CSIRO makes the case for hyperspectral prospecting for REEs
Rare earth elements (REE) spectra. Image Kuse (1967) edited by AuScope
CSIRO’s principal research scientist in CSIRO’s Mineral Resources business unit, Carsten Laukamp, told the Australian Earth Sciences Convention last week that explorers could do well to search for REE deposits in Australia’s huge stores of hyperspectral data.
Value of spectral data
Carsten Laukamp presented a series of charts that show many of the 14 REEs have visible signals in the Very Near Infrared (VNIR) spectral range, which is one of the spectral ranges measured by the CSIRO-invented NCRIS enabled HyLogger that has been commercialised with Australian METS Corescan Pty Ltd.
He noted there are some limitations, mainly around the presence of iron oxides which could partially mask the spectral signatures of REEs.
But he showed just how valuable spectral data can be in exploration by showing the results of spectral scans in the National Virtual Core Library of drillcore from several of Australia’s REE deposits and occurrences. REEs at Mt Weld, both carbonate-hosted and phosphate-hosted, have a very strong spectral signature. Less prominent but still with distinctive spectral signatures are Browns Range, Nolans Bore and Toongi.
NVCL is funded by AuScope and CSIRO and strongly supported by the Australian state and territory geological surveys who have created the NVCL by logging huge volumes of open file drill core.
Light rare earth element Neodymium has the strongest reflectance spectral signature of all the REEs. Carsten even showed a satellite spectral image where DLR and Teledyne Browne Engineering detected a Nd-signatures at Mt. Pass, California, from space. The more valuable heavy rare earth elements also have their own signatures, which he set out in a very useful table of the absorption wavelengths of all the REEs and selected transition metals.
Carsten’s presentation also makes the point that spectral data can be used to determine a number of the common minerals that host REEs. The type of host mineral, especially for clay-hosted REEs, can make-or-break the economics of a potential development.
“Spectral data can be used at a very early stage of exploration to decide if a target is worth pursuing. If the REES are sitting in xenotime or phosphates, then that could be all good. But if they are in certain clay minerals, they might be more difficult to process.
The economic viability of a mineral deposit depends very much on the host minerals of the respective commodity, and the industry has been waking up to this over the past five years.
It’s not only the geochemistry, but the mineralogy is very important,”
Representative VNIR-SWIR reflectance spectra of major REE-element bearing minerals. Image: Carsten Laukamp
Carsten said the strength of spectral data as an exploration tool was its ability to be used across a full range of scales – from microscopic examination of samples, through field detection with hand-held instruments and, of course, satellite imagery.
“It’s a great tool to link other datasets such as geochemistry and geophysics across different scales.”
He said Australian explorers had a greater opportunity than peers in other countries to leverage spectral data for discovery because of the NVCL and the community of experts that had developed around this core piece of precompetitive infrastructure.
“The Australian community of what we like to affectionately call spectral nerds is one of the biggest and broadest of its kind. It is made up of people with many different backgrounds and interests in a wide range of commodities, including copper, nickel, lithium, REEs and vanadium.
Geological surveys in other countries are looking at the large spectral community in Australia and our publicly available spectral data sets and are working towards establishing their own virtual core libraries.”
Carsten does not claim that spectral data has all the answers. He said it was one piece of an exploration tool kit for REEs, and it had its limitations, whether it was the masking potential of iron or the fact that some REEs just can’t be recognised in the spectral data.
“But more research is being done by CSIRO and others, and there is the real possibility that the revolutionary improvements in sensors over the past decade will continue and open up even more possibilities for exploration and ore body characterisation.”
AUTHOR
Re-published with permission from David Upton - Precompetitive Review
Edited by Philomena Manifold
MORE INFORMATION
National Virtual Core Library (NVCL)
Access NVCL Data
Dr Carsten Laukamp