America mines rare earths. The problem has always been what comes next

The Mountain Pass mine in California’s Mojave Desert is the only active rare earth mine in the United States, and in 2025 it pulled roughly 50,000 metric tons of rare earth oxides out of the ground, close to a tenth of world supply. Until early 2024, though, most of that ore left the country anyway, shipped to China for the step that actually turns it into something usable. Digging rare earths out of rock, it turns out, is the easy part. Separating the tangle of 17 nearly identical elements into pure, industrial-grade materials is the part almost no one outside China has learned to do at scale.

Mining produces ore. Separation produces a product

Rare earth ore comes out of the ground as a mixed concentrate, with most or all of the 17 elements bound together in the same mineral grains, alongside radioactive thorium and uranium as unwanted companions. That concentrate is nearly worthless to industry as is.

A magnet needs pure neodymium and praseodymium; a laser or a nuclear-reactor control rod needs pure europium or gadolinium. Turning a barrel of mixed concentrate into 99-percent-pure separated oxides is a distinct, capital-intensive branch of chemical engineering, and it is where nearly all of the industry’s value, and nearly all of its bottleneck, actually sits.

China accounts for roughly 60 percent of global rare earth mining but close to 90 percent of global separation and processing capacity, according to the Center for Strategic and International Studies. It’s a gap that explains why mining announcements alone tell you almost nothing about future supply.

17 elements that chemistry barely tells apart

The reason separation is so hard is buried in the periodic table. The 15 lanthanides, plus the chemically similar scandium and yttrium, nearly all exist as trivalent ions whose radii differ by a matter of hundredths of a nanometer across the entire series — a quirk of their filled inner electron shells known as the lanthanide contraction. The Department of Energy’s own materials scientists describe the 17 rare earth elements as simply “chemically similar,” and note that purifying just one ton of a single element creates tons of acidic and radioactive waste.

Because ordinary chemistry barely distinguishes one lanthanide from its neighbor, separation plants rely on complex solvent-extraction circuits using kerosene and phosphonic acid extractants. Running through dozens to several hundred sequential stages, each step nudges one element slightly further from the rest. That’s why building a separation plant looks nothing like opening a mine. It demands specialized chemists, enormous processing infrastructure, and a place willing to accept the resulting waste stream.

Advertisement Advertisement

How China built a monopoly on the hard part

China’s grip on separation was not inevitable. The United States led the world in rare earths through the mid-1990s, and the government-funded Ames Laboratory, created out of Manhattan Project uranium-purification work, pioneered much of the era’s separation science in the 1950s, work that seeded American radar and laser technology, according to CSIS. Over the following three decades, though, China invested steadily in separation infrastructure and technical expertise while the West let its own capacity lapse, partly because Western environmental and worker-safety rules made handling the same radioactive and acidic waste streams far more expensive at home.

By 2023, China refined an estimated 99 percent of the world’s heavy rare earths, and in that same year it moved to ban exports of separation technology itself, not just the finished materials. The effort was aimed squarely at keeping rivals from ever catching up.

The handful of plants now trying to change that

A small cluster of projects is now trying to close that gap.

Advertisement Advertisement

MP Materials brought a light rare earth separation line online at Mountain Pass, exited 2025 with a run-rate near 4,000 metric tons a year of separated neodymium-praseodymium oxide. The company is racing to commission a heavy rare earth line there by mid-2026, backed by a Pentagon deal that made the Defense Department its largest shareholder and guaranteed it a $110-per-kilogram price floor.

Energy Fuels has converted part of its White Mesa uranium mill in Utah, the country’s only fully licensed conventional uranium mill, into a rare earth separation line. Officials expect commercial-scale heavy rare earth output by the end of 2026.

Lynas, based in Australia with a plant in Malaysia, became the first company outside China to produce commercial dysprosium and terbium oxide, in mid-2025, and is building a Pentagon-funded heavy rare earth facility in Texas.

Between them, according to industry analysis published in September 2026, Lynas and MP Materials still represent essentially the entire non-Chinese commercial oxide supply, at 7,260 and roughly 4,000 tonnes a year respectively.

Advertisement Advertisement

Why the gap won’t close this decade

Even a fully realized pipeline does not close the distance quickly. Trade analysts tracking the commissioning schedule, MP Materials’ heavy line in mid-2026, Energy Fuels’ by the end of 2026, an Australian refinery in 2027, project that combined ex-China separation capacity will stay well below 20 percent of global supply through 2028. Even so, it’s believed that China will keep refining most of the world’s rare earths well past that.

The International Energy Agency’s own 2026 outlook is not much more optimistic about the decade beyond: even by 2035, announced non-Chinese refining capacity is expected to reach only about two-thirds of projected mine output, and magnet-making capacity only about a third of that. The Pentagon set a public target of a complete “mine-to-magnet” supply chain by 2027. On the industry’s own commissioning calendar, that milestone looks less like a near-term achievement than the opening stage of a much longer rebuild.

This article was constructed with the assistance of artificial intelligence and published by a member of The Washington Times' AI News Desk team. The contents of this report are based solely on The Washington Times' original reporting, wire services, and/or other sources cited within the report. For more information, please read our AI policy or contact Steve Fink, Director of Artificial Intelligence, at sfink@washingtontimes.com

The Washington Times AI Ethics Newsroom Committee can be reached at aispotlight@washingtontimes.com.