A machine learning model has been able to predict the locations of minerals on Earth — potentially even on other planets — possibly paving new ways to find rare raw materials used in technologies such as rechargeable batteries.
Researchers from the Carnegie Institution of Sciences sought to create a tool for finding specific minerals by taking advantage of patterns found in mineral associations. Typically, this is used to better understand the history of our planet, but it could also be used to extract these minerals — specifically those that are rare or even contested on Earth.
The machine learning model used data from the Mineral Evolution Database, which includes 295,583 mineral localities of 5,478 mineral species to predict previously unknown mineral occurrences based on association rules.
Researchers tested the model by exploring the Tecopa basin in the Mojave Desert and was able to predict the locations of geologically important minerals such as:
- Uraninite alteration
- Rutherfordine
- Andersonite
- Schröckingerite
- Bayleyite
- Zippeite
Additionally, the model located promising areas for critical rare Earth element and lithium minerals like monazite-(Ce), and allanite-(Ce) and spodumene.
The full research can be found in the journal PNAS Nexus.