USING DETAILED SAMPLING, GEOCHEMICAL ANALYSIS, AND MACHINE LEARNING ANALYSIS TO CHARACTERIZE LITHIUM BEARING SEDIMENTS IN THE WESTERN US AND BETTER UNDERSTAND THE FORMATION OF ECONOMIC VOLCANO-SEDIMENTARY LITHIUM DEPOSITS

The lithium (Li) used in battery technology is currently sourced primarily from high-altitude saline brines and spodumene-bearing pegmatites. In the coming years, Li-bearing volcano-sedimentary (VS) deposits are expected to become an increasingly important source of Li, with several VS Li deposits under development throughout the Basin and Range. Chief among these is the Lithium Americas Thacker Pass Project hosted in caldera lake sediments of the ~16.3 Ma McDermitt Caldera, Nevada. Because this type of deposit is critical to obtaining a secure domestic supply of Li in the United States, we initiated a study to ascertain the occurrences and formation of VS Li deposits and their key characteristics. Over the past three years, we traveled to >500 sedimentary basins in 21 states across the western U.S., carrying out observational field work, detailed sampling, and geochemical analyses from stratigraphic sections in the various basins. Through this work we created a database of over 1,500 samples from sedimentary basins in the western US. In combination with machine learning analysis, this data allows us to characterize features of each basin (e.g., chemistry, grain size, sedimentary history), correlate those features to the amount of Li trapped in the sediment to understand what creates an economic sedimentary Li deposit, and ultimately create a predictive model for Li concentrations in samples without complete chemical data. While sediments of the McDermitt Caldera contain anomalously high amounts of Li, few other sedimentary basins in the U.S. contain economic levels of Li (>1000 ppm). Basins that do contain >1000 ppm Li are predominantly characterized by Miocene-aged small (< 1,000 km²) lake deposits formed in closed systems containing a high proportion of fine-grained clay material. Sedimentary samples from the various basins were primarily defined by high variation in Ca and Mg based on PCA analysis, with Li most highly correlated with the behavior of Mg, K, V and Al. The physical property that most strongly correlates with Li concentration is grain size, with coarser material holding considerably less Li across the various systems. We interpret these preliminary analyses and observations to be indicative of high volcanic and/or hydrothermal input to create the reducing environment of the lacustrine systems that deposited economic concentrations of Li.