Laser Induced Breakdown Spectroscopy (LIBS) for whole rock geochemistry

To meet the ever-increasing global demand for metals, new mineral deposits need to be discovered. However, as the exploration frontier progressively moves towards deeper buried prospective areas, sampling is becoming costlier and deposit discovery rates are in decline. New technologies and analytical procedures able to deliver fast, reliable, and cheaper samples, analysis, and data, compared to current procedures are therefore, crucial in decreasing the risk of exploration targeting and can represent a step change in mineral exploration. In this study we demonstrate how to emulate laboratory whole-rock geochemical data for major elements (Al, Ca, Fe, K, Mg, Na, Si, and Ti) by averaging LIBS spot analyses performed over single, 1-mm spaced transects, that imitate downhole trajectories, along similar to 1-meter length drill core intervals for chemically, texturally, and mineralogically diverse rocks. Data was collected using a benchtop prototype LIBS instrument that could be reconfigured to fit within a >= 75 mm diameter drill hole. The prototype comprises a customised 8 ns pulsed 532 nm Nd:YAG laser with maximum pulse energy of 40 mJ at 5 Hz; two high-resolution spectrometers that together cover a spectral range from 190 to 830 nm, hence allowing all elements to be analysed; a motorised stage; and optics. Estimated LIBS geochemistry shows strong correlations with laboratory whole rock geochemistry for the selected major elements, particularly for Si, Al, and Na, and to a lesser extent K, although the LIBS estimation can breakdown for elements found in low concentrations (e.g., 0.06 wt%). Factors including the variability in mineralogy and element deportment, or roughness of the analysed surface can result in under- or over-estimation in LIBS whole rock geochemistry. In addition, for more efficient data collection processes, investigations of the optimal number of LIBs spot analysis showed that optimal similar to 30 to similar to 560 spot analyses per meter are required to emulate whole rock geochemistry (e.g., 1 % error and 95 % confidence). The strongest influence on the optimal number of LIBS analysis required for estimation of whole rock geochemistry is the range of element concentration and to a lesser extent the number of minerals that host an element as well as grain size. The results of this study show promise in development of strategies for rapid whole rock geochemical analysis down a drill hole.