In the scope of the exploration of the Lunar surface and assessment of its resources, reliable and quantitative methods need to be employed. The majority of currently known information about the Lunar surface composition comes from the retrieved samples and remote sensing from the Lunar orbit. In this study we introduce a novel matrix-based calibration methodology for Laser Induced Breakdown Spectroscopy (LIBS) by coupling its data to Scanning Electron Microscopy – Energy Dispersive X-Ray spectroscopy (SEM-EDX) reference data. The accuracy of the proposed methodology and its relevancy for Lunar exploration is showcased by the analysis of a lunar meteorite sample, where the SEM-EDX data from two diverse areas of the sample are used as training data and one area is used for the validation of the results of the methodology. Comparing the elemental composition from the SEM-EDX validation area and calculated LIBS data of the same area, the overall bias of the composition between all elements reached less than −0.5 wt% and the average mean absolute percentage error (MAPE) reached less than 15%. The methodology also achieved reliable and precise detection of matrices present in less than 0.1% of the total area of the sample enabling quantification of rare minerals like ilmenite or chromite. The usage of the methodology may improve the level of acquired LIBS results on extraterrestrial bodies from semi-quantitative to quantitative. This methodology is designed to be universally applicable to any LIBS and spatially resolved reference method relationship, thus increasing the potential impact outside of the extraterrestrial or even geological applications. • Comprehensive 2D characterization of a lunar meteorite sample. • Quantitative imaging of the large-scale cross-section of the meteorite sample. • Novel clustering-based calibration approach, merging SEM-EDX into LIBS. • Chemistry- and matrix-based calibration. • Reliable chemistry quantification of a heterogeneous sample including trace minerals.
Holub et al. (Sun,) studied this question.