Classification of Copper Minerals by Handheld Laser-Induced Breakdown Spectroscopy and Nonnegative Tensor Factorisation.

Laser-induced breakdown spectroscopy (LIBS) analysers are becoming increasingly common for material classification purposes. However, to achieve good classification accuracy, mostly noncompact units are used based on their stability and reproducibility. In addition, computational algorithms that require significant hardware resources are commonly applied. For performing measurement campaigns in hard-to-access environments, such as mining sites, there is a need for compact, portable, or even handheld devices capable of reaching high measurement accuracy. The optics and hardware of small (i.e., handheld) devices are limited by space and power consumption and require a compromise of the achievable spectral quality. As long as the size of such a device is a major constraint, the software is the primary field for improvement. In this study, we propose a novel combination of handheld LIBS with non-negative tensor factorisation to investigate its classification capabilities of copper minerals. The proposed approach is based on the extraction of source spectra for each mineral (with the use of tensor methods) and their labelling based on the percentage contribution within the dataset. These latent spectra are then used in a regression model for validation purposes. The application of such an approach leads to an increase in the classification score by approximately 5% compared to that obtained using commonly used classifiers such as support vector machines, linear discriminant analysis, and the k-nearest neighbours algorithm.

Diagnostic of laser-induced plasma using Abel inversion and radiation modeling.

A method based on matching synthetic and experimental emissivity spectra was applied to spatially resolved measurements of a laser-induced plasma ignited in argon at atmospheric pressure. The experimental emissivity spectra were obtained by Abel inversion of intensity spectra measured from a thin plasma slice perpendicular to the plasma axis. The synthetic spectra were iteratively calculated from an equilibrium model of plasma radiation that included free-free, free-bound, and bound-bound transitions. From both the experimental and synthetic emissivity spectra, spatial and temporal distributions of plasma temperature and number densities of plasma species (atoms, ions, and electrons) were obtained and compared. For the best-fit synthetic spectra, the temperature and number densities were read directly from the model; for experimental spectra, these parameters were obtained by traditional Boltzmann plot and Stark broadening methods. In both cases, the same spectroscopic data were used. Two approaches revealed a close agreement in electron number densities, but differences in plasma excitation temperatures and atom number densities. The trueness of the two methods was tested by the direct Abel transform that reconstructed the original intensity spectra for comparing them to the measured spectra. The comparison yielded a 9 and 13% difference between the reconstructed and experimental spectra for the numerical and traditional methods, respectively. It was thus demonstrated that the spectral fit method is capable of providing more accurate plasma diagnostics than the Boltzmann plot and Stark broadening methods.