Kvantifikace minerálního složení hornin pomocí Ramanovy spektroskopie

Abstract

This diploma thesis was focused on the use of Raman spectroscopy in combination with chemometric methods to determine the content of albite, chlorite, muscovite and quartz in selected samples of sedimentary rocks (in claystones and clay slates). Partial least squares regression was chosen for the creation of chemometric calibration models. Raman spectra were obtained by measurements using three lasers (at wavelengths of 532 nm, 633 nm, and 780 nm) and two gratings (each providing a different spectral resolution). The partial least squares regression algorithm was unable to find a correlation between the spectral information and albite content nor between spectral information obtained using a full-range grating (grating with a lower spectral resolution) and the contents of all studied minerals. Other calibration models were used to predict chlorite, muscovite and quartz contents. The values of relative errors for determining the mineral content ranged from 0,5 % to 62,2 %.