Abstract

In this study, a machine learning method and a piece of cassiterite trace element composition data were used to find fingerprint trace elements that distinguish different tin (Sn) mineralization types and build tools for exploring primary Sn deposit exploration. The trace element dataset of cassiterite from the granite-related Sn metallogenic system was built using the following two approaches: (1) by analyzing the cassiterite samples from nine Sn deposits in Myanmar using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) and (2) collecting published data. The resulting dataset contains 661 records of 12 trace elements in cassiterite from 4 mineralization types: pegmatite, greisen, quartz-vein, and skarn. The spider diagrams of trace elements and the principal component analysis indicate that cassiterite samples from the same one mineralization type are clustered together and have unique geochemical characteristics. Using recursive feature elimination with cross-validation and simulation, tantalum (Ta), niobium (Nb), manganese (Mn), hafnium (Hf), iron (Fe), scandium (Sc) and Sc/Ta, Sc/Hf, and Sc/Mn were selected as the fingerprint elements and ratios, respectively. The cluster distribution in the biplots of the fingerprint elements and ratios indicates that these fingerprints are sensitive to the mineralization type. The distribution in the biplots also reveals that there is an evolutionary sequence of magmatic–hydrothermal fluids from pegmatite to skarn at the cassiterite crystallization environment. This data-driven study improves our understanding of the isomorphism between Sn and ions with similar charge and ionic radii in cassiterite from different hydrothermal environments. The complementary relationships between vanadium (V), Nb, and Ta are also identified. The element V is preferred to form a charge balance pair with Sc in cassiterite from skarn, whereas Ta and Nb are preferred to constitute a charge balance pair in cassiterite from the quartz-vein, pegmatite, and greisen. Our findings demonstrate that trace element compositions of detrital cassiterite grains from stream sediments can be used as an exploration tool to discover concealed primary Sn deposits and to evaluate the economic value based on the grade-tonnage model in the preliminary stage of mineral exploration.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call