Washington is funding a range of homegrown separation technologies rather than copying Chinese methods. The bet is that innovation closes the gap faster than imitation.
The United States does not know how to refine rare earths at scale. China does. That gap is not a matter of scientific ignorance. It is a matter of manufacturing know-how accumulated over forty years, embedded in thousands of Chinese technicians, hundreds of specialized facilities, and an export-control regime that Beijing has been steadily tightening since 2023.
Washington’s answer is not to replicate the Chinese process. It is to fund a wide field of alternative separation technologies and back enough of them simultaneously that at least some succeed. The strategy is deliberate. It carries real risk. And the timeline to closing the gap runs through the end of this decade at the earliest.
Why the Gap Exists
China’s dominance in rare earth refining was not accidental and it will not be quickly reversed.
The standard Chinese method involves dissolving ore in strong acid, mixing the solution with kerosene and chemical extractants such as P507 and P204, and passing it through hundreds of mixing and settling tanks. A stronger acid then strips out individual elements for precipitation. The process requires up to 200 cycles of solvent extraction to achieve commercial purity for certain elements. Chinese firms have spent more than two decades refining every variable in that sequence.
In December 2023, Beijing banned exports of rare earth processing and magnet-making technologies. That move foreclosed the option of purchasing Chinese equipment or hiring Chinese engineers to transfer the process. In June 2026, China cut off supplies of dual-use items to US producers MP Materials and USA Rare Earth. Last month, it tightened exit controls to stop skilled technicians — including rare earth specialists — from leaving the country.
The effect is that Western firms seeking to master solvent extraction at commercial scale must pay the full cost of learning what China learned by doing, without access to the people or equipment that made the Chinese process as efficient as it is.
Audrey Robertson, assistant secretary of energy at the Department of Energy’s Office of Critical Minerals and Energy Innovation, stated the position plainly at the Financial Times Metals and Mining Summit in London on October 15. The US developed intellectual property over the past twenty years, she said, but did not develop the manufacturing know-how. The country built ideas but not factories. That is the gap it now faces.
More Read
The Three Paths US Refiners Are Taking
Rather than directing investment toward a single method, Washington is backing three distinct separation technologies. Each addresses the know-how problem differently.
Solvent extraction remains the baseline approach. MP Materials, NioCorp’s Elk Creek project in Nebraska, and Aclara Resources are all pursuing it. The method is proven and commercially understood. The challenge for US operators is sourcing the chemical extractants — P507 and P204 — that Chinese producers have used for decades. Whether to source those chemicals from China or pay a premium for US and European substitutes is an unresolved cost question that will affect project economics across this group.
Continuous ion chromatography, or CIC, is the technology US refiners dropped in the 1950s because it was too expensive when rare earths were cheap. The method dissolves ore in acid, dilutes the solution with water, and runs it through resin-filled columns where each element moves at its own speed and separates naturally. ReElement Technologies, USA Rare Earth, and the Colorado School of Mines’ new demonstration facility are all pursuing this route. The DOE selected the Colorado School of Mines on June 2 for a share of $134 million in funding to build a rare earth demonstration facility processing mining and industrial waste into rare earth metals. With rare earths now priced at crisis levels in some markets, the economics that made chromatography uncompetitive in 1950 no longer apply.
Electrochemical separation is the least commercially proven of the three approaches. Rare earths are dissolved in molten salt at approximately 700 degrees Celsius and an electric current is applied. Each element converts to metal at a slightly different voltage and can be collected separately. Phoenix Tailings is the primary US developer pursuing this route at commercial scale.
The Projects Moving to Construction
Two facilities illustrate how the US refining strategy is moving from laboratory to construction.
On October 21, Phoenix Tailings announced Freedom Separation: a $375 million rare earth separation facility to be built in Rochester, New York. At design capacity, the plant will process more than 18,000 tonnes of feedstock annually and produce over 4,400 tonnes of separated rare earth oxides. First production is targeted for 2028. Phoenix Tailings secured a $500 million conditional loan from the Pentagon and a $66 million DOE grant in June 2026.
The second facility is Aclara Resources’ planned heavy rare earth separation complex at the Port of Vinton in Louisiana. Announced on October 13, the project carries an estimated price of $494 million. At capacity, it will produce approximately 1,350 tonnes of yttrium oxide per year, alongside neodymium-praseodymium, dysprosium, terbium, gadolinium and samarium oxides. The expected production window is the second half of 2028. Aclara will use large-scale solvent extraction, processing mixed rare earth carbonate shipped from its projects in Brazil and Chile. The extractant sourcing question — Chinese-origin chemicals or premium substitutes — remains publicly unresolved.
The DOE also invested in workforce capacity. On August 7, it launched the $100 million PROSPECT initiative, aimed at doubling the number of mining and minerals graduates in the US within two years. Robertson identified workforce shortage as a direct constraint on the sector’s ability to execute. A facility that can be financed and permitted still requires engineers, process chemists and plant operators who know how it works.
Japan’s Yttrium Problem Sharpens the Stakes
The consequences of supply concentration are already visible in the yttrium market.
After Japanese Prime Minister Sanae Takaichi said last November that a Chinese attack on Taiwan could pose an existential threat to Japan, Beijing responded on January 6 by banning exports of dual-use items — including certain rare earths — to Japanese military users. Yttrium exports from China have since stalled globally.
Yttrium coats jet engines and chipmaking equipment. US defense supply chains depend on Japanese components that contain it. The Financial Times reported on October 21 that Japan could run out of yttrium within months and had asked Washington to help press Beijing to ease its curbs. According to Argus Media, yttrium oxide prices outside China have increased 179-fold in two years to $1,150 per kilogram, compared with $9 per kilogram inside China.
Japan’s industry ministry confirmed the supply disruption publicly this week. Aclara’s Louisiana facility, if it delivers to schedule, would produce 1,350 tonnes of yttrium oxide annually from 2028 — a meaningful contribution to a market currently experiencing near-total Chinese supply closure. That production date is still two years away.
What the Industry Should Watch
Phoenix Tailings’ construction execution on Freedom Separation. Electrochemical separation has not been demonstrated at the scale Freedom Separation requires. The gap between laboratory validation and 4,400 tonnes per year of commercial output is where novel processes historically fail. The rate of construction progress through 2027 will indicate whether the technology is tracking toward its 2028 production target.
Aclara’s extractant sourcing decision. Whether the Louisiana facility uses Chinese-origin P507 and P204 or more expensive alternatives will affect its operating economics and its exposure to further Chinese export restrictions. That decision has not been publicly disclosed and will matter to anyone assessing the project’s supply-chain resilience.
The CIC demonstration facilities’ output. The Colorado School of Mines facility funded in June 2026 is a demonstration plant, not a commercial operation. Its purpose is to prove that CIC works at a scale that justifies commercial investment. Results from the demonstration phase will set the credibility of CIC as a route to commercial rare earth separation in the US.
China’s further export restriction moves. Beijing has tightened its rare earth export controls in stages: the December 2023 technology ban, the June 2026 dual-use item cutoff, and last month’s technician exit restrictions. Each step closed off another avenue through which Western refiners might have closed the know-how gap more quickly. Further restrictions — on specific chemicals, additional elements, or additional country targets — remain a live possibility and would directly affect project economics across all three technology paths.
The workforce ramp. Robertson’s PROSPECT initiative targets doubled graduate output within two years. The bottleneck is not only financial or technological. It is human. The US will need process engineers and plant operators who understand rare earth chemistry at industrial scale. Whether the education pipeline produces them on the timeline the refinery construction schedule requires is a real variable that will constrain how quickly any new facility reaches nameplate capacity.
Where the Strategy Stands
Michel Van Hoey, a senior partner at McKinsey, assessed the challenge directly at the London summit. The rare earth and magnet value chain represents process capability the West transferred to China and no longer possesses. China has continued to advance that capability over forty years. Closing the gap requires genuine reinvention, not incremental improvement.
The US is pursuing that reinvention across three technology tracks simultaneously, with government capital deployed across DOE grants, Pentagon loans, and the newly established Office of Critical Minerals and Energy Innovation. The total announced support across facilities, demonstration projects and workforce programs runs into the billions. The strategy accepts that not every approach will succeed. It bets that enough will.
The first commercial facilities targeting this outcome are now in announced or funded stages for 2028 delivery. Between the announcement and the commissioned plant, two years of construction, technology validation and supply-chain development remain. That distance is where the outcome will be decided.
Sources:
- Jeff Pao, “US reinvents rare earth refining as China guards know-how” — Asia Times, October 23, 2026
- DOE Office of Critical Minerals and Energy Innovation — energy.gov/cmei
- “Phoenix Tailings to build $375M rare earth plant in New York” — Mining.com
- “Aclara jumps on US REE plant plan” — Northern Miner
- “Digging Deep: Inside the Race for Domestic Critical Minerals” — Foley Hoag, October 2026
- ““China bans export of rare earth processing technologies” — The Register, December 22, 2023
