Temperature based segmentation for spectral data of laser-induced plasmas for quantitative compositional analysis of brass alloys submerged in water

This study describes a method to quantify the composition of brass alloys submerged in water using laser-induced plasmas. Principal component regression (PCR) analysis and partial least squares (PLS) regression analysis are applied to spectral measurements of plasmas generated using a long-ns duration pulse. The non-linear effects of excitation temperature fluctuations on the signals are treated as systematic errors in the analysis. The effect of these errors on the analytical performance is evaluated by applying PCR and PLS with a temperature segmented database. The results of the analysis are compared to conventional methods that do not consider the excitation temperature and it is demonstrated that the proposed database segmentation improves accuracy, with root-mean square errors of prediction (RMSEP) of 2.7% and 2.8% for Cu and Zn in the PCR model and 2.9% and 1.8% for Cu and Zn in the PLS model, respectively. The results indicate that systematic effects contribute to fluctuation of underwater plasmas, where appropriate database segmentation can improve the performance of the PCR and PLS methods.