"Relying on a single, centralized mine represents a risk": US wants rare earth independence from China — but can DARPA actually make it work?

• US targets processing bottleneck rather than searching for new rare earth deposits
• Parallel extraction concept seeks profitability despite higher domestic labor and environmental costs
• Distributed processing model attempts reducing reliance on single vulnerable mining sites

China is responsible for much of the world’s rare earth refining capacity, giving it control over supply chains during trade disputes. That advantage was built by handling the costly and messy processing stage at scale, often with lower costs and fewer environmental restrictions.

The United States has spent years trying to rebuild its rare earth supply chain, but mining alone hasn’t fixed the core problem. Processing remains the sticking point, and as Data Centre Dynamics reports, that’s where the Defense Advanced Research Projects Agency is placing a high-risk bet.

Rare earth elements aren’t, as the name suggests, truly rare, and the US already has access to large volumes of ore. DARPA's new Smash program moves away from finding new deposits and toward solving the processing bottleneck.

SMASH - Proposers Day - YouTube

Watch On

Near-zero-waste separation

DARPA’s approach centers on what it calls near-zero-waste separation across the periodic table. The goal isn’t just rare earth elements but up to 80 stable elements that could be recovered from existing ore and waste streams.

“So the challenge is processing, not mining,” said Julian McMorrow, Smash lead and program manager at DARPA’s Microsystems Technology Office. “We want to develop technologies to take the industry from wasting over 99 percent of its feedstock to making use of the entire feedstock.”

Traditional mining wastes enormous amounts of material during refinement. More than two tons of ore and 13 tons of water can produce just one kilogram of copper, leaving most of the original material discarded.

Smash explores a parallel processing model that attempts to extract nearly everything from a shovel of dirt at once. That concept borrows ideas from industries such as petroleum refining, where multiple outputs are separated efficiently from a single input.

The program also reflects concerns about relying on a single major site such as the Mountain Pass mine which once dominated global rare earth output but struggled when refining costs became uncompetitive.

DARPA notes that concentrating production in one location creates vulnerability if disruptions occur. A distributed model using varied feedstocks, including mining waste and recycled materials, could cut that exposure.

Smash will run as a 48-month effort split into two phases. The first will focus on proof-of-concept experiments, while the second will move toward working prototypes suitable for industrial mining environments.

Even if the technology succeeds in laboratory settings, scaling it economically could be tricky. Achieving profitability while maintaining strict environmental and labor standards will be the real test.

Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.