Temperature determination in non-LTE plasmas

The temperature determination of plasmas in LTE, or in reality close-to-LTE, from the ratio of line intensities is a straightforward method which in principle depends only on the accuracy of line intensity measurements, the accuracy of the values of transition probabilities of the lines and the energy differences of upper levels in relation to the temperature. Even in these "simple" circumstances it turns out that the use of the "Boltzmann plot" method leads to large practical problems. Methods that include in some way the ground state density or ion density are much more favorable. Errors or uncertainties in the (electron) temperature may lead to a misinterpretration of the plasma state. In inductively coupled plasmas (ICP), close-to-LTE, it is possible to determine the electron temperature from the Saha equation with measurements (by Hß broadening) of the electron density and of the densities of highly excited states. In non-LTE circumstances it is improper to use the "Boltzmann plot" method. One has to account for the stage of departure from equilibrium of the plasma. This stage is determined by the dominant excitation equilibrium and can be classified into the corona, excitation saturation, capture radiative cascade or partial LTE regime. In all these circumstances it is possible to obtain information from line intensities, since these are determined by the electron temperature and density, as well as the density of the ground and/or ion state. It is clear, however, that a determination of all these parameters asks for more information than is provided by line intensity measurements. In the excitation saturation regime it is sometimes possible to determine the electron temperature by taking into account the specific dependency of the density of levels on the principal quantum number of the levels.