The mineral exploration industry requires new methods and tools to address the challenges of declining mineral reserves and increasing discovery costs. Laser-induced breakdown spectroscopy (LIBS) represents an emerging geochemical tool for mineral exploration that can provide rapid, in situ, compositional analysis and high-resolution imaging in both laboratory and field and settings. We demonstrate through a review of previously published research and our new results how LIBS can be applied to qualitative element detection for geochemical fingerprinting, sample classification, and discrimination, as well as quantitative geochemical analysis, rock characterization by grain size analysis, and in situ geochemical imaging. LIBS can detect elements with low atomic number (i.e., light elements), some of which are important pathfinder elements for mineral exploration and/or are classified as critical commodities for emerging green technologies. LIBS data can be acquired in situ, facilitating the interpretation of geochemical data in a mineralogical context, which is important for unraveling the complex geological history of most ore systems. LIBS technology is available as a handheld analyzer, thus providing a field capability to acquire low-cost geochemical analyses in real time. As a consequence, LIBS has wide potential to be utilized in mineral exploration, prospect evaluation, and deposit exploitation quality control. LIBS is ideally suited for field exploration programs that would benefit from rapid chemical analysis under ambient environmental conditions.

The Cadia East porphyry deposit, located approximately 20 km south of Orange, New South Wales, Australia, contains a significant resource of copper and gold. This resource is hosted within the Forest Reefs Volcanics and is spatially and temporally associated with the Cadia Intrusive Complex. To extract ore, the underground mine currently uses the block cave mining method. The Cadia East geotechnical model provides data inputs into a range of numerical and empirical analysis methods that make up the foundation for mine design. These data provide input into the construction of stress models, caveability models, ground support design, and fragmentation analysis. This geotechnical model encompasses two commonly used rock classification systems that quantify ground conditions: (1) rock mass rating (RMR) and (2) rock tunneling quality index (Q index). The RMR and Q index are calculated from estimates of rock quality designation (RQD), number of fracture sets, fracture roughness, fracture alteration, and fracture spacing. Geologists and geotechnical engineers collect information used to produce these estimates by manually logging sections of drill core, a time-consuming task that can result in inconsistent data. Modern automated core scanning technologies offer opportunities to rapidly collect data from larger samples of drill core. These automated core logging systems generate large volumes of spatially and spectrally consistent data, including a model of the drill core surface from a laser profiling system. Core surface models are used to extract detailed measurements of fracture location, orientation, and roughness from oriented drill core. These data are combined with other morphological and mineralogical outputs from automated hyperspectral core logging systems to estimate RMR and the Q index systematically over contiguous drill core intervals. The goal of this study was to develop a proof-of-concept methodology that extracts geotechnical index parameters from hyperspectral and laser topographic data collected from oriented drill core. Hyperspectral data from the Cadia East mine were used in this case study to assess the methods. The results show that both morphological and mineralogical parameters that contribute to the RMR and Q index can be extracted from the automated core logging data. This approach provides an opportunity to capture consistent geologic, mineralogical, and geotechnical data at a scale that is too time-consuming to achieve via manual data collection.

This study evaluates the effectiveness of four methods for estimating mineralogy from micro X-ray fluorescence (µXRF) elemental maps: two optimized/supervised methods (linear programming and Random Forests) and two unsupervised methods (k-means clustering and Uniform Manifold Approximation and Projection (UMAP) combined with Hierarchical Density Based Spatial Clustering of Applications with Noise (HDBSCAN)). µXRF is a non-destructive, high-resolution imaging technique that provides spatially resolved elemental data and bridges the scale gap between micro- and macroscales in mineralogical studies. Analysing samples from porphyry copper deposits (representing a range of textures and mineralogical complexities), each method was assessed based on its ability to accurately represent the mineralogy and texture of each sample relative to a petrographic image. Compared to traditional petrographic assessment, Random Forests yielded the lowest normalized root mean squared error (NRMSE) values (0.09–0.38), while k-means (0.06–0.44), linear programming (0.07–0.44) and UMAP–HDBSCAN (0.18–0.64) showed more variable and generally higher error values across the samples. k-Means and linear programming produced results within minutes, Random Forests in ∼10 minutes and UMAP–HDBSCAN in ∼20 minutes. The results demonstrate that supervised methods, particularly Random Forests, deliver better mineralogical estimates, especially in complex mineral assemblages. However, unsupervised methods (particularly k-means) offer a faster initial assessment, making them valuable for preliminary analysis. Combining different methods using known mineral compositions or training datasets could enhance cluster-to-mineral correlation. The integration of supervised, optimization and unsupervised approaches for mineralogy from µXRF data can enhance the robustness and efficiency of mineralogical interpretations, adding significant value across mineral exploration and processing.