BRENNEN

Industries / Mining and materials / Mining and quarrying / Rare earth processing

Metal making

Metal making turns separated rare earth oxides into metals and alloys, mostly by passing an electric current through a bath of molten salt, so that magnet makers and other users can melt them into products.

5

Milestones achieved, from 1875 to 2024.

2

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1875

    Cerium metal by electricity

    Chemists had failed to isolate pure cerium; even heating cerium chloride with potassium did not work. In 1875 two American chemists obtained a pure specimen by passing an electric current through molten cerium chloride, by the Royal Society of Chemistry's account. Electrolysis in molten salts, now usually rare earth oxide dissolved in molten fluorides, is still how many rare earth metals and alloys are made, by a SINTEF research paper.

    edu.rsc.orgzenodo.org

    Credit

    William Hillebrand and Thomas Norton

  2. 1903

    Sparks from cerium and iron

    Carl Auer von Welsbach found that an alloy of cerium and iron catches fire in air when shavings are scraped from it. He filed patents on the alloy in many countries in 1903, production started that year at the Treibacher Chemische Werke in Carinthia, Austria, and from 1907 he sold lighters to go with it, by the Austrian Patent Office's account. Flints of this "Auer metal" are still made in Treibach, and every flint-wheel lighter works on the same principle.

    patentamt.attreibacher.com

    Credit

    Carl Auer von Welsbach; Treibacher Chemische Werke (now Treibacher Industrie)

  3. 1953

    Pure heavy rare earth metals

    At the Ames Laboratory in Iowa, which had made uranium metal for the Manhattan Project under Spedding, A. H. Daane and Frank Spedding reported a way to make solid yttrium, terbium, dysprosium, holmium, erbium and thulium "in high purity and high yields": they reduced each metal's chloride or fluoride with calcium metal in tantalum containers under an inert gas. Ytterbium would not reduce fully this way. The lab's Materials Preparation Center still supplies high-purity rare earth metals to researchers and industry.

    osti.govosti.govameslab.govameslab.gov

    Credit

    A. H. Daane and Frank H. Spedding, Ames Laboratory

  4. 2018

    Smelter gases measured in China

    Perfluorocarbons (PFCs) such as tetrafluoromethane (CF4) are greenhouse gases that last for thousands of years; aluminum smelting releases them when a cell runs short of dissolved oxide. A team measured the exhaust of rare earth electrolysis cells at three Chinese companies for 24 to 60 hours each, while they made praseodymium-neodymium, lanthanum and dysprosium-iron. They found 26.66 to 109.43 g of CF4 per tonne of metal, plus C2F6 and traces of other PFCs, equal to 211.60 to 847.41 kg of CO2 per tonne: lower than aluminum electrolysis for most metals and similar for dysprosium-iron. The 2006 IPCC guidelines for national greenhouse gas inventories had no method for this source.

    LessonFluoride electrolysis gives off the same long-lived gases as aluminum smelting, and a source without an inventory method goes uncounted.

    ueaeprints.uea.ac.uk

    Credit

    Bofeng Cai and 13 co-authors (Resources, Conservation and Recycling, 2018)

  5. 2024

    Metal for GM in Texas

    MP Materials began commercially producing rare earth metal for General Motors in 2024 at its Independence plant in Fort Worth, Texas, from concentrate mined at Mountain Pass, California, Automotive News reported via GM Authority. An inventory of rare earth metal plants outside China by the analyst Jack Lifton, published in September 2026, lists the plant as operating, with commercial metal production since 2024 and no disclosed metal tonnage.

    gmauthority.cominvestornews.com

    Credit

    MP Materials

Watch for

  1. Korean dysprosium and terbium metal

    The Korean Metals Plant at Ochang, built by Australian Strategic Materials (ASM) and owned through ASM by Energy Fuels since August 2026, makes neodymium-praseodymium metal and neodymium-iron-boron alloy cast as thin strip for magnet makers. It is reported to have made its first commercial sales of dysprosium and terbium metals, the heavy rare earths that keep magnets working when hot. Energy Fuels gives 1,300 t a year of alloy capacity, but ASM reported only about 45 t of alloy and metal dispatched in the second quarter of 2026.

    Achieved whena magnet maker names the plant as the source of dysprosium or terbium metal in magnets it sells, or audited reports show sustained heavy rare earth metal output.

    investornews.com

    Claimed by

    ASM and Energy Fuels, as compiled by Jack Lifton, September 2026

  2. NdPr metal from Saskatoon

    REalloys, a Nasdaq-listed company, says it is fully funded from its own cash, having committed US$58.3 million to this project and a planned heavy rare earth metal plant in Utah, to upgrade the Saskatchewan Research Council's rare earth plant in Saskatoon, Canada, to 525 t a year of neodymium-praseodymium metal plus 30 t of dysprosium oxide and 15 t of terbium oxide. Work was due to start in the third quarter of 2026, and REalloys expects to take in commercial metal from the third quarter of 2027.

    Achieved whenthe plant reports metal shipped to an outside customer at a rate of hundreds of tonnes a year.

    miningweekly.com

    Claimed by

    REalloys, August 2026

Open

  1. 01

    Count metal made outside China

    A free quarterly table, from 2025 on, of the rare earth metal and alloy actually shipped by each plant outside China that makes them, such as Santoku and Shin-Etsu in Japan, the Korean Metals Plant, Less Common Metals in the UK, MP Materials in Texas and the Saskatchewan Research Council plant. Each figure is kept apart from nameplate capacity, linked to the filing or release it comes from, and labeled with its basis (metal, alloy or oxide equivalent).

    The brief
    Credit · unclaimed
  2. 02

    Settle the smelter gas gap

    A public test of the two published ways of sizing the CF4 and C2F6 that rare earth metal electrolysis gives off, against what the atmosphere shows. A laboratory estimate and a set of plant measurements differ several hundredfold; multiply each by yearly metal output for 2010 to 2025 and set both beside atmospheric estimates of China's CF4 emissions, to see which, if either, fits.

    The brief
    Credit · unclaimed

Open 01 · the brief

Count metal made outside China

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A free quarterly table, from 2025 on, of the rare earth metal and alloy actually shipped by each plant outside China that makes them, such as Santoku and Shin-Etsu in Japan, the Korean Metals Plant, Less Common Metals in the UK, MP Materials in Texas and the Saskatchewan Research Council plant. Each figure is kept apart from nameplate capacity, linked to the filing or release it comes from, and labeled with its basis (metal, alloy or oxide equivalent).

Why it is still open

Lifton's September 2026 inventory lists the plants and their claimed capacity but calls itself "a documented inventory, not a guaranteed census", says public disclosures are uneven, and gives one output figure, ASM's 45 t in a quarter, as "a useful reminder that nameplate capacity is not output". Searches for this brief found plant-level output only in paid market reports.

investornews.com

Built with what exists

Company quarterly reports and stock exchange filings (ASX, Nasdaq, NYSE, TSX); Python; an AI model built for pulling figures out of long documents, such as Muse Ember (see the Note "The jobs agents repeat"), with every number checked by a person against its page.

  1. First step

    Read the 2026 quarterly reports of Energy Fuels or ASM, USA Rare Earth (owner of Less Common Metals) and MP Materials, and record what each says about metal and alloy tonnage and on what basis.

  2. Done looks like

    Eight quarters, 2025 to 2026, for at least eight plants, each cell either a sourced figure or marked "not disclosed", published as a CSV file with a one-page comparison of shipped tonnage against claimed capacity.

Open 02 · the brief

Settle the smelter gas gap

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A public test of the two published ways of sizing the CF4 and C2F6 that rare earth metal electrolysis gives off, against what the atmosphere shows. A laboratory estimate and a set of plant measurements differ several hundredfold; multiply each by yearly metal output for 2010 to 2025 and set both beside atmospheric estimates of China's CF4 emissions, to see which, if either, fits.

Why it is still open

H. Vogel and B. Friedrich estimated in a 2018 conference paper, from laboratory measurements of about 74 g of CF4 per kilogram of metal and up to 35,000 t of metal a year, that rare earth electrolysis gives off about 25 million t of CO2 equivalent a year, or up to 10 million t in a medium case. The plant measurements of Cai and colleagues the same year found 26.66 to 109.43 g of CF4 per tonne of metal (see "Smelter gases measured in China"). A December 2025 preprint by Benjamin Püschel and 16 co-authors, using measurements from global monitoring networks, finds global CF4 emissions 2.6 ± 0.3 times what the EDGAR inventory reports; it names rare earth smelting as a source that "could contribute substantially" but calls current estimates "highly uncertain" and does not separate its share. A 2018 study of East Asian CF4 had already noted that rare earth production "may also release substantial amounts of CF4".

link.springer.comegusphere.copernicus.orgatmos-chem-phys.net

Built with what exists

The emission factors in the two 2018 papers; production figures from the USGS Mineral Commodity Summaries; the public NOAA and AGAGE CF4 records and the regional estimates in Püschel and colleagues' preprint; Python for a Monte Carlo estimate of uncertainty.

  1. First step

    Multiply each paper's factor by Vogel and Friedrich's 35,000 t a year of metal and compare both results with China's CF4 total in the preprint.

  2. Done looks like

    Yearly totals for 2010 to 2025 under each set of factors, with 90% ranges and every input traced to a source, and a stated finding on which set is consistent with the atmospheric estimates for China.