Applications of stimulated Raman scattering microscopy to mineralogy: Real-time quantification of spodumene in lithium-bearing ores

Lithium is a critical component in rechargeable battery chemistries, leading to international interest in efficient ore characterization methods. Current industry-standard automated mineralogy systems (EDX) provide high-resolution mapping but are challenged by long analysis times and intensive sample preparation. Here we demonstrate the application of Stimulated Raman Scattering (SRS) microscopy to the rapid mineral-specific quantification of spodumene and other major species within lithium-bearing ores. By combining spectral-focusing SRS with Partial Least Squares Discriminant Analysis (PLS-DA), we identified a reduced set of high-information Raman peaks which facilitate rapid material classification. This subsampling approach allows for a two-order of magnitude reduction in the number of required spectral points while maintaining classification accuracy. These results show quantitative agreement with QEMSCAN measurements, achieving for the samples under study an average spodumene area percentage of 23.96 ± 5.82%, compared to the reference value of 23.01%. Combined with complete analysis times of under an hour and minimal sample preparation, this new method provides a viable pathway to real-time feedback in process mineralogy.