Raman Quantitative Models of Silicate Ternary Mixtures: Factoring in the Effect of Glass Content

ABSTRACT Raman spectroscopy is a powerful nondestructive tool for analyzing the mineral composition of planetary surface soils. However, quantifying the modal mineralogy remains a challenge, especially for lunar regolith due to the low Raman scattering cross sections of key silicate minerals and the interference from glass components. This study developed quantitative models for determining the mineral modes of lunar soil simulants, factoring in the effect of glass content on Raman‐based mineral quantification. The simulant materials, including feldspar, olivine, augite, and glass, were mixed in varying mass fractions to create 24 different mixtures. A Gaussian–Lorentzian fitting method was employed to analyze the Raman spectra and extract peak areas, which were quantitatively correlated with the mineral proportions in the mixtures. The quantitative models were built using the linear relationship between Raman peak areas and mineral proportions. These models demonstrated high accuracy with correlation coefficients ( R 2 ) exceeding 0.94 and root mean square errors (RMSEs) less than 5.16 wt.% for the glass‐free samples. Further analyses showed that the inclusion of up to 40 wt.% glass reduced only the area of Raman peaks, particularly for augite, but the models remained reliable for predicting mineral content. This work provides a reliable approach for quantitative mineral analysis of lunar soils, offering valuable insights for future planetary exploration missions.