease in size and increase in number, appearing to move along the direction of the local electric field. This behavior continued until there were multiple strings of very small plasma masses, each resembling a string of pearls and traveling along the convection currents inside the tube. At some point as the pressure continued to increase, strings of pearls would whip around like "jump ropes" and branch out into the shape of a "pitchfork." More detailed discussion of this behavior can be found at University of Miami's Nonlinear Dynamics Laboratory website. We observed that multiple plasma strings began to repel each other and form periodic patterns around the interior of the glass tube. This additional pattern-formation behavior is closely related to the dielectric barrier discharge research published previously by our group. Below are videos featuring the behavior we observed.
|Examples with small (above), medium (above right), and large (right) coil separations. (Double click on each picture to view video.)|
Second Experiment: recreate the striations found in the initial experiment in a pressure controlled system
This experiment involved the plasma discharge cell circuit discussed above, the same 400V power supply, a pressurized supply of krypton gas, a gas regulator, a pump, a baratron pressure gauge, and additional vacuum parts. We used an old champagne bottle for our gas cell. Below is a photo diagramming the final setup.
Diagram of Vacuum Setup
With this setup, we were able to control the pressure inside the tube by adjusting inlet and outlet valves. Unfortunately, the vacuum setup didn't seem to be perfectly clean, and we therefore didn't have the purest krypton possible. By pouring LN2 into the "kick" of the inverted champagne bottle we were able to freeze some impurities, effectively forming a temporary getter in the vessel. We were able to produce the central mass of plasma as well as smaller plasma balls while monitoring the gas pressure. However, we were unable to reproduce the string-of-pearls striations that we found most interesting in the sealed cell.
Third Experiment: determine the plasma intensity's dependence upon pressure and the plasma ball diameter's dependence upon pressure.
This experiment involved the same setup as the second experiment. However, we also used a video camera with which we zoomed in tight on the region of the champagne bottle where the plasma was created, as well as on the pressure gauge. From each video recording we grabbed frames and analyzed them with Scion Image software. By tracing a line over one plasma ball in Scion Image, we could obtain a graph of optical intensity versus pixel number. This graph could be analyzed for intensity and diameter.
Screen Snapshot of frame analysis with Scion Image software
Scion Image plot with measurements qualitatively described
Excel graphs of our measurements
The the relatively large uncertainties in these measurements are reflected in the vertical scatter of the graphs. The only thing that is sytematically verifiable is that the size of the plasma balls generally decreases as the pressure increases. It may be easier to see this in the videos taken than in the graphs. Below are some of the frames that were pulled from the video footage.
Four images of plasma masses at various pressures between two and eight torr, as measured by a baratron gauge.