"Electron temperature measurements using the line-intensity ratio on the CTCC spheromak "
Y Kato et al.

Summary:

The CTCC-II experiments involve the creation of spheromaks with a toroidal current of 90 kA, an electron temperature of 20-80 eV, a density of 2-6e13 cm^-3, and a lifetime of about 1 ms. Using VUV spectroscopy, it was determined that after the inner surface of the flux conserver was coated with titanium (to reduce impurity concentrations), the dominant impurity in the plasma was oxygen. Emission line intensity ratios can be used as a measure of plasma electron temperature; in this study OIV, OV, and OVI lines were observed. However, because OIV and OV have metastable levels, their line ratios are dependent on the electron density, so care must be taken in interpreting the results.

Expected line intensity ratios were calculated analytically using cross-sections and impact parameters from previous papers. The intensity ratio of two lines of the same ionization stage whose upper levels are populated by collisional excitation from the ground level depends only on the electron temperature. This does not hold for ions with metastable levels, since their upper levels can be populated by electrons excited from a metastable level instead of from the ground level. Line intensity ratio measurements are appropriate for determining a temperature which is less than the energy difference between the upper levels of the two lines being considered. With this consideration in mind, the following three pairs of lines were chosen: OIV 79.0 nm and 23.9 nm, OV 63.0 nm and 17.2 nm, and OVI 103.2 nm and 15.0 nm.

Two VUV spectrometers were used: a grazing incidence one to measure in the region between 10 and 60 nm, and a normal incidence one to observe lines between 50-150 nm. Both could take data simultaneously, so the line ratios could be used to calculate the electron temperature for a single shot. The electron temperature was also measured by Thompson scattering with a ruby laser. The VUV spectrometers were calculated using the "branching-ratio" technique. This involves using pairs of lines (one in the visible and one in the UV) originating from the same upper level in an ion. A visible spectrometer is calibrated using two standard lamps and then used to measure the visible lines in each pair. If the transition probabilities are known, these measurements can then be compared to measurements of the UV lines by the VUV spectrometer to transfer the calibration from the visible to the VUV spectrometer.

Typical electron temperatures were found using observations of OVI to vary between 30-50 eV during the course of a shot. Measurements of OIV implied a temperature range of 20-40 eV. The OVI values were interpreted to correspond to the core of the plasma, and the OIV values to the periphery. The plasma lifetime depends on the product of the average electron density and total impurity density; the lifetime was also found to be longest for higher temperature runs. Thompson scattering measurements at t = 0.2 and 0.4 ms gave an electron temperature of 20-80 eV, in agreement with the temperature calculated from line ratios. Considerations of the lines of sight over which measurements were averaged led to the conclusion that the spheromaks had a flat or hollow temperature profile, with the temperature off the magnetic axis the same or higher than on the axis.

Previous studies of the CTCC spheromak found a stepwise decay in the plasma fields due to kink instabilities and relaxation--measurements of the time evolution of the electron temperature were used to further study this phenomenon. The temperature was found to peak during relaxation events. Variations in the electron temperature over time were shown to be larger in magnitude than the calibration errors for the experiment.

Applications to my research:

This paper is extremely relevant to my research at SSX--like the authors I am using measurements of line ratios with VUV spectroscopy to calculate the electron temperature in a spheromak plasma. The paper provides some useful theory on the use of the line ratio technique and its limitations. It is interesting that the authors were able to use line ratios for measurements without the aid of detailed computer simulations--this was achieved by only considering line pairs that were produced by ions of the same ionization state during electron transitions back to the ground level. This is a useful method for reducing the error in measurements (for example, there is no density dependence using their method), but we may not be able to implement it in SSX due to a limited number of lines that are strong enough for us to observe. We also cannot take advantage of their method for calibrating the VUV spectrometer, since we have only a single monochromator.