Laser Spectroscopy for Material Characterization: Chemical Analysis Using Laser-Induced Breakdown Spectroscopy (Libs)

Abstract

Laser-Induced Breakdown Spectroscopy (LIBS) is a powerful tool for performing chemical analysis measurements of materials, such as slurries, soils, plastics and powder samples. The LIBS technique has proven to be sensitive, selective and robust for rapid, in situ analysis. The focus of this dissertation is the optimization of laser spectroscopic sensing methodologies for material characterization. The applications of the LIBS technique to slurry samples is very challenging due to the water content (~80%). A new sample preparation method called “spin-on-glass” was adopted to reduce the water content in slurry samples and improve the LIBS signal. The feasibility of using the new sampling method with a LIBS system was tested by applying multivariate analysis to the LIBS spectral data. The calibration results demonstrated that the LIBS technique with the new sampling method could successfully predict the elemental concentrations of slurry samples qualitatively and quantitatively. The possibility of developing a LIBS-based sensor system for total carbon quantification in soil samples was studied. The soil samples were studied in pellet form and the calibration models were developed by using simple linear regression (SLR) and multiple linear regression (MLR) analysis. It was found that both SLR- and MLR-based calibrations successfully predicted the carbon concentration in an unknown sample with relative accuracy (RA) within 8%. The LIBS experimental setup was designed, developed and tested for the determination of elemental impurities in plastic calibration standards that are used in dual-energy computed tomography (CT) scanning for petrophysical applications. Univariate calibration (UC) and multiple linear regression (MLR) analysis were used to develop calibration models. From this study, it was concluded that MLR improved the calibration results and data derived from the LIBS analysis enhanced the predictive capabilities of dual-energy CT scanning in general. A comparative study was performed for quantification of strontium (Sr) in an aluminum (Al) batch with both the atomic and molecular LIBS emissions. The calibration models were developed using SLR analysis and the limits of detection (LOD) were obtained. The study confirmed that molecular LIBS could be used for quantification of Sr in a binary mixture.