Identification of particulate materials by correlation analysis using a microscopic laser induced breakdown spectrometer

Abstract

The goal of this work was the instant identification of particulate geological materials by comparison with spectral libraries stored in a computer. The libraries consisted of representative spectra from different groups of powdered samples. Special attention was paid to identification of particles with very similar chemical composition, such as certain iron ores. Chemical speciation analysis was also carried out. Iron and iron oxides were shown to be reliably identified using statistical correlation methods. Both linear and rank correlations were applied. Both correlation methods yielded probabilities of correct identification close to unity for almost all studied samples. This technique should have applications in the metallurgical, mining, and semiconductor industries, and in medical, environmental and forensic sciences.