Simulation of emission spectra from nonuniform reactive laser-induced plasmas

Abstract

We demonstrate that chemical reactions leading to the formation of AlO radicals in plasmas produced by ablation of aluminum or Ti-sapphire with ultraviolet nanosecond laser pulses can be predicted by the model of local thermodynamic equilibrium. Therefore, emission spectra recorded with an echelle spectrometer and a gated detector were compared to the spectral radiance computed for uniform and nonuniform equilibrium plasmas. The calculations are based on analytical solutions of the radiation transfer equation. The simulations show that the plasmas produced in argon background gas are almost uniform, whereas temperature and density gradients are evidenced in air. Furthermore, chemical reactions exclusively occur in the cold plume periphery for ablation in air. The formation of AlO is negligible in argon as the plasma temperature is too large in the time interval of interest up to several microseconds. Finally, the validity of local thermodynamic equilibrium is shown to depend on time, space, and on the elemental composition. The presented conclusions are of interest for material analysis via laser-induced breakdown spectroscopy and for laser materials processing.

Two-step procedure for trace element analysis in food via calibration-free laser-induced breakdown spectroscopy

“…The values of gate delay and gate width applied in both measurements were chosen based on the knowledge acquired in previous investigations. These studies evidenced that accurate compositional measurements of oxygen-containing materials such as glass and sapphire are possible only for recording times < 1 µs, when the electron density is large enough to ensure collisional equilibrium [25,26]. For later recording times, the atom with the largest energy gap between excited states, namely oxygen, was shown to run out of equilibrium, leading to an overestimation of the oxygen fraction.…”

“…We consequently expect for the here chosen early measurement time of 0.35 µs Boltzmann equilibrium distributions for all elements of the organic samples. The choice of the late measurement delay is founded on calibration-free analysis under equal experimental conditions, showing that accurate compositional analyses of metals can be performed with measurement times as late as several microseconds [26]. We stress that, even when the equilibrium conditions are not fully satisfied, the resulting measurement errors of the minor and trace elements are expected to be small compared to statistical variations due to the heterogeneity of the samples [33].…”

“…For this measurement, the gate width ∆t gate is set small enough so that the variations of temperature T and electron density n e during the time of observation are small compared to their absolute values. We have shown in previous studies that a gate width ∆t gate ≤ t gate /2 is a reasonable compromise between model validity and signal-to-noise ratio [25,26]. The late measurement is performed with a large gate width to reach highest signal-to noise ratio required for trace element detection.…”

“…Consequently, calibration-free laser-induced breakdown spectroscopy (CF-LIBS) was mostly applied to measure elements with sufficiently high concentration values. In addition, the application of CF-LIBS to organic materials encounters a supplementary difficulty that arises from matrix elements C, H, O and N. According to the atomic structures including large energy gaps between the electronic excitation levels, equilibrium Boltzmann distributions are hardly achieved for these elements [25,26]. The matrix elements are therefore generally ignored and CF-LIBS measurements are limited to a partial analysis of the elemental composition of food [17,12,18,11,19].…”

The effect of hyperfine splitting on Stark broadening for three blue-green Cu i lines in laser-induced plasma

“…Intensities were normalized to unity to outline temporal evolution of profiles of the Cu I lines in (Hermann et al 2015), we used only one side of the profile for fitting with Lorentzian function (Figure 5, b). We combined Lorentzian function with a linearly sloping background to fit the line at 5218.20 Å to diminish slight asymmetry (Figure 5, c) of the profile caused by the wing of strong Mg I 5183.6 Å.…”