Data fusion method to identify light elements bearing minerals from the Li-Beauvoir granite based on SEM, µXRF, and µLIBS

• Demonstrated successful data correlation across multiple characterization methods. • µLIBS effectively identified lithium-bearing mineral phases. • Reduction of the overall analytical volume using spectrum deconvolution and data correlation. • Classified mineral map exhibits textural consistency with automated mineralogy tool. Understanding the modal mineralogy, composition, and grain properties (such as size, texture, and liberation) is crucial for optimizing mineral processing techniques, particularly in separation processes such as froth flotation, which depend on mineral surface properties. Current methods for mineralogical analysis, such as automated mineralogy tools ( e.g. , MLA, QEMSCAN), provide detailed insights but are time-consuming and costly. Emerging techniques, including micro-X-ray fluorescence (µXRF), offer more accessible alternatives but come with limitations, particularly in spatial resolution and spectral interference between elements. This study presents a novel method combining Scanning Electron Microscopy (SEM) imaging, µXRF, and Micro Laser Induced Breakdown Spectroscopy (µLIBS) analyses, applied to the Beauvoir granite, known for its light element-bearing minerals with substantial lithium, beryllium, and fluorine contents. By calibrating µLIBS mapping with in situ Laser Ablation Inductively Coupled Plasma Mass Spectrometry (LA-ICPMS) assays and using a custom Python-based classifier for spectral data, the method achieves accurate mineralogical mapping, particularly for light element-bearing phases. Results demonstrate strong agreement with commercial automated mineralogy tools and improved classification of complex mineral phases. In this study, lepidolite was effectively identified from muscovite in coarse flotation concentrates, the first one representing about 65 % of the volume, whereas the latter represents from 7 to 16 % of the sample. This indicates the potential of the method for advancing mineral processing and metal recovery. In addition, thisstudy suggests that the low resolution of some analytical methods (∼20 µm) could be improved if fused with a high-resolution mapping method, such as SEM (∼1 µm). This approach is particularly relevant as light elements gain strategic importance, especially within the context of European mineral resource strategies.