Cracking the clay code as CSIRO’s multiscale mineralogy reveals how rare earths hide in WA sands and why smart mineralogy is the new exploration edge

When it comes to rare earths, the real magic isn’t in the elements, it’s in the minerals that host them. That was the message from Dr Gabriel Berni, team leader at CSIRO, who presented “Multiscale mineralogy of clay-hosted REE deposits in Western Australia – practical applications for exploration, domaining and resource modelling” at the Critical Minerals Conference 2025 in Perth.

The paper, co-authored with Heta Lampinen, Rong Fan, Michael Verrall, Nathan Reid, Robert Thorne, Mario Iglesias-Martinez, Catherine Spaggiari, Yoram Teitler, Jo Miles, and Mark Pearce, explored how advanced mineralogical tools are giving explorers a new lens on Australia’s clay-hosted rare earth element (REE) systems.

Getting the scale right

Gabriel’s talk cut through the complexity of rare earth mineral systems with a simple point: scale matters. “We approached the deposits from a multiscale perspective, from regional mapping right down to the mineral grain,” he explained. “That’s the only way to accurately define mineralisation boundaries, understand how the elements are hosted, and predict their processing behaviour.”

The study compared two Western Australian REE prospects, Splinter Rock and Balladonia, both within the Albany–Fraser Orogen, but representing distinct geological settings. Using a combination of HyLogger-3 hyperspectral scanning, X-ray diffraction (XRD), and scanning electron microscopy (SEM) paired with TIMA automated mineralogy, the team dissected the deposits from surface regolith to bedrock.

Splinter Rock beneath the surface

At Splinter Rock, mineralisation sits beneath marine and lacustrine sediments overlying the Booanya Granite. Here, the rare earths are hosted by secondary phosphates and carbonates, the products of deep weathering processes dating back to the Cretaceous.

The CSIRO team differentiated key regolith boundaries including transported cover, saprolite, and saprock using distinct spectral and geochemical signatures. “Bound water and IR reflectance spectra helped us map the transition between transported material and in-situ regolith,” said Gabriel. “We could even identify the saprolite–saprock contact by its sharp potassium jump.”

This level of resolution, he noted, isn’t just academic. It underpins the accuracy of domaining models and resource estimations. “If you start wrong in defining your sampling domain,” he warned, “you’re going to be wrong all the way through.”

Balladonia the ionic frontier

Balladonia, meanwhile, tells a different story. Here, REE enrichment occurs within clay fractions of coarse-grained sands rather than in the underlying regolith. Hosted in palaeovalley sediments above the Hiltaba Granite, the mineralisation is associated with fine-grained clays and titanium oxides, a signature of ionic-type systems similar to those mined in China.

By integrating particle size analysis, spectral data, and mineral mapping, Gabriel’s team found that REE grades correlate strongly with the fine clay fraction. The conclusion: ionic adsorption is the dominant enrichment mechanism. “We’re seeing that the rare earths are loosely bound to the clay surface, which opens up potential for in-situ recovery using mild leaching conditions,” Gabriel explained.

For explorers, that’s a crucial insight. It points to extraction pathways that are both technically and environmentally more feasible than hard-rock mining.

From discovery to domain modelling

Beyond the mineralogy, the real innovation lies in workflow. By combining continuous spectral scanning, handheld validation, and automated mineralogy, the CSIRO workflow allows explorers to map boundaries, identify mineral hosts, and predict metallurgical behaviour in real time. “You don’t need to wait until the resource stage to understand your ore,” Gabriel said. “These tools give you actionable intelligence early, during exploration and drilling.”

He also highlighted CSIRO’s LandScape+® and EM data processing capabilities, which help explorers model regolith thickness and palaeovalley preservation, both key factors in identifying where clay-hosted systems are best developed.

Shifting the exploration mindset

Gabriel’s work challenges the traditional perception that REE exploration is a geochemical guessing game. Instead, it’s about seeing the mineral system as a spectrum, not a single layer. “We’re dealing with deposits that don’t always show clean stratigraphic control,” he said. “You need to integrate every scale of observation, from spectral reflectance to micro-texture, if you want to understand where the value lies.”

For mining professionals, the takeaway is clear. As the energy transition accelerates and Australia positions itself as a non-Chinese supplier of rare earths, the competitive edge will go to explorers who can characterise deposits before they’re defined. In Gabriel’s words, “the smarter you are about your mineralogy, the faster you’ll be about your metallurgy.”