**4. Discussion**

According to the obtained results, both presented methods provide about the same values of electronic densities, but the second method leads to temperature values higher (around a 1.7 factor) than those obtained by the first procedure, as can be seen in Table 5. In this table, values of 5, 3, and 2 μs for the delay times and 500 ns for the time gate (with the best time of integration of the light) were used.

The authors consider that this deviation (that must not be considered as dramatic provided the general uncertainties considered in the kind of plasma diagnosis envisaged in the paper) may be attributed to the inherent inaccuracy of the procedure used to estimate the Hα-line shifts. Finer adjustments of the profile of the line (beyond a Voigt profile and considering asymmetries) are expected to yield more precise values of the temperatures using the described combinations of Stark widths and Stark shifts as described in the second method. In view of the promising character of this method for a practical determination of plasma parameters in one single step, these kinds of improvements are under way and are the subject of upcoming research.

The most important feature of the presented development and its comparison to the standard procedures is that within the experimental uncertainties, the authors consider that the described second method is sufficiently precise and practical to be implemented at an industrial level of LSP applications.

**Table 5.** Electronic temperatures obtained from Boltzmann plot compared with the temperatures estimated from red shifts of the Hα-line obtained in this work at 5, 3, and 2 μs delay times and 500 ns of time gate.


In view of the results obtained with the 500 ns window, we think that this second method based on both the Stark broadening and Stark shift of the Hα-line allows for the diagnosis of the plasma in industrial conditions, with good precision for the electron density and an acceptable uncertainty for temperature
