Rare Earth Elements

The 17-Element Basket: Why “Rare Earths” Are One Market With 17 Prices

This deep-dive treats rare earths as an aggregate basket, distinct from the standalone USGS chapters on heavy rare earths, scandium, and yttrium that this hub covers separately. The lanthanide series runs from lanthanum (atomic number 57) through lutetium (71); adding scandium (21) and yttrium (39) — both chemically similar to the lanthanides because of comparable ionic radii and a shared preference for the trivalent (3+) oxidation state — gives the conventional 17-element group, sometimes reduced to 16 in commercial discussion because promethium (61) is virtually absent in nature, being radioactive with a short half-life (rare-earth element reference summary; USGS Fact Sheet 2014-3078).

1. Light, medium, and heavy: the three-tier classification

USGS's operating convention, used across its own commodity chapters, divides light rare earths (LREE) as the lanthanide elements from atomic number 57 (La) through 64 (Gd), and heavy rare earths (HREE) as atomic numbers 65 (Tb) through 71 (Lu), with yttrium grouped into the heavy category despite its low atomic number because of its similar ionic radius and geochemical behavior (USGS 2020 Minerals Yearbook, rare earths). A three-way split used elsewhere separates a middle band of medium/intermediate REEs (MREE) — samarium through holmium in some schemes, or specifically europium and gadolinium in USGS's own heavy-REE chapter framing — from light (La–Sm/Nd) and heavy (Tb–Lu plus Y) groups (USGS MCS 2026, Rare Earths (Heavy); Geological Society of London briefing note). The dividing line varies by source: some class La through Sm as light and Eu through Lu plus Y as heavy; the U.S. Congressional Research Service's 2026 summary places La, Ce, Pr, Nd, Pm, Sm, Eu, and Gd in the light group and Tb, Dy, Ho, Er, Tm, Yb, Lu, Sc, and Y in the heavy group (Congressional Research Service, IF13171, updated June 2026).

2. The 4f electron shell: the chemistry that makes separation hard

The scientific basis for the light/heavy divide is electron configuration in the inner 4f shell. USGS explains the split is grounded in the fact that “the LREEs [have] unpaired electrons in the 4f electron shell and HREEs [have] paired electrons in the 4f electron shell” (USGS 2020 Minerals Yearbook, rare earths). Because all 17 elements share the same +3 valence state and near-identical ionic radii — the lanthanide contraction shrinks ionic radius smoothly and predictably as atomic number rises — they substitute for one another in crystal lattices and co-precipitate together in nature rather than forming separate, element-specific minerals. This “geochemical coherence” is why no economic ore body yields a single rare earth in isolation: every mined ton of rare-earth ore contains a full, non-selective mixture of the basket, in proportions set by the host mineral and deposit type, not by market demand (Virginia Department of Energy, Geology of Rare Earth Elements).

3. Abundance is not the constraint — concentration and separation are

Despite the name, most rare earths are not rare in the Earth's crust: cerium, at roughly 43–66 parts per million depending on the estimate, is more abundant than copper (about 27–28 ppm), and even the scarcest naturally occurring REE, thulium, at roughly 0.28–0.3 ppm, is still far more abundant than silver (0.05–0.08 ppm) (Geological Society of London briefing note; Virginia Department of Energy, Geology of Rare Earth Elements). What is genuinely rare is a mineral deposit where rare earths occur in sufficiently high, economically minable concentrations — typically requiring magmas that have undergone heavy fractionation, such as carbonatites, alkaline igneous complexes, and certain weathered clay horizons (Virginia Department of Energy, Geology of Rare Earth Elements). This is the structural reason the industry's real bottleneck sits downstream of mining, in chemical separation (Section 3), rather than in the mining step itself.

4. The host minerals: bastnaesite, monazite, xenotime, and ionic clays

Four ore types supply essentially all commercial rare earths. Bastnaesite ((Ce,La)(CO₃)F) is light-REE-dominant and the principal ore at both Bayan Obo, China, and Mountain Pass, California, containing roughly 74.8% theoretical REE-oxide content by weight (Virginia Department of Energy, Geology of Rare Earth Elements). Monazite ((Ce,La,Nd,Th)PO₄) also concentrates the light REEs but carries a substantially higher heavy-REE fraction (roughly two to three times that of bastnaesite) and, critically, co-hosts up to 30% thorium by weight, making it both a heavy-mineral-sand byproduct source and a radioactive-waste-management challenge (Oxford Academic, “What Are Rare Earths?”). Xenotime (YPO₄) is the most heavy-REE-enriched of the major ores, at roughly 61.4% theoretical REE-oxide content, but is a minor, generally byproduct source globally (Virginia Department of Energy, Geology of Rare Earth Elements). Ionic-adsorption clays — found overwhelmingly in southern China's Jiangxi, Guangdong, Fujian, and Hunan provinces — are weathered laterite deposits where rare-earth ions are loosely adsorbed onto clay-mineral surfaces rather than locked in a crystal lattice, permitting low-cost in-situ leaching with ammonium sulfate solutions and yielding an unusually heavy-REE-rich product; these deposits are the world's dominant source of commercial dysprosium and terbium (IEA Global Critical Minerals Outlook 2025 — China processing dominance; Nature Communications, origin of heavy rare earth mineralization in South China, 2017).

Global Supply Concentration: China Mines 69%, Refines 85–90%, and Makes 90%+ of Magnets

World mine production and reserves by country, 2024–2025

Source: USGS MCS 2026. Brazil holds the world's second-largest reserves at 11 million tonnes, and Australia's 36.3 million tonnes is the third largest, though Australia notes only 3.3 million tonnes are JORC-compliant reserves under stricter Australian reporting standards (USGS Mineral Commodity Summaries 2025 — Rare Earths). Canada holds no current commercial production but reports over 830,000 tonnes of reserves and more than 14 million tonnes of measured-and-indicated resources (USGS MCS 2026).

China's three-region mining base: Bayan Obo, Sichuan/Shandong, and the southern ionic clays

China's domestic output is not homogeneous; it comes from three geologically distinct production centers. Bayan Obo, in Inner Mongolia, is the world's single largest rare-earth deposit, a bastnaesite/monazite iron-ore-associated body dominated by light REEs and mined as a byproduct of iron ore by Baotou Steel (Baogang) (Rare Earth Mining News, China Rare Earth Mining: Key Deposits & Policy, 2026). Sichuan province's Maoniuping and Mianning deposits form a second major light-REE bastnaesite source, while Shandong hosts additional light-REE resources feeding China's eastern refining base (Rare Earth Mining News, 2026). Southern China — principally Jiangxi, with additional deposits in Guangdong, Fujian, Hunan, and Yunnan — hosts the ionic-adsorption clay deposits that are the world's dominant source of heavy rare earths, especially dysprosium and terbium, exploited through in-situ ammonium-sulfate leaching (Rare Earth Mining News, 2026; Nature Communications, 2017). China's Ministry of Industry and Information Technology (MIIT) regulates output nationally through a biannual mining-quota system: the 2025 total mining quota was set at 270,000 tonnes REO and the smelting/separation quota at 255,000 tonnes, both year-on-year increases, allocated primarily across six state-owned enterprise groups with a small allowance for approved private operators (Rare Earth Mining News, 2026).

Myanmar: the informal, unverified heavy-REE swing supplier into China

Burma (Myanmar) is USGS's own reported second- or third-largest producer at 22,000–27,000 tonnes REO in 2024–2025, but this figure is estimated indirectly from Chinese import data because Myanmar's own mining sector lacks reliable government reporting; USGS notes world totals exclude some informal, unverified Burmese output that independent trackers estimate could add a further 220,000–230,000 tonnes if fully counted (USGS MCS 2026; REEtracker, Rare Earth Mine Production 1950–2025). Myanmar's Kachin State ionic-clay mining, much of it in militia-controlled territory, feeds directly into China's heavy-REE separation plants across the border, making Myanmar effectively an extension of the Chinese heavy-REE supply chain rather than an independent source.

Downstream concentration is worse than mining concentration

Mining concentration alone understates China's leverage. Separately, refining (chemical separation into individual oxides, Section 3) is estimated by industry analysts at roughly 85–90% Chinese share, and NdFeB permanent-magnet manufacturing capacity is estimated above 90% Chinese share, meaning China's effective control of the value chain is materially higher than its 69% mining share implies (Rare Earth Exchanges, “Rare Earth Separation Is the Real Chokepoint,” 2026; IEA Global Critical Minerals Outlook 2025 — refining concentration). The U.S. is 67% net import reliant for most REEs and 100% net import reliant specifically for scandium and yttrium as of 2025 (Congressional Research Service, IF13171).

The Real Chokepoint: Solvent-Extraction Separation, Not Mining

1. Why separation, not mining, is the bottleneck

Because all 17 elements share the same +3 valence and near-identical ionic radii, they cannot be separated by the physical or single-stage chemical methods used for most other metals. Industrial separation relies on countercurrent solvent extraction: an aqueous feed solution of dissolved rare-earth ions is repeatedly contacted with an organic extractant (commonly di-2-ethylhexyl phosphoric acid, or the P507/HEHEHP/PC-88A family for NdPr purification and heavy-REE work) across dozens to hundreds of mixer-settler stages, exploiting the minute differences in extraction equilibrium between adjacent elements to progressively concentrate each one (Hydrometallurgy/ScienceDirect, critical review of REE solvent extraction; Rare Earth Exchanges, 2026). Heavy REEs such as terbium, dysprosium, holmium, erbium, ytterbium, lutetium, and yttrium have even more similar chemical properties to their immediate neighbors than light REEs do, requiring more extraction stages, higher reagent consumption, and materially greater process precision, which is why heavy-REE separation capacity is scarcer and more concentrated than light-REE separation capacity worldwide (Rare Earth Exchanges, 2026).

2. China's separation dominance: an estimated 90% light, 98% heavy

Industry analysis estimates China controls roughly 90% of global light rare-earth separation capacity and as much as 98% of heavy rare-earth separation capacity, a concentration built over four decades of process refinement, reagent-supply-chain development, and engineering-talent accumulation that is difficult to replicate quickly (Rare Earth Exchanges, 2026). A separate industry estimate puts overall (light-plus-heavy blended) non-Chinese refining concentration at roughly 86% Chinese share as of mid-2026, with capital increasingly flowing toward non-China processing projects in response (IEA Global Critical Minerals Outlook 2025 — refining concentration).

3. Lynas: the only non-Chinese company operating full commercial separation

Lynas Rare Earths' Malaysian facility (Section 4) is the sole non-Chinese operation running a complete, commercial-scale cracking-leaching-and-solvent-extraction circuit for both light and, since 2024, heavy rare earths at meaningful volumes; Lynas produced 6,558 tonnes of NdPr in fiscal 2025, up from prior-year levels, and posted record quarterly NdPr output above 2,000 tonnes for the first time in the June 2025 quarter (Yahoo Finance, MP/Lynas production comparison, Sep 2025; Lynas Rare Earths, Quarterly Report, period ended 30 June 2025). MP Materials and Energy Fuels are only now bringing NdPr and heavy-REE separation online at much smaller scale (Section 4), meaning the West effectively has one and a half commercial separation operations against China's dozens.

4. China's technology export controls now target the separation process itself

Recognizing that separation know-how, not raw ore access, is its structural advantage, China's October 2025 controls explicitly extended beyond metals and compounds to equipment, reagents, and technical know-how for rare-earth mining, smelting and separation, metal smelting, magnetic-material manufacturing, and recycling, per MOFCOM Announcement No. 62 of 2025, covering “production line assembly, debugging, maintenance, repair and upgrade” technology (White & Case, 13 Oct 2025; Taylor Wessing, 9 Oct 2025). Separately, Announcement No. 56 of 2025 extended controls to “machinery, auxiliary chemicals and raw materials used in rare-earth separation and metallurgical refining” (Taylor Wessing, 9 Oct 2025). Analysts note this forces Western developers to build independent separation know-how from scratch rather than licensing it, extending the realistic timeline for non-Chinese heavy-REE self-sufficiency (IEA Global Critical Minerals Outlook 2025 — China processing dominance).

Building an Alternative: MP Materials, Lynas, Energy Fuels, and the Rest of the Non-China Supply Chain

1. MP Materials: Mountain Pass to Fort Worth, an integrated U.S. mine-to-magnet chain

MP Materials' Mountain Pass, California mine — America's only rare-earth mine and processing operation of scale — produced a record 50,692 metric tons of rare-earth oxides in concentrate in 2025, up 12% year-on-year, and a record 2,599 metric tons of separated NdPr oxide, up 101% year-on-year, reflecting the ramp of newly commissioned separation circuits (MP Materials, Q4/FY2025 results, 26 Feb 2026). Momentum accelerated into 2026: Q1 2026 NdPr production hit a record 917 metric tons, up 63% year-on-year, with sales up 117% to 1,006 metric tons, crossing the 1,000-tonne quarterly threshold for the first time (MP Materials, SEC filing, Q1 2026 results). In July 2025 MP Materials signed a Department of Defense partnership anchored by a $110/kg NdPr price floor and simultaneously ceased all rare-earth-oxide sales to China, redirecting Mountain Pass output entirely into the U.S. supply chain (Rare Earth Mining News, market outlook, Mar 2026). Downstream, its Independence facility in Fort Worth, Texas produced its first commercial-scale NdFeB magnets in December 2025, supplying General Motors, with a second, larger 10X facility in Northlake, Texas scheduled to break ground in 2026, backed by more than $200 million in state and local incentives; combined, Independence and 10X target roughly 10,000 metric tons per year of U.S. magnet output (Yahoo Finance, MP Materials 2025 production records, Mar 2026; Rare Earth Mining News, Mar 2026).

2. Lynas Rare Earths: Mt Weld ore, Kalgoorlie concentration, Kuantan separation

Lynas operates the world's largest rare-earth mining and processing business outside China, structured across three sites. Ore is mined at Mt Weld near Laverton, Western Australia — a bastnaesite-rich carbonatite deposit with an Ore Reserve of 18.9 million tonnes at 8.3% total rare-earth oxide grade and a Mineral Resource of 55.2 million tonnes at 5.3% TREO, supporting a mine life exceeding 35 years at current NdPr capacity or 20+ years at the expanded 12,000-tonne-per-annum NdPr target (EPA Western Australia, Mt Weld FAQ; Lynas Rare Earths, Mt Weld project page). Concentrate is processed at a cracking facility in Kalgoorlie, Western Australia, before final separation at the Lynas Advanced Materials Plant (LAMP) in Kuantan, Malaysia, the largest rare-earth separation plant outside China (Lynas Rare Earths, Kalgoorlie project page). Lynas's NdPr production reached 3,407 tonnes in the most recently reported half-year period, up 15% year-on-year, with average selling price rising sharply from A$44.6/kg in the first half of FY25 to A$68.4/kg in the first half of FY26 — and A$85.6/kg in the December 2025 quarter alone — a move Lynas management attributed directly to MP Materials' DoD-backed $110/kg NdPr floor price “reshaping the market” (Rare Earth Mining News, Lynas operations profile, Apr 2026).

3. Lynas's U.S. and Malaysian expansion: Kalama and a second Malaysian facility

Under a 2021 Defense Production Act Title III technology investment agreement, Lynas is establishing heavy-rare-earth separation capability in Seadrift, Texas (widely referenced by the associated town of Kalama-linked planning in some coverage, though the DoD-funded U.S. HREE facility is sited in Texas), with a subsequent $120 million DoD contract awarded in 2022 to build a commercial heavy-rare-earths separation facility in the United States (U.S. Department of War, DPA Title III award to Lynas, 1 Feb 2021; Mining.com, Lynas $120m DoD contract, 13 Jun 2022). Separately, in Malaysia, Lynas announced plans in October 2025 to build a new rare-earth facility to add capacity alongside the existing Kuantan LAMP site (MarketScreener, Lynas Malaysia expansion, 28 Oct 2025). A flotation-circuit expansion at Mt Weld reached 70% of nameplate capacity in the first half of fiscal 2026, part of the broader push toward the 12,000-tonne NdPr target (Rare Earth Mining News, Apr 2026).

4. Energy Fuels: monazite byproduct integration from uranium milling

Energy Fuels processes monazite — a mineral that contains both rare earths and uranium — at its White Mesa Mill in Blanding, Utah, the only facility in the U.S. combining uranium milling with rare-earth separation. Its Phase 1 NdPr separation circuit has capacity for up to 1,000 metric tons per year, with Phase 2/3 expansion targeting 4,000–6,000 tonnes and adding dysprosium and terbium production (Energy Fuels, Rare Earths & Monazite). Feedstock security is being built through the Donald Project joint venture in Victoria, Australia (with Astron Corporation), a monazite-and-xenotime heavy-mineral-sand deposit expected to supply 7,000–8,000 tonnes per year of rare-earth-element concentrate in its first phase, rising to 13,000–14,000 tonnes in phase two, plus the Toliara project in Madagascar (via the Base Resources acquisition) and the Bahia project in Brazil (Mine Australia, Issue 57, Jul 2025; PR Newswire, Energy Fuels/Base Resources acquisition, 21 Apr 2024). A $725 million, 20-year conditional loan commitment from the U.S. Office of Strategic Capital is funding White Mesa Mill's expansion (Investing.com, Energy Fuels rare earth magnet strategy, 25 Jun 2026).

5. Vietnam's Dong Pao and India's IREL: state-controlled, underdeveloped resources

Vietnam's Dong Pao mine in Lai Chau province is Southeast Asia's largest known rare-earth deposit, but Vietnamese REO output has fallen sharply, from an estimated 300 tonnes in 2024 to just 150 tonnes in 2025, despite the country holding 3.5 million tonnes of reserves — the world's fifth-largest — reflecting stalled development and licensing disputes rather than resource scarcity (USGS MCS 2026; Wikipedia, Dong Pao mine). In India, state-owned IREL (India) Limited mines monazite-bearing beach sands and is the country's sole integrated rare-earth producer, with 2025 output around 2,900 tonnes REO; in mid-2025 the Indian government reportedly asked IREL to suspend a 13-year-old rare-earth export agreement with Japan to prioritize domestic supply amid China-related supply concerns (Wikipedia, Indian Rare Earths / IREL; USGS MCS 2026).

China's 2025 Export-Control Escalation: April's Seven Elements, October's Extraterritorial Reach, and the November Truce

1. The 4 April 2025 announcement: seven elements, dual-use rationale

China's Ministry of Commerce (MOFCOM) and General Administration of Customs (GAC) issued Announcement No. 18 of 2025 on 4 April 2025, imposing export-licensing controls on samarium, gadolinium, terbium, dysprosium, lutetium, scandium, and yttrium and their metals, alloys, oxides, compounds, and specified permanent-magnet materials, citing China's Export Control Law and “international nonproliferation obligations” (Holland & Knight, 4 Apr 2025; MOFCOM Announcement, English translation). A MOFCOM spokesperson framed the measure explicitly around the seven “medium and heavy rare earth-related items” and cited the Export Control Law and related regulations as the legal basis (MOFCOM Spokesperson's Remarks, 10 Apr 2025). Crucially, this was not an outright ban: exporters must obtain case-by-case licenses, and the Center for Strategic and International Studies noted the measure “institutionalizes discretion” — Beijing can throttle supply simply by slowing or denying license approvals without ever announcing a formal embargo (CSIS, 14 Apr 2025). The controls included a de minimis rule requiring a Chinese export license for any foreign product containing 0.1% or more Chinese-origin controlled rare-earth content by value (Rare Earth Exchanges, 30 Oct 2025).

2. October 2025: extraterritorial jurisdiction and five more elements

On 9 October 2025, MOFCOM issued Notice No. 61 of 2025 and companion Announcements No. 55–58, expanding controls to five additional rare earths — europium, holmium, erbium, thulium, and ytterbium — bringing the controlled list to 12 of the 17 REEs, effective 8 November 2025 (China Briefing, 10 Nov 2025). For the first time, MOFCOM asserted extraterritorial jurisdiction: foreign entities exporting rare-earth permanent-magnet materials or sputtering targets manufactured outside China from one non-China country to another now require a MOFCOM license if the product contains 0.1% or more Chinese-origin controlled rare-earth content by value, or was produced using specified Chinese rare-earth processing technology (White & Case, 13 Oct 2025; CSET/Georgetown, MOFCOM Notice 2025 No. 61, English translation). Companion measures (Announcements No. 55–58) simultaneously extended controls to superhard materials, rare-earth processing equipment and raw materials, and lithium-battery anode materials — a broadening well beyond rare earths alone (Taylor Wessing, 9 Oct 2025).

3. November 2025: a one-year suspension, not a repeal

Following diplomatic engagement ahead of a Trump-Xi meeting, China suspended the October expansion — the five additional elements and the broadest extraterritorial provisions — for one year in November 2025, while explicitly leaving the April 2025 seven-element licensing regime fully in force (Rare Earth Exchanges, 30 Oct 2025; USGS MCS 2026, Rare Earths (Heavy)). USGS's own February 2026 rare-earths chapter confirms this sequencing precisely: “In April 2025, China tightened its export controls… In October, China expanded its rare-earths export controls… In November, China suspended the October export controls for 1 year. The April export controls remained in effect, although China began to issue general export licenses to selected exporters” (USGS MCS 2026). Analysts characterize this as a rollback to the April baseline rather than genuine relief: “Beijing retains the legal and regulatory tools to throttle rare earth exports at will… simply by delaying or denying license approvals” (Rare Earth Exchanges, 30 Oct 2025).

4. Why neodymium itself was never controlled — but NdFeB magnets still are

Notably, neither the April nor the October controls placed neodymium or praseodymium themselves under license — the controls target the seven (later twelve) medium/heavy elements plus specific finished magnet products. However, the controlled product list explicitly includes “NdFeB permanent magnet materials containing terbium” and “NdFeB permanent magnet materials containing dysprosium,” meaning the vast majority of commercial high-performance neodymium magnets — which are routinely doped with dysprosium and/or terbium to raise their operating temperature — fall under the licensing regime even though neodymium is uncontrolled (China Briefing, 10 Nov 2025). This is the mechanism by which China's control of seven elements translates into effective control of the global magnet trade.

Prices & Benchmarks: 17 Elements, 17 Curves, One Dominant Reference — NdPr FOB China

Source: USGS MCS 2026, Argus Media, Argus Non-Ferrous Markets assessments, free-on-board basis.

1. NdPr FOB China: the dominant price reference

Because neodymium and praseodymium together drive roughly 90% of rare-earth-magnet value (Section 7), NdPr oxide and NdPr metal FOB-China assessments — published by Fastmarkets, Argus Media, and Asian Metal — function as the industry's de facto benchmark, the closest thing rare earths have to an LME-style reference price (Fastmarkets, NdPr oxide 99% FOB China assessment; Fastmarkets, NdPr metal FOB China assessment). There is no futures contract or exchange-cleared instrument for any rare earth; all pricing is physical, assessment-based, and dominated by Chinese domestic transaction data because China remains the largest producer and consumer.

2. The MP Materials $110/kg floor price: a structural break from the FOB-China reference

MP Materials' July 2025 Department of Defense agreement introduced a $110/kg NdPr price floor for its U.S. production — a price-support mechanism explicitly designed to insulate a domestic producer from Chinese price volatility. Lynas management directly credited this floor with “reshaping the market,” citing the doubling of its own realized NdPr price from A$44.6/kg to A$85.6/kg within roughly two quarters (Rare Earth Mining News, Lynas profile, Apr 2026). This marks the emergence of a bifurcated pricing structure: a Chinese FOB benchmark that continues to set the global marginal price, and a growing non-China premium tier anchored by the U.S. government-backed floor.

3. Dysprosium-terbium and heavy/medium REO baskets: thin, opaque, and China-set

Unlike NdPr, dysprosium and terbium have no comparably liquid published benchmark; pricing for Dy and Tb oxide, and for broader medium-rare-earth-oxide (MREO) and heavy-rare-earth-oxide (HREO) baskets, is thinner, less transparent, and effectively set by Chinese domestic transaction levels because nearly all heavy-REE separation capacity sits in China (Section 3). Squatex's 2026 compilation places dysprosium around $452/kg and terbium around $1,720/kg, a roughly 25× and 6.5× premium to NdPr respectively (Squatex, 2026).

4. China's rare-earth price index and 2026 momentum

China maintains its own composite rare-earth price index, which analysts reported hitting 266 in mid-2026 — interpreted by some as a genuine demand signal and by others as policy-driven signaling ahead of quota announcements — with NdPr oxide assessed at ¥733.9–753.9/kg (roughly $110–113/kg) as of early July 2026, notably converging toward the MP Materials floor price (Rare Earth Exchanges, 1 Jul 2026).

End Uses: Magnets Are ~48% of Volume but the Dominant Value and Strategic Driver

1. Magnets: NdFeB dominance and the Dy/Tb high-temperature dopants

Neodymium-iron-boron (NdFeB) permanent magnets are the largest and fastest-growing rare-earth end use, consumed in electric-vehicle traction motors, wind-turbine direct-drive generators, and a vast range of consumer electronics and industrial motors. USGS identifies magnets as “the estimated leading global use” for rare earths overall, even though catalysts, not magnets, is the leading domestic U.S. end use, reflecting the United States' still-nascent magnet manufacturing base relative to its refining capacity (USGS MCS 2026). Dysprosium and terbium are added in small percentages to NdFeB magnets specifically to raise their Curie temperature and coercivity, allowing them to retain magnetic performance at the high operating temperatures found in EV motors and wind generators — this is precisely why China's export controls target dysprosium- and terbium-containing magnets even while leaving neodymium itself uncontrolled (Section 5).

2. Catalysts: the largest U.S. domestic use, driven by petroleum refining

Cerium- and lanthanum-based catalysts are used heavily in fluid catalytic cracking (FCC) at petroleum refineries and in vehicle catalytic converters, and USGS reports this as the leading domestic end use of rare earths within the United States (USGS MCS 2026). Squatex's 2026 global demand breakdown places catalysts at roughly 15.7% of global rare-earth consumption by volume, the second-largest category after magnets (Squatex, 2026).

3. Polishing, glass, and ceramics: mature, cerium/lanthanum-dominant markets

Cerium oxide is the workhorse polishing compound for glass, optics, and semiconductor-wafer finishing, representing about 10.5% of global rare-earth demand (Squatex, 2026). USGS separately notes the glass sector uses lanthanum, cerium, and erbium (in descending order of consumption) for polishing, colorizing, and UV-absorbing glass applications, while other REEs see smaller-volume use across ceramics and glazes (USGS 2020 Minerals Yearbook, rare earths).

4. Phosphors, batteries, and high-value niche uses

Phosphors — combinations of lanthanum, cerium, terbium, yttrium, and europium oxides — convert ultraviolet radiation into visible light in fluorescent lighting and were historically a larger driver of heavy-REE demand before LED lighting displaced fluorescent technology, and now represent a smaller, roughly single-digit-percent share of global demand (Oxford Academic, “What Are Rare Earths?”). Nickel-metal-hydride and other rare-earth-containing battery chemistries, along with metallurgical alloys such as mischmetal (a cerium-lanthanum alloy used in lighter flints and steel/alloy additives), round out USGS's other-use categories (USGS MCS 2026). High-unit-value niche uses concentrate the heavy REEs: laser crystals commonly use neodymium- and yttrium-based hosts doped with dysprosium, erbium, thulium, or ytterbium, and lutetium is used in positron-emission-tomography (PET) medical-imaging detector crystals (USGS 2020 Minerals Yearbook, rare earths). The U.S. Department of Energy identifies dysprosium, neodymium, terbium, europium, and yttrium specifically as critical for clean-energy technologies (Squatex, 2026).

The Western Policy Response: DPA Title III, the EU's CRMA Benchmarks, and the Recycling Build-Out

1. U.S. policy: DPA Title III, Executive Order 14241, and the REEshore Act

Beyond the MP Materials, Lynas, and Energy Fuels awards detailed in Section 4, the U.S. government's core legal tool is Defense Production Act Title III, which President Trump's March 2025 Executive Order 14241 (“Immediate Measures to Increase American Mineral Production”) explicitly delegated to the Secretary of Defense for accelerating domestic critical-mineral production, including rare earths (White House, Executive Order 14241). Legislatively, the REEshore Act, reintroduced in bipartisan form, would prohibit the Department of Defense from using rare-earth metals processed or refined in China for defense-system contracts entered into or renewed on or after 31 December 2026, with waivers permitted only in limited circumstances (King & Spalding, REEshore Act summary). In February 2026, the White House announced Project Vault, a $12 billion initiative to establish a U.S. Strategic Critical Minerals Reserve, alongside the formation of a critical-minerals trade zone and a critical-minerals ministerial process (Congressional Research Service, IF13171, updated June 2026). The FY2025 National Defense Stockpile potential acquisitions included 300 tons of NdPr oxide, 450 tons of NdFeB magnet block, and 60 tons of samarium-cobalt alloy, though FY2026 Annual Materials Plan figures were not yet released as of the February 2026 USGS summary (USGS MCS 2026).

2. The EU's Critical Raw Materials Act: rare earths as a Strategic Raw Material

The EU's Critical Raw Materials Act (Regulation (EU) 2024/1252, in force since May 2024) designates rare earths for magnets — specifically neodymium, praseodymium, terbium, dysprosium, gadolinium, samarium, and cerium — as a Strategic Raw Material, subject to the Act's headline 2030 benchmarks: EU extraction capacity covering at least 10% of annual consumption, processing capacity covering at least 40%, and recycling capacity covering at least 25%, with no single third country supplying more than 65% of EU consumption of any strategic raw material (European Commission statement, CRMA rationale; European Commission, CRMA overview). The European Commission projects EU rare-earth-metal demand will grow sixfold by 2030 and sevenfold by 2050 (European Commission, CRMA sector page). A European Court of Auditors 2026 review found the EU's domestic mining capacity for all strategic raw materials already sat near 8% of consumption when the 10% target was set, but flagged that for materials like rare earths specifically, the gap to the target remains wide, and criticized the aggregated (non-material-specific) nature of the benchmarks as reducing their practical meaningfulness (European Court of Auditors, Special Report SR-2026-04). A Eurometaux analysis separately flagged Sweden's Norra Kärr deposit as theoretically capable of supplying up to 80% of the EU's 2030 dysprosium needs alone if developed, against current EU domestic rare-earth capacity below 5% (Eurometaux/EESC presentation). In December 2025 the Commission proposed amending the CRMA to increase rare-earth-magnet recyclability and, by Q2 2026, to restrict exports of rare-earth permanent-magnet scrap to keep more secondary material inside the EU (European Commission, COM(2025) 945, 3 Dec 2025).

3. Recycling: HyProMag, Noveon, and Solvay's Loop, from <1% to a policy mandate

Fewer than 1% of rare-earth magnets are currently recycled globally, but government mandates are now the dominant driver of investment in the space (Rare Earth Exchanges, magnet recycling technologies review, Jul 2025). HyProMag, a UK-based recycler, uses a hydrogen-decrepitation process to recover magnet alloy powder from end-of-life motors and hard-disk drives, and has reported piloting-stage progress toward commercial-grade recycled powder blended with virgin material to meet incoming EU CRMA minimum-recycled-content thresholds (Nasdaq/HyProMag press release, 16 Jun 2025). HyProMag's approach is one of only two recycling projects selected from the entire 17-project European Raw Materials Alliance (ERMA) global portfolio, and directly supports the UK's Critical Minerals Intelligence Centre “Vision 2035” target of 20% of UK critical-mineral demand from domestic recycling by 2035 (HyProMag/ERMA portfolio reference, Mar 2026). Noveon Magnetics, a U.S. recycler backed by Department of Defense support, operates an “EcoFlux” magnet-to-magnet recycling process and in November 2025 announced a partnership with Solvay — whose “Loop” rare-earth recycling technology recovers light and heavy rare-earth materials — to jointly supply recycled rare-earth feedstock (PR Newswire, Noveon/Solvay partnership, 12 Nov 2025). The EU's CRMA recycling benchmark rose from an initial 15% proposal to a final 25% target specifically because policymakers judged rare-earth-magnet circularity achievable at scale; modeling cited by the European Parliament suggests systematic magnet recycling could supply up to 19% of new-magnet material requirements, including 10% of neodymium and 11% of dysprosium needs for new wind installations, by 2030 (Intereconomics, EU raw materials strategy analysis).

Environmental Legacy and the WTO Precedent: Baotou's Tailings, Malaysia's Radioactive Waste, and China's 2014 Loss

1. Baotou's Weikuang tailings dam: sixty years of unlined radioactive sludge storage

The Weikuang Dam, an artificial tailings lake built in the 1950s near Baotou, Inner Mongolia, without the waterproof liners that became standard practice in the West by the 1970s, stores decades of waste from rare-earth and iron-ore processing at the nearby Bayan Obo mine complex. A 2025 New York Times investigation found the sludge, during dry winter months, creates dust carrying lead, cadmium, and thorium contamination, while summer rainfall creates a surface water layer that mixes with thorium and other contaminants, driving groundwater contamination roughly seven miles from the Yellow River, a water source for some 150 million people (New York Times, 5 Jul 2025). Peer-reviewed gamma-ray survey work measured average thorium-232 concentrations in the Baotou tailings dam at 321 ± 31 mg/kg — roughly 34.6 times higher than in local soil in nearby Guyang County — confirming the scale of radionuclide concentration in the waste stream (PubMed, in-situ gamma-ray survey of Baotou and Bayan Obo tailings dams, 2016). A 2025 Business and Human Rights Resource Centre summary of the same NYT reporting noted academic studies linking the contamination to intellectual-development disorders and pollution-related illness among children in the region, and that journalists attempting to visit the Weikuang Dam were detained and questioned by police and Baogang security, who described the site as a “business secret” (Business & Human Rights Resource Centre, Oct 2025). China's government has acknowledged the damage directly: a 2012 State Council white paper stated “excessive rare earth mining has resulted in landslides, clogged rivers, environmental pollution emergencies and even major accidents and disasters” (Nuclear-News, citing 2012 State Council white paper).

2. Lynas Malaysia: fourteen years of unresolved radioactive-waste disposal

Since starting operations in 2012, the Lynas Advanced Materials Plant (LAMP) in Kuantan, Malaysia has generated Water Leach Purification (WLP) residue — a radioactive byproduct of the cracking-and-leaching stage of separation containing thorium-232, with a half-life of roughly 14 billion years, at concentrations reported by campaign groups at 1,600–1,953 parts per million (AID/WATCH, Stop Lynas campaign summary; Greenpeace Malaysia, Jun 2026). By February 2023, Lynas had produced over one million metric tons of this waste, with the Malaysian government requiring the company in 2023 to halt cracking-and-leaching operations in Malaysia and either relocate that step or present a permanent-disposal plan (Wikipedia, Lynas corporate history). By 2024, the Malaysian government reported Lynas had produced over one billion kilograms of radioactive waste since 2012 (Straits Times, 14 Nov 2024). In March 2026, Malaysia's Science and Technology Minister announced a renewed operating licence with new conditions: no new permanent disposal facility will be built in Malaysia, and WLP residue production must cease entirely by 2031, with existing accumulated residue to be neutralized or treated under approved methods (Malay Mail, 6 Mar 2026). Malaysian civil-society groups continue to dispute Lynas's technical claims about thorium removal, with the Save Malaysia Stop Lynas coalition stating in June 2026 that even Malaysia's own nuclear regulator “confirmed that no such thorium extraction project ever materialised, despite allocated grants” (Greenpeace Malaysia, Jun 2026).

3. The WTO's 2014 ruling: China's export quotas found illegal, but market power endures

On 26 March 2014, a WTO dispute-settlement panel ruled that China's export duties, export quotas, and export-quota administration requirements on rare earths (along with tungsten and molybdenum) violated Article XI:1 of the GATT 1994 and China's WTO Accession Protocol, rejecting China's defense that the measures were justified as environmental-conservation measures (USTR press release, 26 Mar 2014). The panel specifically found China's restrictions “encourage domestic extraction and secure preferential use of those materials by Chinese manufacturers” rather than serve a genuine conservation purpose, since China did not simultaneously restrict domestic consumption (UK Government, Chinese reaction to WTO ruling, 2014). China's appeal was rejected in August 2014, and China formally scrapped its rare-earth export quota system at the end of December 2014, replacing it with export licensing rather than fixed quantitative limits (BBC News, 5 Jan 2015). The 2014 ruling is the direct legal precedent analysts cite when assessing the 2025 export licensing regime: because China switched from quotas (found illegal) to case-by-case licensing (a discretionary tool not directly addressed by the 2014 ruling), the current controls sit in a different, less clearly WTO-inconsistent legal category, even though their practical market effect — restricting supply to non-Chinese buyers — is similar (CSIS, 14 Apr 2025).