Critical Minerals & Rare Earth

Overview / thesis

Critical minerals are the physical building blocks of modern technology, defense systems, and the energy transition. Without them there are no EV batteries, no wind-turbine or motor magnets, no fighter-jet guidance systems, no semiconductors, and no AI data centers. The word "critical" is not a geological classification — it is an economic and national-security designation. A mineral earns the label when two conditions hold at once: it performs an essential function in products vital to the economy, defense, or energy systems, and its supply chain is geographically concentrated, geopolitically exposed, or otherwise disruptible. The investment story for the entire sector flows from the collision of those two facts with a third: one country — China — controls the chokepoints, and the West has decided, with real money, to build a parallel chain around it.

The one-line thesis

The structural bet is not on rocks in the ground. The world has enough ore; what it does not have is enough non-Chinese processing. China's grip tightens as you move downstream — roughly 60-70% of rare earth mining, ~90% of separation, ~93-94% of permanent-magnet manufacturing, and ~99-100% of heavy-rare-earth refining. That asymmetry is the whole game. The single insight that recurs across every source is that processing and magnet-making, not mining, are the bottleneck and the highest-value, deepest-moat layer of the chain. Companies building midstream capacity — separation, metals reduction, magnet manufacturing — are more strategically valuable than pure miners, and the market is being forced to re-rate them as Western governments underwrite the buildout. The corollary "so what": vertical integration from mine-to-magnet plus government backing (equity, price floors, offtake) is the configuration that wins; mining assets without downstream integration carry limited strategic value.

China's dominance is illustrated by its own arithmetic: it produces ~300,000+ tonnes/year of NdFeB magnets against the entire non-Chinese world's <15,000 tonnes — roughly a 20:1 ratio. Two decades ago China held ~50% of permanent-magnet production; by 2024 that share was 94%. Separation is hard for a reason rooted in chemistry — the 17 rare earths have nearly identical ionic radii and must be teased apart over 100-300+ sequential solvent-extraction stages ("sorting 17 shades of white paint"). China has perfected industrial-scale separation over 30+ years and holds an estimated ~25,000 rare-earth patents. That is why its ~90% processing share matters more than its ~70% mining share.

Why it matters now — the catalyst stack

What turned a slow-burning vulnerability into an active, fundable thesis is China weaponizing the chain and the West responding with capital, not just rhetoric. The analyst line that captures the regime change: "In 2010, the alternatives were embryonic. In 2026, they are real, funded, and already being built." Each episode of Chinese coercion makes it easier to justify subsidies, accelerate permitting, and mobilize capital.

The coercion is concrete. China's export-control escalation ran from gallium/germanium licensing (July 2023) and graphite (Oct 2023), to controls on 7 heavy REEs plus all compounds, metals, and magnets (April 4, 2025), to extraterritorial reach over "internationally made" products using Chinese REEs (Dec 1, 2025), to a complete export ban to the U.S. on gallium, germanium, antimony, and superhard materials (Dec 2024). Effects were immediate: European NdFeB magnet prices hit 6x Chinese domestic prices during April-May 2025 and automakers cut production; antimony trioxide surged 228% since January 2024 to ~$39,000/tonne; gallium and germanium shipments to the U.S. fell to near zero. On January 6, 2026 China banned "dual-use" exports to Japan and began halting rare-earth and magnet shipments to Japanese firms — the sharpest weaponization since the 2010 Senkaku embargo (full-year cost to Japan estimated at ¥2.6 trillion / ~0.43% of GDP).

The clock everyone watches: China's temporary one-year suspension of the October 2025 controls expires in November 2026, after which Beijing retains full legal authority to reinstate the bans. Beijing called the pause a "temporary adjustment" — the regulatory framework stays intact and reactivatable on minimal notice.

The Western response is now institutional and capitalized. Project Vault (Feb 2026) is a $12B strategic stockpile ($10B EXIM direct loan — EXIM's largest ever — plus ~$2B private capital from GM, Stellantis, Boeing, Alphabet, Corning, GE Vernova, with Hartree/Traxys/Mercuria handling purchases), warehousing all 60+ USGS-listed minerals. FORGE (Forum on Resource Geostrategic Engagement, successor to the Minerals Security Partnership) launched at the Feb 4, 2026 Critical Minerals Ministerial with 54 nations, a preferential trading bloc, and coordinated price floors to shield producers from China's market-flooding. The "One Big Beautiful Bill Act" (July 4, 2025) appropriated $2B for the National Defense Stockpile, $5B for an Industrial Base Fund, and authorized up to $350B in DOE financing; DoD can now take direct equity positions. The MP Materials template — $400M DoD equity, $150M loan, and a $110/kg NdPr price floor (contract-for-difference) with 10-year 100% offtake — is the blueprint that breaks China's ability to crash prices to kill Western competitors. A Section 232 investigation confirmed the U.S. is 100% import-reliant for 12 critical minerals and >50% for 29 more, with a July 13, 2026 negotiation deadline that is a binary catalyst: failure means tariffs and/or minimum import-price floors, either of which hard-codes a price floor for domestic producers.

Sizing the opportunity

The combined critical-minerals market is ~$325B (2024) — comparable to iron ore (IEA). The build by mineral:

Mineral	2024 Est.	2030 Projected	CAGR
Copper	~$180B	~$250B+	5-7%
Lithium	~$22B	~$55-70B	15-20%
Nickel (battery)	~$35B	~$55B	8-10%
Rare earths	~$12-15B	~$25-30B	10-15%
Graphite	~$25B	~$40B	8-10%
Cobalt	~$15B	~$20B	5-6%

The NdFeB permanent-magnet market — the strategic heart of the chain — is ~$16-20B (2024), projected to reach $28-59B by 2030-2034. Value-chain revenue pools sit roughly at: mining ~$150B+ (20-50% gross margins, cycle-dependent), processing/cracking-leaching ~$40-60B (25-35%, China 70-90% consolidated), separation/refining ~$30-50B (30-50%, China 90%+ in REEs), magnets/cathodes/cells ~$100-120B (15-25%, China 94% in magnets), and end products in the trillions. The deepest moats and the most entrenched Chinese dominance both sit in the midstream — which is precisely where the Western strategy aims to build parallel capacity. Adjacent sizing worth carrying: the DLE (direct lithium extraction) market is growing at a 19.6% CAGR through 2036, and the market value of recycled battery metals grew ~11-fold between 2015 and 2023.

The demand drivers

The tailwinds are secular and stack on top of each other:

Tailwind	Duration	Magnitude
EV adoption (17M in 2024 → 40M+ by 2030)	10+ years	Enormous
Grid-scale storage (GWh → TWh)	10+ years	Large
AI & data centers (power doubling every 2-3 yr)	5-10 years	Growing
Defense spending (NATO 3%+ GDP)	5-10 years	Significant
Western supply-chain buildout (FORGE, Vault, IRA, DPA)	5-10 years	Transformational
China export controls (force diversification)	3-5 years	Price floor + urgency
Declining ore grades globally	Permanent	Structural cost inflation

Per-unit intensity makes the demand concrete: a single EV battery needs ~8 kg lithium, ~35 kg nickel, ~14 kg cobalt, ~65 kg graphite; a single-motor EV needs ~100 g dysprosium and ~9 g terbium. Defense is the demand source that cannot be value-engineered away because the periodic table does not negotiate: an F-35 carries ~920 lbs of rare-earth materials (including ~50 lbs of SmCo magnets that survive 350°C continuous), a Virginia-class submarine 4,200+ kg. The underappreciated convergence is AI + critical minerals — data-center power demand drives copper, and GaN power electronics drive gallium. Defense modernization, EVs, wind, and automation remain robust demand regardless of short-term geopolitical friction.

The structural origin — why China got here first

Three converging forces created the modern industry: materials science (1980s-2000s — NdFeB magnets ~10x stronger than ferrite, lithium-ion at 3-5x the density of lead-acid); China's strategic foresight (Deng Xiaoping, 1992: "The Middle East has oil, China has rare earths" — decades of state investment built dominance); and the simultaneous arrival of the energy transition and the AI revolution in the 2020s, which multiplied demand. China didn't out-innovate so much as out-invest and out-last, using low-cost ionic-clay deposits for heavy REEs and aggressive export pricing to drive Western competitors out of business. Chinese magnet-maker margins (JL MAG) compressed from ~10% in mid-2022 to ~3% in 2024 under overcapacity and price wars — itself the structural opening for Western producers with government support and captive defense demand.

The Western architecture and the chokepoint map

The cross-cutting strategic frame (from Steve Zissouu's Meridian Report) is a USA-anchored chain built on unprecedented intergovernmental cooperation, with a trilateral spine: Australia (feedstock) → USA (refining) → Japan (downstream OEM / magnets / offtake). Australia is the feedstock powerhouse (Lynas Mt Weld, Iluka Eneabba, Arafura Nolans, the Energy Fuels/Astron Donald Project), with monazite already produced as a byproduct of existing mineral-sands operations. Japan is the most strategically prepared ally — it treated the 2010 embargo as a wake-up call, cut Chinese REE dependence from ~90% to ~58-70%, and built advanced processing focused on capability over scale, manufacturing ~4,500 tonnes of NdFeB magnets annually (the bulk of non-Chinese supply). The U.S.-Australia Framework (Oct 20, 2025) committed at least $1B from each side within six months, anchoring an $8.5B Australian project pipeline. (Company- and country-specific detail belongs on dedicated pages; see UUUU, MP, LYC.AX, ALB, FCX, AREC and the Western-supply-chain and country-relationship sections.)

The chokepoint cascade is the map every investor should hold in mind:

Mining (~70%) → Processing (~90%) → Magnets (~93-94%) → Heavy REE (~99-100%)

Barriers to entry rise the same way — mining is the easiest to replicate; magnet manufacturing (Hitachi-era IP, $300-500M plant cost, decades of process know-how, multi-year customer qualification cycles) is the hardest. Magnet manufacturing is the ultimate chokepoint, which is why a first Western-scale magnet producer is the critical variable for the whole thesis.

The key debates

Three honest tensions sit at the center of the thesis and should not be smoothed away:

1. Can the West actually break the dominance — and on what timeline? The consensus answer is no, not soon. Realistic projections have China holding 75%+ market share through 2030; breaking dominance is a 5-10+ year project even at maximum effort. Lynas's May 2025 first-ever commercial separated Dy/Tb outside China is strategically pivotal but "a drop in the ocean of demand." The bet is on the direction of travel, priced over years, not a fast displacement.

2. Is government support a durable moat or a crutch? Western midstream is structurally high-cost with slim margins (CSIS calls it "a high-cost endeavor with slim profit margins"; MP ran negative materials-segment EBITDA during ramp). Western producers need sustained policy backing — price floors, offtake, equity — to compete with subsidized Chinese output. The price-floor debate already exposed allied friction: reports that the U.S. administration stepped back from guaranteed minimum prices knocked Australian mining stocks. If political will or price floors prove insufficient, the economics don't stand on their own yet.

3. Demand-side and cycle risk cut against the bull case. Headwinds: technology substitution (sodium-ion, ferrite/rare-earth-free motors — inferior but capable of capturing the low end), structural oversupply (Indonesian nickel, Chinese lithium) pressuring prices, 10-15 year Western permitting timelines that slow any supply response, and the chance the China pause becomes permanent (a short-term sentiment dampener that removes urgency). Cobalt is being actively engineered out of cathodes. NdPr prices fell 20%+ in 2024. The avoid-list that falls out of this: Indonesian nickel pure-plays, early-stage DLE without DOE backing, and cobalt pure-plays.

Where the alpha concentrates

If processing is the bottleneck, the differentiated angles are the ones that attack it. Two recur as underpriced. First, recycling and urban mining as strategic (not merely environmental) supply: IEA projects recycling cuts new-mine needs by 40% for Cu/Co and 25% for Li/Ni by 2050, the EU Battery Regulation mandates 80% Li and 95% Co/Cu/Ni recovery by end-2031, and magnet-to-magnet recycling (HPMS/HDDR) plus e-waste recovery create non-Chinese feedstock that bypasses Chinese ion-adsorption-clay HREEs. Second, brownfield / waste-stream recovery — pulling minerals from coal ash (~11 million tonnes of REEs in U.S. coal ash, ~8x known reserves; Appalachian ash the richest at 591 mg/kg), red mud (>4 billion tonnes globally, ~2x REE enrichment vs. input ore, scandium holding ~95% of the value), and acid mine drainage (treatment sludge ~2,600x more concentrated than raw drainage) — which sidesteps the decade-long permitting that kills new hard-rock mines, reaching commercial operation in 3-7 years on negative-cost feedstock. Novel separation technology (chromatography, RapidSX, Flash Joule Heating) is the enabling layer: FJH claims >90% purity and yield in seconds, 87% less energy, 84% fewer emissions, and 54% lower operating cost than conventional hydrometallurgy.

Net: the sector is in early-cycle recovery (Feb 2026) off the 2023-2024 trough across lithium and rare earths, with copper in structural deficit, uranium in a strong bull market, and China's export controls plus the FORGE/Project Vault response putting a policy floor under prices. The direction is clear; the timeline is long; the differentiated bets are in the midstream the West is being forced to fund.

How it works

Critical minerals are the physical building blocks of modern technology, defense systems, and the energy transition. Without them there are no EV batteries, no wind turbine magnets, no fighter jet guidance systems, no semiconductors, and no AI data centers. "Critical" is not a geological classification — it is an economic and national-security designation. A mineral earns the label when two things are true at once: (1) it performs an essential function in products vital to the economy, defense, or energy, and (2) its supply chain is geographically concentrated, geopolitically risky, or otherwise vulnerable to disruption. The USGS 2025 Final List (published Nov 6, 2025) expanded to 60 minerals — the 50 from 2022 plus boron, copper, lead, metallurgical coal, phosphate, potash, rhenium, silicon, silver, and uranium, with 15 rare earth elements listed individually. The EU's Critical Raw Materials Act (March 2024) names 34 CRMs plus 17 Strategic Raw Materials and sets 2030 targets: 10% domestic extraction, 40% domestic processing, 25% from recycling, and no more than 65% dependence on any single third country.

The category exists because of three converging forces. Materials science (1980s–2000s) produced rare-earth permanent magnets roughly 10x stronger than ferrite and lithium-ion chemistry that stores energy at 3–5x the density of lead-acid. China's strategic foresight (1990s–2010s) — Deng Xiaoping in 1992: "The Middle East has oil, China has rare earths" — converted decades of state investment into dominant market share. And the energy transition plus the AI revolution (2020s) multiplied demand for the same handful of elements simultaneously. The physical stakes are concrete: a single EV battery requires roughly 8 kg lithium, 35 kg nickel, 14 kg cobalt, and 65 kg graphite; every wind turbine, EV motor, and precision-guided missile uses NdFeB magnets; gallium and germanium are essential to advanced chips, 5G, and defense electronics; and grid-scale storage runs on lithium-ion and vanadium redox-flow chemistries that are themselves mineral-intensive.

Why the chemistry creates the moat

The single most important first principle in this sector: the processing bottleneck is more important than the mining bottleneck. The world has enough ore; it does not have enough non-Chinese processing. China's dominance is not really about deposits — it is about the chemical engineering that turns mixed ore into separated, purified, usable material. That is why China's ~90% processing share matters more than its ~70% mining share, and why companies building midstream capacity are more strategically valuable than pure miners. China's grip tightens at every step downstream: Mining (~70%) → Processing/Separation (~90%) → Magnets (~93–94%) → Heavy REE refining (~99–100%).

Rare earth elements (REEs) are 17 metallic elements — the 15 lanthanides plus scandium and yttrium. They are not truly "rare" in the crust; they rarely concentrate into economically extractable ore bodies. They split into Light Rare Earths (La, Ce, Pr, Nd, Sm — more abundant, easier to process) and Heavy Rare Earths (Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, Y — scarcer, harder to separate, far more valuable). The reason separation is hard is atomic: REEs have nearly identical ionic radii and chemical properties, so isolating one from the next is "like sorting 17 shades of white paint." Traditional separation requires hundreds to thousands of sequential solvent-extraction (mixer-settler) stages. China has perfected industrial-scale separation over 30+ years and holds roughly 25,000 rare-earth patents — an IP moat as deep as the chemical-engineering one.

NdFeB magnet physics: neodymium aligns atomic magnetic domains to deliver a maximum energy product (BHmax) of ~50–52 MGOe, about 10x stronger than ferrite. The limiting factor is heat. Plain NdFeB loses coercivity exponentially above 60–80°C and suffers irreversible magnetization loss above 150°C. Adding dysprosium or terbium at 7–10% by weight — they substitute for neodymium atoms at grain boundaries, creating the Dy₂Fe₁₄B phase with a dramatically stronger anisotropy field — raises the Curie temperature so the magnet survives the high-temperature environment of an EV traction motor or a jet engine. A typical single-motor EV needs roughly 100 g of dysprosium and ~9 g of terbium. This is why heavy rare earths are the acute vulnerability and why China's ~99–100% control of HREE separation is the hardest chokepoint in the entire sector.

The atomic-level irreplaceability. Several of these elements cannot be engineered around because the periodic table does not negotiate:

• Samarium (Z=62) has six unpaired 4f electrons giving exceptionally high magnetocrystalline anisotropy. Samarium-cobalt (SmCo) magnets hold their magnetism to 350°C continuously (excursions to 550°C), versus NdFeB's ~150°C ceiling, with a temperature coefficient of only −0.03%/°C versus NdFeB's −0.12%/°C. Energy density 16–32 MGOe. Missile nose cones routinely exceed 200°C in supersonic flight; the F-35 carries ~50 lbs of SmCo magnets for turbomachine pumps, actuators, and missile components.
• Dysprosium and terbium (Z=66, 65): high magnetic susceptibility from their 4f configurations; the additives that keep EV and fighter-jet motors from self-destructing under heat.
• Gallium (Z=31): forms gallium nitride (GaN) with a 3.4 eV bandgap — more than 3x silicon's 1.1 eV — enabling breakdown voltages of 40–80 V/µm (vs silicon's 20), operation to 400°C (vs silicon's 150°C), and 5–10x the power density in the same package. AESA radars carry 300+ transmit/receive modules, each needing a GaN amplifier capable of ~200W; at X-band (8–12 GHz) silicon simply overheats. Gallium also melts at 29.76°C (in your hand); GaAs (1.42 eV direct bandgap) serves LEDs, solar cells, and 5G RF; it is extracted as a byproduct of the Bayer alumina process (bauxite holds ~50 ppm Ga). Semiconductor-grade gallium reaches 99.99999% (7N) via zone refining.
• Germanium (Z=32): a 0.67 eV bandgap creates a natural long-pass optical filter — visible photons (1.8–3.1 eV) are absorbed while infrared (0.08–0.6 eV) passes at up to 97% transmission when AR-coated. Over 60% of lens elements in mid-to-low-end IR systems are germanium; its refractive index of 4.0 enables thin compact lenses for night vision, FLIR, and missile seekers. Semiconductor-grade germanium reaches 99.999999999% (11N). It is a byproduct of zinc smelting (GeO₂ fiber-optic dopant; transparent 2–12 µm).
• Antimony (Z=51): a metalloid that hardens lead-antimony ammunition alloys (1–5% Sb) and, as Sb₂O₃, acts as a flame-retardant synergist — during combustion it forms halogenated antimony species that scavenge hydrogen radicals and break the fire's chain reaction. US ammunition alone consumes >600 tons of antimony annually. Emerging use in liquid-metal batteries.
• Tungsten (Z=74): highest melting point of any metal (3,422°C) and a density of 19.25 g/cm³ (2.3x steel). Tungsten carbide (WC) has a Mohs hardness of 9–9.5. These properties make it irreplaceable for APFSDS kinetic-energy penetrators, bunker-busters, cutting tools, and drill bits — the only comparable material is depleted uranium, which is radioactive. Ore minerals: wolframite and scheelite (0.3–1.5% WO₃).
• PGMs (Pt, Pd, Rh, Ir, Ru, Os): noble metals with exceptional catalytic properties — Pt/Pd for catalytic converters, Pt for hydrogen fuel-cell catalysts, Ir for PEM electrolyzer anodes. South Africa's Bushveld Complex holds ~70% of global Pt and ~80% of Rh reserves; ore grades 3–10 g/tonne; Pt is ~0.005 ppm in the crust.

Lithium is the lightest metal (Z=3); its high electrochemical potential makes it ideal for rechargeable batteries. Two source types: hard rock (spodumene, LiAlSi₂O₆), grade 1–2.4% Li₂O, concentrated to 6–7.6% Li₂O by flotation, dominant in Australia; and brine, 200–4,000 mg/L lithium, solar-evaporated over 12–24 months to ~6,000 ppm, dominant in Chile/Argentina. Cobalt stabilizes NMC/NCA cathode crystal structure to prevent thermal runaway; ~73% comes from the DRC, often as a copper byproduct, and the industry is reducing cobalt content (NMC 111 → 811 → 955). Nickel raises energy density: sulfide ores (0.7–3% Ni) are underground-mined, floated, and smelted (Canada/Russia/Australia); laterite ores (1–2% Ni) are open-pit and HPAL-processed (higher cost, Indonesia/Philippines). Graphite is carbon in hexagonal layers; battery anodes intercalate lithium between graphene sheets. Ore grades 3–25% carbon, floated to 80–95%, with battery-grade requiring 99.95%+ purity.

How the rare-earth process actually works — mine to magnet

The chain runs: Mining → Cracking & Leaching → Separation → Metals Reduction → Magnet Making → OEM.

• Step 1 — Mining (feedstock). Monazite (a phosphate mineral carrying both light and heavy REEs) from mineral sands (Chemours, Tronox, Iluka); bastnäsite at Mountain Pass and Bayan Obo; ion-adsorption clays in southern China and Myanmar (HREE-rich, low-cost surface mining, higher environmental impact). Hard-rock deposits use standard drill-blast-load-haul. The output is raw ore or a mixed concentrate with all 17 REEs jumbled together. Mining has the lowest barriers to entry — geology, capital, and permits — and rare earths are geologically abundant.
• Step 2 — Cracking & Leaching. Monazite is dissolved in NaOH or H₂SO₄ at elevated temperature to produce a mixed rare-earth solution or carbonate. Radioactive thorium must be managed throughout — the handling license is itself a barrier. Beneficiation (crush, grind, flotation) lifts concentrate to ~50–70% REO (Mixed Rare Earth Concentrate / carbonate).
• Step 3 — Separation. The critical bottleneck. The mixed solution is fed through 100–300+ (up to thousands of) mixer-settler stages with organic solvents; each stage preferentially extracts one REE. Output: individual high-purity oxides (99.5–99.99%). China holds ~90% of global capacity. Novel approaches challenge the SX paradigm — ReElement uses Ligand Assisted Displacement (LAD) chromatography (resin-filled columns, claimed 100x efficiency, 80% less waste, far smaller footprint), and UCU (Ucore) is developing RapidSX column-based separation (at least 3x faster than conventional SX with a dramatically reduced footprint).
• Step 4 — Metals Reduction. REE oxides are converted to metals via molten-salt electrolysis or metallothermic reduction — energy-intensive. China controls ~90% of rare-earth metal production; the only scaled Western producer is Less Common Metals (UK), acquired by USA Rare Earth.
• Step 5 — Magnet Manufacturing. The highest-barrier step. NdFeB alloy → strip casting → hydrogen decrepitation → jet milling (3–5 µm powder) → magnetic-field pressing → sintering at 1,000–1,100°C (under vacuum) → machining → coating → magnetization. A modern sintered-NdFeB plant costs $300–500M and requires complex IP (Hitachi patents), decades of process know-how, and multi-year customer qualification cycles. China controls ~93–94% of sintered NdFeB, up from ~50% two decades ago.
• Step 6 — OEM Integration. Magnets go into EV traction motors, wind turbines, defense systems, industrial robots, MRI machines, and consumer electronics.

Alloy chemistry at the downstream end: NdFeB is roughly 30% Nd / 65% Fe / 5% B (EVs, wind, electronics); SmCo is samarium + cobalt (aerospace/defense high-temp).

How the lithium process works — deposit to battery

Three distinct routes with very different physics and economics:

HARD ROCK: Mining → Crushing → Flotation (SC6 concentrate 6–7% Li₂O) → Calcination 1050°C → Acid Roast → Leaching → Purification → Li₂CO₃ or LiOH
           Flotation recovery 70–85%. Greenbushes (WA) is the highest known grade at 2.4% Li₂O.

BRINE:     Pumping → Solar Evaporation 13–24 months → Impurity Removal (Na, K, Mg, B) → Na₂CO₃ Precipitation → Li₂CO₃
           Recovery 20–40%. Land footprint ~5 km² per 25,000 tpa LCE. Salar de Atacama runs 1,400–1,800 ppm Li.

DLE:       Pumping → Selective Adsorption (hours) → Desorption with dilute acid → Concentration → LiOH / Li₂CO₃
           Recovery 80–95% (vs 20–40% for evaporation). Three tech families: adsorption (LiAl₅O₈ sorbents),
           ion exchange (resins, works <100 ppm), membrane (nanofiltration/electrodialysis, pilot stage).
           DLE market CAGR ~19.6% through 2036.

Direct Lithium Extraction (DLE) is the disruption to watch: it skips the 12–24-month evaporation ponds, roughly doubles-to-quadruples recovery, and shrinks land footprint — the recovery jump from 20–40% to 80–95% is the technology-disruption indicator for the lithium sub-sector.

Brownfield recovery — extracting minerals from waste streams

A parallel "waste-to-minerals" pathway is emerging because new hard-rock mines take 10–15 years to permit while byproduct recovery at existing permitted industrial facilities can reach commercial operation in 3–7 years, with feedstock at zero-or-negative cost and infrastructure already in place. This is the explicit logic of DOE's $275M DE-FOA-0003583 program ("Mines & Metals Capacity Expansion — Piloting By-Product Critical Minerals and Materials Recovery"), which targets advancing technologies from TRL 4-5 to TRL 7.

• Gallium and germanium are "scattered/dispersed" metals — they rarely form independent ore minerals and instead substitute into common sulfides (gallium for aluminum, germanium for zinc in sphalerite). This is the "byproduct trap": output cannot scale independently of primary aluminum/zinc/copper production. During copper smelting germanium almost entirely vaporizes into flue-gas residues; zinc smelting residues contain ~0.5 wt% germanium and ~0.4 wt% gallium. Historically only ~1.4% of available gallium and ~0.7% of germanium were ever extracted from host ores.
• Red mud (Bayer-process bauxite residue) is generated at 2.5–4 tonnes per tonne of aluminum; global stockpiles exceed 4 billion tonnes, growing ~150 million tonnes/year. Because REEs don't dissolve in NaOH, they concentrate in the residue at an enrichment factor of ~2x the input ore — typically 500–1,700 ppm total REEs (3,000–4,000 ppm in some feedstock, ~25x upper-continental-crust levels), with scandium holding ~95% of the rare-earth economic value. Red mud also runs 38–50% iron oxide and 7–10% silica, with the valuable elements locked inside refractory silicon-aluminum crystal phases.
• Coal ash from Appalachia has the highest US REE concentrations. Heavy REEs preferentially associate with organic matter, so cleaned high-organic coal is HREE-enriched. Combustion concentrates non-volatile REEs by 4–10x; Appalachian Basin ash averages 591 mg/kg REE (vs 403 for Illinois Basin, 264–337 for Powder River). An estimated 11 million tonnes of REEs sit in US coal ash — roughly 8x known national reserves.
• Acid mine drainage (AMD): raw drainage is dilute, but treatment sludge holds on average 2,600x more REEs than the raw water; ~6,000 t/yr of REEs flow through Appalachian AMD sites. Pilot work (WVU / Mount Storm) suggests Appalachia could yield ~800 t/yr of REEs — about the defense industry's entire annual requirement.
• Phosphogypsum (fertilizer waste) is another REE-bearing stream (Mosaic / Rainbow Rare Earths' Uberaba Project, Brazil — 24.5% NdPr content).

Flash Joule Heating — the process breakthrough

Flash Joule Heating (FJH), invented in Dr. James Tour's lab at Rice University (seminal Nature paper, January 2020; key patent US 12,054,391 B2) and licensed exclusively to Metallium Ltd., passes direct or pulsed electrical current through resistive material, converting electricity to heat via the Joule effect and reaching over 3,000°C within milliseconds (typically 10 ms to 1 s) at nearly 100% electrical-to-thermal efficiency. Combined with chlorination (FJH-Cl₂) it achieves single-step separation: transition metals form volatile chlorides at ~1,000°C and boil off, while rare-earth oxides only react above ~1,350°C, converting to stable oxychlorides — yielding >90% purity and >90% yield in seconds versus hours for conventional acid leaching.

The economics versus traditional hydrometallurgy: 87% reduction in energy use, 84% fewer GHG emissions, 54% lower operating costs, and complete elimination of water and acid consumption (acid use down ~87% where any is used). For refractory red mud specifically, FJH restructures the silicon-aluminum crystal matrix before acid ever touches it — mixing red mud with conductive additives (carbon black/graphite) and applying rapid pulses to 2,000–3,000°C reconstructs the locked phases, so subsequent leaching with dilute 0.1 M HCl achieves >80% extraction versus 40–70% on untreated material with concentrated acid. Documented improvements: +72% neodymium recovery, +103% scandium recovery, +514% titanium recovery. The staged flow is feedstock → FJH pretreatment (Metallium) → acid leach → solvent extraction → separated oxides (ElementUS hydrometallurgy), with FJH sitting at the junction that turns an economically marginal process into a viable one. First commercial site: Chambers County, Texas, commissioning Q1 2026.

Recycling and urban mining — the secondary supply chain

Recycling is becoming a strategic supply source, not just an environmental one. The market value of recycled battery metals grew ~11-fold between 2015 and 2023; the IEA projects recycling reduces new-mine needs by ~40% for Cu and Co and ~25% for Li and Ni by 2050. The EU Battery Regulation mandates 80% Li and 95% Co/Cu/Ni recovery by end-2031.

Three lithium-battery recycling routes, with sharply different physics:

Route	Temperature	Recovers	Li Recovery	Advantage
Pyrometallurgical	800–1,200°C	Co, Ni, Cu as mixed alloys	<50% (Li lost to slag)	Low capex, handles mixed feedstock
Hydrometallurgical	Ambient–100°C	Individual metals via SX/precipitation	80–95%	Higher purity, recovers Li
Direct recycling	Moderate	Preserves cathode crystal structure	>95% potential	Lowest energy, fastest, cheapest

NdFeB magnet recycling (HPMS/HDDR) is a magnet-to-magnet shortcut that bypasses most of the mine-to-magnet chain: expose magnets to H₂ gas at room temperature → the Nd-rich grain boundary absorbs H₂ to form NdH₂ with ~3x volume expansion → the magnet decrepitates to powder → heat to ~800°C under H₂ (disproportionation) → remove H₂ under vacuum (recombination) to reform Nd₂Fe₁₄B grains. This is the pathway targeted for e-waste and end-of-life magnets, and it reduces dependence on Chinese ion-adsorption-clay HREEs.

The unit economics and cost structure

Value capture is shaped like a pyramid and the margins are not where intuition expects. The combined critical-minerals market was ~$325B in 2024 (comparable to iron ore, per the IEA). By layer:

Layer	Revenue Pool	Gross Margins	Consolidation	Barriers to Entry
Mining	~$150B+	20–50% (cycle-dependent)	Moderate	Capital ($500M–5B), permits (5–15 yr), geology
Processing (cracking/leaching)	~$40–60B	25–35%	Very high (China 70–90%)	Chemistry, environmental permits, radioactive handling
Separation/refining	~$30–50B	30–50%	Very high (China 90%+ REEs)	SX IP, 100–300+ stages, qualification
Magnets/cathodes/cells	~$100–120B	15–25%	High (China 94% magnets)	Customer qualification, IP, capital
End products (OEMs)	Trillions	Varies	Fragmented	Brand, engineering, scale

The crucial subtlety: separation/refining carries the deepest moat and the best theoretical margins (30–50%), but in the real, China-dominated market the midstream is a "high-cost endeavor with slim profit margins." Separation/refining absorbs ~80% of mine-to-metal capital cost. MP Materials posted negative materials-segment EBITDA during its midstream ramp (Q4 2024); processor gross margins realistically run 10–25%. Chinese magnet makers like JL MAG saw margins compress from ~10% (mid-2022) to ~3% (2024) on overcapacity and price wars — which is precisely the structural opening for Western producers backed by government price floors and captive defense demand. Upstream margins are violently cyclical: Lynas earned a 62% gross margin at the 2022 peak but fell to 19–28% in FY2024–25 on depressed NdPr prices, against a ~35% historical median for integrated producers. The lowest-cost Chinese producers operate near ~$11/kg REO, which is why Western competition is impossible without government support.

The capacity gap quantifies the moat: China produces ~300,000+ tonnes/year of NdFeB magnets while the entire non-Chinese world produces under 15,000 tonnes — roughly a 20:1 ratio. The downstream NdFeB magnet market itself is ~$16–20B (2024), projected to $28–59B by 2030–2034.

Representative product economics: NdFeB magnets sell at $50–200/kg; separated oxides at NdPr ~$60–80/kg, Dy oxide ~$250–400/kg, Tb oxide ~$800–1,200/kg; lithium chemicals Li₂CO₃ ~$10–12/kg and LiOH ~$11–14/kg; copper ~$4.00–4.50/lb; gallium ~$400–800/kg; germanium ~$1,500–2,500/kg.

The technical metrics that matter

Metric	What It Measures	State of the Art	Investor Relevance
REE separation purity	Individual oxide purity	99.5–99.99%	Higher = magnet qualification
NdFeB BHmax	Magnet energy product	~52 MGOe	Motor efficiency
Lithium ore grade	Li₂O in rock	1.0–2.4% (Greenbushes best)	Lower cost at higher grade
DLE recovery rate	% Li extracted from brine	80–95% (vs 20–40% evaporation)	Technology-disruption indicator
Battery energy density	Wh/kg at cell level	~300 Wh/kg (NMC 811)	More range = more mineral demand
Copper ore grade	% Cu in ore	0.5–1.0% (declining)	Structural cost inflation
Nickel HPAL recovery	% Ni from laterite	83–90% at 225–270°C	Indonesia capacity expansion
Graphite battery purity	% carbon in anode	99.95% (thermal purification)	Spheroidization yield only 30–50%
Gallium purity	Semiconductor grade	99.99999% (7N) via zone refining	Near-total China monopoly
Germanium purity	Semiconductor grade	99.999999999% (11N)	Byproduct of zinc smelting
Recycling recovery	% from end-of-life batteries	95%+ Ni/Co; 80–95% Li	EU mandate: 80% Li, 95% Co/Ni by 2031

Two metrics carry hidden cost: declining copper ore grade (0.5–1.0% and falling) is permanent structural cost inflation, and graphite spheroidization yield of only 30–50% means roughly half the purified material is wasted producing anode-grade spherical graphite.

Key terminology

Term	Definition
NdFeB	Neodymium-iron-boron permanent magnet — strongest permanent magnet type
NdPr oxide	Combined neodymium-praseodymium oxide — key magnet intermediate
TREO	Total Rare Earth Oxide — aggregate REE content of a deposit/product
Monazite	Phosphate mineral carrying light and heavy REEs — key REE feedstock
Spodumene	Lithium-bearing mineral (LiAlSi₂O₆) — primary hard-rock lithium source
DLE	Direct Lithium Extraction — extracts Li without evaporation ponds
Solvent extraction (SX)	Liquid-liquid separation for isolating individual REEs
Mixed rare earth carbonate (MREC)	Intermediate with multiple REEs, before separation
Cracking and leaching	Chemical process to break down monazite/bastnäsite and release REEs
BHmax	Maximum energy product — key magnet performance metric (MGOe)
Curie temperature	Temperature above which a magnet loses its permanent magnetism
FJH	Flash Joule Heating — DC pulse to >3,000°C in milliseconds for single-step separation
HPAL	High-Pressure Acid Leach — nickel laterite processing at 225–270°C
HPMS/HDDR	Hydrogen Processing of Magnet Scrap / Hydrogenation-Disproportionation-Desorption-Recombination — magnet-to-magnet recycling

Subsectors

The critical minerals sector is not one market. It is a stack of distinct sub-areas, each with its own technology, its own set of players, and its own investment angle. The recurring theme across all of them: mining is not the bottleneck — processing, separation, and magnet-making are. China's grip tightens dramatically as you move downstream (mining ~70% → separation ~90% → magnets ~93-94% → heavy REE refining ~99-100%). What follows enumerates each subsector, including the specific minerals families, the geopolitical control layer, the Western response, the Western refining/processing technologies, and the policy/procurement monitor that tracks it all. Company-specific detail lives on ticker pages — see the wikilinks.

1. Rare Earth Supply Chain (Mine → Magnet)

The flagship subsector. Rare earths are 17 metallic elements (15 lanthanides + scandium + yttrium), split into Light REEs (La, Ce, Pr, Nd, Sm — more abundant, easier to process) and Heavy REEs (Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, Y — scarcer, harder to separate, more valuable). They are not geologically rare; they rarely concentrate into economically extractable ore bodies.

The technology / the chain. Six stages: Mining (feedstock) → Cracking & Leaching → Separation (SX/RapidSX) → Metals Reduction → Magnet Making → OEM integration.

• Feedstock types: monazite (phosphate mineral, from mineral sands — Chemours, Tronox, Iluka), bastnasite (Mountain Pass and Bayan Obo), ion-adsorption clays (southern China + Myanmar, HREE-rich).
• Cracking & leaching: monazite dissolved in NaOH or H2SO4 at elevated temperature, producing mixed rare earth carbonate; radioactive thorium must be managed.
• Separation — the critical bottleneck. REEs have nearly identical ionic radii and chemistry; separating them "is like sorting 17 shades of white paint." Conventional solvent extraction (SX) requires 100-300+ mixer-settler stages (some sources cite hundreds to thousands). China has perfected this over 30+ years and holds roughly 25,000 rare earth patents. Output: individual high-purity oxides (99.5-99.99%).
• Metals reduction: oxides → metal via molten salt electrolysis (energy-intensive); only scaled Western metal producer is Less Common Metals (UK, acquired by USA Rare Earth). China controls ~90% of metal production.
• Magnet making — the ultimate chokepoint. NdFeB alloy → strip casting → hydrogen decrepitation → jet milling (3-5 μm) → field pressing → sintering (1,000-1,100°C) → machining → coating → magnetization. A modern sintered NdFeB plant costs $300-500M. China controls 94% of sintered NdFeB (up from ~50% two decades ago). The physics: NdFeB hits BHmax ~50-52 MGOe (~10x ferrite); adding Dy or Tb raises Curie temperature for high-temp use (EV motors, jet engines).

Who plays. Western-aligned miners: MP (Mountain Pass, CA — only active US REE mine, ~42,400 MT, ships concentrate to China until Phase II), LYC.AX Lynas (Mt Weld WA, ~14,000-19,000 MT REO/yr, largest non-China producer), ILU.AX Iluka (Eneabba refinery), UUUU Energy Fuels (White Mesa monazite-to-oxide). China-aligned: China Northern Rare Earth (600111.SS, Bayan Obo, ~83% of China's light REE, ~40% global), China Rare Earth Group (~60-70% of China's heavy REE), Shenghe Resources (separator/trader). Separators outside China: LAMP Malaysia (Lynas, world's largest ex-China at 10,500+ t/y NdPr), Mountain Pass, Kalgoorlie (Lynas C&L), Sillamäe Estonia (Neo Performance). Magnet makers: Chinese (JL MAG 38,000 MT/yr targeting 60,000, Beijing Zhong Ke San Huan, Ningbo Yunsheng); Western emerging (MP Fort Worth, USA Rare Earth USAR Stillwater OK 10,000 MT target, VAC Vacuumschmelze Sumter SC, Neo Performance Narva Estonia). The capacity gap is stark: China ~300,000+ tonnes/yr NdFeB vs the entire non-Chinese world <15,000 tonnes — a 20:1 ratio.

Product economics. NdFeB magnets ASP $50-200/kg; NdPr oxide ~$60-80/kg; Dy oxide ~$250-400/kg; Tb oxide ~$800-1,200/kg. Value capture by stage (gross margins): mining 15-35%, metal production 5-15%, magnet manufacturing 8-18% (compressed by Chinese overcapacity — JL MAG margins fell from ~10% in 2022 to ~3% in 2024), motor/component 20-40% (highest, defense premiums). Global NdFeB magnet market ~$16-20B (2024), projected $28-59B by 2030-2034.

Investment angle. The midstream (separation + magnets) has the deepest moats and is where China's dominance is most entrenched, so the Western strategy is to build parallel midstream capacity. The processing bottleneck matters more than mining: "The world has enough ore; it doesn't have enough non-Chinese processing." Two parallel Western REE chains are forming — Chain 1 (Energy Fuels monazite pathway) and Chain 2 (ReElement mixed-carbonate/recycling pathway) — both converging at Vulcan Elements for magnet manufacturing. Heavy rare earths (Dy, Tb) remain the acute vulnerability; true non-China HREE magnet supply-demand balance is unlikely before 2028-2030.

2. Antimony, Tungsten & Other Strategic / Defense Minerals

Beyond rare earths sits a cluster of individually critical, often defense-irreplaceable minerals, each China-dominated and each now a US procurement priority. The recurring logic: certain electron configurations and atomic properties produce magnetic, optical, or structural characteristics no engineering workaround can replicate — "the periodic table does not negotiate with geopolitics."

Antimony (atomic 51). ~48-56% China mine output (down from 75-83% in 2011-2015) but dominates refining; global production ~83,000 t/yr. Uses: lead-antimony ammunition hardener (1-5% by weight; US ammo consumes 600+ tons/yr), antimony trioxide (Sb2O3) flame-retardant synergist, lead-acid batteries, emerging liquid-metal batteries. Under Chinese export licensing since September 2024; antimony trioxide surged 228% since January 2024 to ~$39,000/t. The only US smelter: UAMY US Antimony (Thompson Falls, Montana — won a $245M IDIQ sole-source DLA contract, ~17x its 2024 revenue of $14.9M). America's only antimony reserve: PPTA Perpetua (Stibnite, Idaho — 148M lbs proven/probable antimony + 4.8M oz gold; the site supplied 90% of US antimony in WWII). Only producing antimony mine: USAS Americas Gold & Silver (Galena Complex, Idaho).

Tungsten (atomic 74) — "the hardest chokepoint." Highest melting point of any metal (3,422°C); WC hardness 9-9.5 Mohs; density 19.25 g/cm³ (2.3x steel). Uses: cutting tools, drill bits, APFSDS armor-piercing penetrators, bunker-busters. China produces ~80%+ (~67,000 MT in 2024); the US has had no commercial tungsten production since 2015. Export controls imposed Feb 2025 spiked APT prices 20%+. Non-Chinese sources: Vietnam (Masan Nui Phao), Austria (Wolfram Mittersill), South Korea (Almonty Sangdong, restarting). US development: American Tungsten (TUNG, IMA Mine Idaho); Kazakhstan friend-shoring (QazMoly ~$240M US-funded project, Severniy Katpar via DFC up to $700M).

Gallium & Germanium — the "byproduct trap." Not mined directly: gallium is a byproduct of alumina/Bayer-process bauxite (~50 ppm Ga), germanium of zinc smelting. Output cannot scale independently of the primary metal, which is what makes diversification so hard. Gallium (atomic 31): China ~98%; GaN (3.4 eV wide bandgap) enables radar AESA modules impossible with silicon (40-80 V/µm breakdown, 400°C operation, 5-10x power density). Germanium (atomic 32): China ~60-68%; 0.67 eV bandgap creates a natural IR long-pass filter for night vision/FLIR/missile guidance (>60% of mid-low IR lens elements). Both under Chinese export controls since July/August 2023; December 2024 complete ban to the US; shipments hit zero. Teck Resources (Canada) is the largest non-Chinese germanium producer (~20 t/yr).

Samarium, Dysprosium, Terbium (defense heavy/specialty REEs). Samarium (atomic 62): SmCo magnets survive 350°C continuously / 550°C excursions vs NdFeB's 150°C ceiling — the only magnets that survive missile flight; F-35 carries ~50 lbs of SmCo. Dy (atomic 66) and Tb (atomic 65) added at 7-10% by weight to NdFeB form the Dy2Fe14B grain-boundary phase that stops EV and fighter motors self-destructing above 150°C; a single-motor EV needs ~100g Dy + ~9g Tb. China controls 99% of Dy/Tb processing. F-35 = 400-920 kg total rare earths; Virginia-class submarine 4,200+ kg.

Other tracked strategic minerals. Scandium (aerospace alloys; NioCorp $10M DPA award, RZ Resources, Sunrise Energy Metals), fluorspar (Ares Strategic Mining ARS.V won $168.9M IDIQ), indium, bismuth, vanadium, manganese, tellurium (Rio Tinto Kennecott ~20 t/yr, ~3% global), niobium, PGMs (Pt/Pd/Rh/Ir/Ru/Os — South Africa's Bushveld holds ~70% Pt and ~80% Rh reserves), cobalt (DRC ~73-74%), nickel (sulfide vs HPAL-laterite), graphite (China ~90%+ anode processing), copper (structural deficit), uranium, lithium. The USGS 2025 Final List (Nov 6, 2025) expanded to 60 minerals (50 from 2022 + boron, copper, lead, met coal, phosphate, potash, rhenium, silicon, silver, uranium); the EU CRMA designates 34 CRMs + 17 SRMs.

Investment angle. These are mostly single-mineral, government-demand-driven stories where a DLA/DPA contract or a price floor is the re-rating trigger, not commodity price. Defense pull-demand (US DoD's January 2027 requirement that all defense-contractor magnets be non-China) and the inability to substitute make the supply-security premium durable. Avoid Indonesian nickel pure-plays (structural HPAL oversupply) and cobalt pure-plays (cathodes de-cobalting).

3. China Export Controls & Coercion

This is less a market than a forcing function — the geopolitical layer that creates urgency and a price floor under everything else. China's mineral dominance is treated explicitly as a strategic weapon (Deng Xiaoping, 1992: "The Middle East has oil, China has rare earths").

The escalation timeline. July 2023 gallium/germanium licensing → Oct 2023 graphite restrictions → Dec 2024 complete ban to US on gallium, germanium, antimony, superhard materials → April 4, 2025 controls on 7 heavy REEs + all compounds/metals/magnets → Oct 9, 2025 expanded to "parts/components/assemblies" containing Chinese REEs → Dec 1, 2025 extraterritorial reach to "internationally made" products using Chinese REEs → Jan 1, 2026 restrictions on unprocessed rare earths. A temporary 1-year pause on the Oct 2025 controls runs until November 2026 — explicitly a "temporary adjustment," with the framework intact and reactivatable on minimal notice.

Impact. European NdFeB magnet prices hit 6x Chinese domestic prices during April-May 2025; automakers cut production. The dual-use export control list contains over 1,100 items across 7+ categories. In 2022, F-35 production was temporarily halted when a Chinese-origin magnet was found in the supply chain.

The Japan coercion case (January 2026) — the clearest weaponization since the 2010 Senkaku embargo. On January 6, 2026 China's MOFCOM banned dual-use exports to Japan; two days later it halted rare earth and magnet shipments to Japanese firms. Trigger: PM Sanae Takaichi's November 2025 statement that a Chinese attack on Taiwan could be a "survival-threatening situation" for Japan. Japan sources ~60-63% of total REE imports from China but approaches 100% dependence on Chinese heavy rare earths (Dy, Tb). Nomura estimated a 3-month restriction costs Japan ~¥660 billion ($4.2B) and 0.11% GDP; a full year ~¥2.6 trillion and 0.43% GDP.

The strategic read for investors. Each coercion episode accelerates rather than delays Western diversification: "In 2010, the alternatives were embryonic. In 2026, they are real, funded, and already being built." Realistic projections still have China holding 75%+ market share through 2030; breaking dominance takes a decade-plus even at maximum effort.

4. Western Refining & Processing Technology (incl. Flash Joule Heating)

The subsector with the most genuine technological differentiation, because the whole Western bet hinges on building midstream capacity without 30 years of accumulated Chinese SX know-how. Several distinct process technologies compete.

Flash Joule Heating (FJH). Developed in Dr. James Tour's lab at Rice University (seminal Nature paper Jan 2020; key patent US 12,054,391 B2, granted Aug 6, 2024), licensed exclusively worldwide to Metallium Ltd (MTM / OTCQX: MTMCF, an Australian-listed company operating in the US via Flash Metals USA). FJH passes pulsed/direct current through resistive material, hitting >3,000°C in milliseconds. Combined with chlorination (FJH-Cl2), transition metals form volatile chlorides at ~1,000°C and boil off while REE oxides react above ~1,350°C to oxychlorides — single-step separation at >90% purity and >90% yield in seconds vs hours for acid leaching. LCA data (Xu et al. 2025, PNAS): 87% less energy, 84% fewer GHG emissions, 54% lower operating cost, ~100% water and acid elimination. First commercial site: Gator Point Technology Campus, Chambers County (Anahuac) Texas, Q1 2026 commissioning; additional permitted sites in Westport MA and Harrisonburg VA. FJH also feeds UCU.V Ucore's RapidSX (Metallium converts RE-bearing materials into chloride intermediates) and de-bottlenecks red mud processing (boosts ElementUS extraction from 40-70% to >80%; documented +72% Nd, +103% Sc, +514% Ti recovery).

Chromatographic separation (LAD). AREC ReElement Technologies uses Ligand Assisted Displacement chromatography — feedstock through a resin-filled column — claiming 100x the efficiency of SX with 80% less waste and a far smaller footprint, consistently 99.5%+ purity. Feedstock-flexible (mined concentrates, mixed carbonates, recycled magnets, e-waste, C5 concentrates). Noblesville IN scaling to 250+ MT/yr; Marion IN "Supersite."

RapidSX. UCU.V Ucore — column-based SX at least 3x faster than conventional, dramatically reduced footprint. $22.4M US Army Contracting Command facility in Alexandria, Louisiana; DPAS DO-B8 priority; demo to USG targeted H2 2026; CDF in Kingston, Ontario; Bokan Mountain (Alaska) HREE in portfolio.

Conventional SX at White Mesa. UUUU Energy Fuels — the only US facility licensed to handle radioactive monazite/thorium, a genuine dual-use (uranium + REE) infrastructure moat. First US commercial NdPr (2024), first US commercial HREE (Dy/Tb/Sm, July 2025), 99.9% Dy oxide (Aug 2025).

Brownfield "waste-to-minerals" recovery. DOE's brownfield-first thesis (the $275M DE-FOA-0003583 program): recover critical minerals from existing industrial waste streams to cut permitting from 10-15 years (new hard-rock mine) to 3-7 years. Feedstocks:

• Red mud (Bayer bauxite residue): 2.5-4 t per t aluminum, >4 billion t global stockpile; ~2x REE enrichment vs input ore, 500-1,700 ppm REE (scandium = 95% of economic value). ElementUS Minerals (DADA Holdings/Noranda + Enervoxa JV) controls 30M+ dry tons at Gramercy LA; $29.9M DPA grant; $10.1M binding LOI with Metallium for FJH; $850M project commitment.
• Coal ash: Appalachian Basin averages 591 mg/kg REE (highest in US); ~11M tonnes REE in US coal ash (~8x known reserves). Ramaco's Brook Mine (METC, Sheridan WY) — "one of the largest unconventional REE deposits in the US," non-radioactive soft carbonaceous strata; Fluor PEA $5.1B NPV, >150% IRR; potentially the world's only primary source for Ga/Ge/Sc (~70% of revenue). First new US REE mine in 70 years.
• Acid mine drainage: treatment sludge holds ~2,600x more REE than raw drainage; Appalachia could yield ~800 t REE/yr (≈ defense's total need). WVU/Core Natural Resources Mount Storm pilot.
• Copper/zinc smelter byproducts: Rio Tinto Kennecott (RIO, Utah) — one of two US copper smelters, recovers 10 products incl. tellurium since 2022; Ga/Ge recovery in research with Missouri S&T.
• Phosphogypsum: Mosaic/Rainbow Rare Earths Uberaba (Brazil).

Why an Australian company holds US tech (the structural arbitrage). Metallium's hybrid structure (ASX listing + US subsidiary) accesses more capital and government support than any purely domestic config: ASX is the world's #2 junior-mining financing source; Australia is a DPA Title III "domestic source," so the entity can tap both Australian (A$5B Critical Minerals Facility, A$1.2B Strategic Reserve, NAIF, NRF) and US defense funding simultaneously; the Oct 2025 US-Australia Framework adds an $8.5B pipeline. Rice runs a portfolio approach — different FJH applications licensed to specialized entities (Metallium for metals, Universal Matter for graphene, Environmental Clean Technologies for PFAS, Flash Hydrogen for batteries).

Investment angle. Process technology is the highest-differentiation, highest-execution-risk slice. The Ramaco-NETL CRADA model (sustained national-lab partnership → patents, validation, grant wins) is the template the ElementUS-Metallium-Rice consortium replicated via exclusive licensing. Catalyst: DOE award announcements move micro/small-caps 20-50% in a day.

5. US Policy & Strategy

The demand-and-financing engine. Not a market subsector but the architecture that underwrites every Western producer's economics.

Stockpile & demand. Project Vault (Feb 2, 2026) — $12B strategic reserve ($10B EXIM direct loan, largest in EXIM history, + ~$2B private), storing all 60+ USGS-listed minerals across US warehouses. Mechanism: manufacturers buy at fixed prices with repurchase agreements, drawing down during disruptions. Partners: GM, Stellantis, Boeing, Alphabet/Google, Corning, GE Vernova, Clarios, Western Digital; traders Hartree Partners, Traxys, Mercuria.

Trading bloc & price floors. FORGE (Forum on Resource Geostrategic Engagement, Feb 4, 2026) — successor to the Minerals Security Partnership; 54-56 nations; preferential trading bloc with coordinated price floors; chaired by South Korea through June 2026. Launched at the inaugural Critical Minerals Ministerial (Feb 4, 2026, State Dept, VP Vance + Sec. Rubio, 50+ nations), which circulated a nonbinding Framework Agreement on Cooperation on Critical Minerals Sourcing and Processing (signatories identify priority projects within 6 months, expedite permitting, run national-security asset-sale reviews). The Pax Silica Declaration (Dec 2025) links minerals to AI/semiconductor security (US, Australia, Japan members; Canada, Taiwan, EU observers; India expected).

Financing operators. DoD / Office of Strategic Capital — "One Big Beautiful Bill Act" (July 4, 2025): $2B FY2025 National Defense Stockpile + $5B Industrial Base Fund + $500M OSC credit, can now take equity. Deployments: $620M Vulcan, $400M MP equity + $150M MP loan (with $110/kg NdPr price floor + 100% offtake of 7,000 MT/yr for 10 years), $80M ReElement, $90M Albemarle, $22.4M Ucore. EXIM ($100B authority): $14.8B+ in LOIs incl. $10B Project Vault, $600M Tronox (with EFA), $2.2B to 7 Australian projects, $120M Critical Metals Corp (Greenland). DFC: $600M Orion Consortium (with UAE ADQ, →$5B target), $565M Serra Verde (Brazil), $75M Ukraine, exploring $700M Kazakhstan tungsten. DOE: $1.4B EnergySource, $225M Standard Lithium, plus the funding pipeline (see §6). CHIPS Act: $50M Vulcan.

Trade & legislative. Section 232 proclamation (Jan 2026) on processed critical minerals: no immediate tariffs but 180-day negotiation deadline (July 13, 2026), tariffs + minimum import price floors reserved — a binary catalyst creating a domestic price floor either way. A Jan 15, 2026 EO authorized price support and contemplated tariff remedies if talks fail within 180 days. Section 232 investigation (Jan 2026) confirmed the US is 100% import-reliant for 12 minerals, 50%+ for 29 more. H.R. 4090 (Critical Mineral Dominance Act, passed House Feb 4) fast-tracks federal-land mining; S. 2860 (Offshore), S. 789, S. 2550 in the Senate. IRA: $7,500 EV credit tied to FTA-partner mineral sourcing (2026 threshold 70%), with the FEOC rule barring any China-processed mineral content — the core economic driver forcing China-bypass chains.

The inter-country architecture. US-anchored, built relationship-by-relationship: US↔Australia (feedstock-processing axis, $1B-each Framework, $8.5B pipeline), US↔Japan (processing-to-magnet, 24-hour Rapid Response Group), Australia↔Japan (oldest non-China REE partnership via JOGMEC/Sojitz), US↔South Korea (POSCO magnet qualification bridge), US↔UAE/Saudi (capital + Maaden processing JV), US↔C5 (Central Asia frontier — ReElement-TMK Uzbekistan tungsten, REalloys-AltynGroup Kazakhstan REE). Steve Zissouu's (Meridian Report) trilateral thesis routes Australia (feedstock) → USA (refining) → Japan/Korea (magnets), deliberately routing around Malaysia, which he frames as political risk.

Investment angle. Government support is determinative — Western projects can't beat subsidized Chinese production without sustained policy backing; the DoD-MP template ($400M equity, $150M loan, $110/kg floor) is the model. Multi-sovereign validation = reduced risk: projects attracting 3+ sovereign financiers (Arafura, with 7+ agencies across 5+ countries) are being designated as Western supply-chain nodes.

6. Minerals Procurement Monitor

The operational tracking discipline that converts the policy firehose into actionable, time-stamped signals — distinct from the strategy narrative because it is a repeatable scanning process with a signal taxonomy and dated catalysts.

What it is. A daily/weekly scan across Federal Register, SAM.gov, DOE EERE Exchange / Grants.gov, DOD Business Defense / DLA, EXIM, ARPA-E, Congress committees, White House, State Dept, and trade-press funding trackers (Inside Government Contracts, InvestorNews, The Meridian Report). Signals are color-coded GREEN (approved capital / contract), AMBER (framework/LOI/negotiation), RED (early RFI/proposed rule).

The signal taxonomy and lead times. RFI (govt sizing market, 12-18 mo) → FOA/NOFO/BAA (budget committed, 6-12 mo) → EXIM LOI (project deemed viable, 6-12 mo) → application deadline (3-6 mo) → award announcement (0-6 mo) → IDIQ contract (revenue imminent) → Section 232 (tariffs likely, 6-12 mo). The conversion LOI → final commitment and RFI → IDIQ are the re-rating triggers.

Live trackers maintained. (a) EXIM LOI cumulative tracker — Graphite One GPHOF $2.07B, American Battery Tech ABAT $900M, American Rare Earths ARR $456M, the 7 Australian projects under the EXIM-EFA "Single Point of Entry," totaling $14.8B+. (b) DOE funding pipeline (~$1.08B pending): $134M REE Supply Chains (CMEI, apps closed Jan 5), $275M Mines & Metals Capacity Expansion (DE-FOA-0003583), $80M Mine of the Future, $500M MESC battery materials (NOI), $50M Critical Minerals Accelerator (AMMTO), $40M ARPA-E RECOVER, plus ARPA-E ROCKS (ore characterization). (c) DLA/NDS stockpile RFI tracker — fluorspar ($168.9M IDIQ awarded), scandium (→ IDIQ + NioCorp $10M DPA), indium (IDIQ), with tungsten, graphite, dysprosium, terbium, samarium, bismuth, vanadium under evaluation/active. (d) Key minerals watchlist with rationale and signal sources.

Dated catalysts to watch. Apr 21, 2026 (Forest Service Locatable Minerals comment deadline); Apr 27, 2026 (BOEM OCS offshore minerals comment close); July 13, 2026 (Section 232 180-day status report); Q1-Q2 2026 (DOE award announcements); June 2026 (FORGE chair rotation from South Korea); Nov 2026 (China export-control pause expiration); Q4 2026 (Energy Fuels commercial HREE target); late 2026/early 2027 (DLA $2B OBBA stockpile deployment).

Investment angle. This is the early-warning layer: government RFIs and LOIs precede revenue by 6-18 months and pre-select the eventual winners. DOE/DLA award-day announcements are the most actionable single events for the micro/small-caps in this sector. The monitor also flags the binary near-term catalysts (Section 232 July deadline, Nov 2026 control-pause expiry) that would re-rate the entire sector.

Value chain

The critical minerals value chain is a pyramid: low-value raw ore at the base, escalating value-add toward magnets and finished components at the top. The single most important fact for an investor is that the bottleneck is not mining — it is processing and magnet manufacturing. The world has enough ore. It does not have enough non-Chinese midstream. Companies building processing/separation/magnet capacity (Energy Fuels, ReElement, Lynas, Vulcan) are more strategically valuable than pure miners, and the Western supply-chain strategy is explicitly an effort to build parallel midstream capacity, not new mines.

The five-stage map (mine to magnet)

The rare earth chain runs feedstock → cracking/leaching → separation → metals reduction → magnet making → OEM. As prose:

1. Mining (feedstock). Monazite from mineral sands (Chemours, Tronox, Iluka); bastnasite at Mountain Pass (MP Materials) and Bayan Obo (China); ion-adsorption clays in southern China and Myanmar (HREE-rich, low-cost surface mining, high environmental impact). Hard-rock mining is standard drill-blast-load-haul, capital-intensive but established. Output: raw ore or mixed rare earth concentrate with all 17 REEs jumbled together.
2. Cracking & leaching. Monazite dissolved in NaOH or H2SO4 at elevated temperature → mixed rare earth solution/carbonate (MREC). Radioactive thorium must be managed. Energy Fuels performs this at White Mesa Mill — the only US facility licensed for radioactive monazite processing.
3. Separation. The critical bottleneck. Mixed RE solution fed through 100–300+ (up to thousands of) mixer-settler stages with organic solvents; each stage preferentially extracts one REE. Separating 17 chemically near-identical elements is "like sorting 17 shades of white paint." Output: individual high-purity oxides (99.5–99.99%). China holds ~90% of capacity and roughly 25,000 rare earth patents built over 30+ years.
4. Metals reduction. Oxides → metals via molten-salt electrolysis or metallothermic reduction. Energy-intensive. China controls ~90% of global rare earth metal production. The only scaled Western metal producer is Less Common Metals (UK), acquired by USA Rare Earth.
5. Magnet manufacturing. NdFeB alloy → strip casting → hydrogen decrepitation → jet milling (3–5 μm) → magnetic-field pressing → sintering (1,000–1,100°C, under vacuum) → machining → coating → magnetization. A modern sintered-NdFeB plant costs $300–500M to build. China controls 94% of sintered NdFeB production, up from ~50% two decades ago.

Then OEM integration: magnets into EV traction motors, wind turbines, defense systems (F-35, submarines, munitions), industrial robots, MRI machines, consumer electronics.

For lithium, the two paths diverge at the feedstock: hard-rock spodumene (mine → crush → flotation to SC6 6–7% Li2O concentrate, 70–85% recovery → calcination → acid roast → leach → purify → Li2CO3/LiOH) versus brine (pump → solar evaporation 13–24 months → impurity removal → Na2CO3 precipitation → Li2CO3, only 20–40% recovery, ~5 km² land footprint per 25,000 tpa LCE). DLE (direct lithium extraction) compresses the brine route to hours at 80–95% recovery and is the disruption to watch.

Three layers — pyramid framing (upstream/midstream/downstream)

The sources also frame the chain in three layers, with a kitchen/farm/restaurant analogy: upstream is farming (find the soil, harvest the crop, deliver raw ore/mixed concentrate); midstream is the kitchen (wash, sort by exact variety, cut to spec — separate all 17 elements that look nearly identical); downstream is the gourmet restaurant (transform prepared ingredients into finished dishes commanding premium prices, where value added per step rises dramatically).

• Upstream — exploration, mining, extraction. Concentration/beneficiation produces a Mixed Rare Earth Concentrate at typically 50–70% rare earth oxide content. Gross margins highly cyclical: Lynas hit 62% at the 2022 peak but fell to 19–28% in FY2024–25 amid depressed NdPr prices; historical median for integrated producers ~35%. Chinese cost advantage runs to $11/kg REO for the lowest-cost producers, making unsubsidized Western competition difficult.
• Midstream — concentration, separation, refining, oxide production. Slim margins, capital-intensive — CSIS calls it "a high-cost endeavor with slim profit margins." MP Materials reported negative EBITDA in its materials segment during midstream ramp (Q4 2024). Estimated gross margin range 10–25%, improving with vertical integration. This is where Chinese dominance is most entrenched.
• Downstream — metal, alloy, magnet, component. Highest-barrier. Alloy types: NdFeB (~30% Nd / 65% Fe / 5% B) for EVs, wind, electronics; SmCo (samarium + cobalt) for aerospace/defense high-temp. Component integration (rotors, stators, power electronics, control systems) carries the highest value-add and defense premiums.

China's grip tightens as you move downstream

The defining structural fact, expressed as a single chain:

Mining (70%) → Processing (90%) → Magnets (93–94%) → Heavy REE (99–100%)

Stage	China market share
Mining	~60–70%
Separation / processing	~87–91%
Heavy rare earth refining	99–100%
NdFeB magnet manufacturing	92–94%
Gallium (refined)	98–99%
Germanium	~68%
Antimony	~48–56%
Battery-grade graphite	~90–98%
Lithium (refining)	~65%

China dominates mining at ~60–70% of global output (~270,000 MT REO in 2024), but that understates the control problem — the midstream is the severe bottleneck. The US exports 95%+ of domestically mined rare earths to Asia for processing because no commercial-scale domestic separation existed; the US is 100% import-reliant for 12–15 critical minerals (Section 232, Jan 2026: 100% reliant on 12, over 50% reliant on 29 more). Antimony has not been mined domestically since 2001.

Per-stage economics and barriers to entry

Two complementary cuts at where margin pools and moats sit.

From the primer (industry-wide revenue pools and gross margins):

Layer	Revenue pool	Gross margin	Consolidation	Barriers to entry
Mining	~$150B+	20–50% (cycle-dependent)	Moderate	Capital ($500M–5B), permits (5–15 yr), geology
Processing (cracking/leaching)	~$40–60B	25–35%	Very high (China 70–90%)	Chemistry, environmental permits, radioactive handling
Separation/refining	~$30–50B	30–50%	Very high (China 90%+ REEs)	SX IP, 100–300+ stages, qualification
Magnets/cathodes/cells	~$100–120B	15–25%	High (China 94% magnets)	Customer qualification, IP, capital
End products (OEMs)	Trillions	Varies	Fragmented	Brand, engineering, scale

From the China-coercion analysis (a more granular value-capture cut by stage, with magnet margins compressed by Chinese overcapacity):

Stage	Gross margin range	Value capture
Mining	15–35%	Commodity, cyclical
Metal production	5–15%	Capital-intensive
Magnet manufacturing	8–18%	Currently compressed by Chinese overcapacity
Motor/component manufacturing	20–40%	Highest value-add; defense premiums

The two cuts disagree slightly on where the fattest margin sits (the primer flags separation at 30–50%; the coercion cut flags downstream motor/component at 20–40% and notes magnet-making margins are currently thin, not fat). The reconciliation: separation has the deepest moat and best structural margin, but magnet-making is the ultimate physical chokepoint even though Chinese price wars have crushed its current profitability. Chinese magnet maker JL MAG saw margins compress from ~10% (mid-2022) to ~3% (2024) on overcapacity — a structural opening for Western producers with government price floors and captive defense demand.

Barriers to entry, highest to lowest:

1. Magnet manufacturing — complex IP (Hitachi patents), $300–500M capex, decades of process know-how, multi-year customer qualification cycles.
2. Metal/alloy production — technical complexity, economies of scale, environmental permitting.
3. Separation/refining — ~80% of mine-to-metal capital cost sits here, specialized chemical engineering, hazardous processes.
4. Mining — lowest barriers; REEs are geologically abundant; the binding constraint is regulatory/permitting (10–15 years in the West).

The magnet capacity gap — the 20:1 ratio

This is the single starkest number in the whole chain. China produces ~300,000+ tonnes/year of NdFeB magnets (one source: ~240,000 t at the 94% share); the entire non-Chinese world produces under 15,000 tonnes — roughly a 20:1 ratio. Japan makes ~4,500 tpa (about 60% of its own 7,500-tonne need); the rest of the non-China world is nascent.

Chinese magnet makers (dominant):

Company	Capacity	Notes
JL MAG Rare-Earth	38,000 MT/yr (targeting 60,000)	Tesla, BYD supplier
Beijing Zhong Ke San Huan	12,000+ MT/yr	Technology leader
Ningbo Yunsheng	15,000+ MT/yr	Vertically integrated

Non-Chinese magnet nodes (emerging/established):

Location	Operator	Capacity target	Status
Benson / RTP, NC (US)	Vulcan Elements	10,000 MT/yr	Grand opening Nov 2025; scaling. See VULCAN (private)
Fort Worth, TX (US)	MP Materials	10,000 MT/yr target	Phase III underway; DoD 100% offtake of 7,000 MT/yr for 10 yr. See MP
Stillwater, OK (US)	USA Rare Earth	10,000 MT/yr target	310,000 sq ft facility; Q1 2026 commissioning. See USAR
Sumter, SC (US)	VAC Vacuumschmelze	~2,000 MT/yr	Late 2025
Estonia (Narva/Sillamäe)	Neo Performance	2,000 → 5,000+ MT/yr	Opened Sep 2025; multi-year Bosch MOU
Japan (6 locations)	Shin-Etsu Chemical	~3,400 tpa sintered	Established; supplies Toyota, Subaru, Suzuki
Japan	Proterial (fmr. Hitachi Metals)	2,500–3,000 tpa	Supplies Toyota, Lexus, Honda, Nissan, Mazda
Japan	TDK	~800 tpa domestic	Established
Kuantan (Malaysia)	Lynas–JS Link	3,000 t/yr NdFeB	RM600M (~$143–145M); near LAMP
Yesan (South Korea)	JS Link	Operational scale	Plant nearing completion
South Korea	Kangwon Energy / Noveon	2,000 MT/yr	Production target 2027; Noveon EcoFlux tech

Magnet manufacturing is "the ultimate chokepoint." The open investor question (Steve's): 10,000 tonnes of Vulcan capacity is ~1% of Chinese capacity — is that enough? China spent 30 years building this. Realistic projections have China holding 75%+ market share through 2030, and domestically-produced non-China heavy-rare-earth-magnet supply-demand balance is unlikely before 2028, possibly not until 2030 or later. Magnet market size: ~$16–20B (2024), projected $28–59B by 2030–2034.

Bill-of-materials / per-unit minerals intensity

Where the demand is bolted, by end-product:

• EV battery (single): ~8 kg lithium, ~35 kg nickel, ~14 kg cobalt, ~65 kg graphite.
• EV traction motor (single-motor EV): ~100 g dysprosium and ~9 g terbium added to the NdFeB magnet (Dy/Tb run 7–10% by weight of the magnet, substituting for Nd at grain boundaries to raise coercivity/Curie temperature).
• F-35 Joint Strike Fighter: ~920 lbs total rare earth materials per aircraft (one source cites 400–920 kg), including ~50 lbs of samarium-cobalt magnets; SmCo holds magnetism to 350°C continuous (vs 150°C for NdFeB) for missile nose cones and turbomachine pumps. 400+ kg distributed across guidance, fin actuators, radar.
• Virginia-class submarine: 4,200+ kg rare earth content (propulsion, sonar, weapons).
• AESA radar (e.g. AN/APG-85, SPY-6): 300+ transmit/receive modules, each a high-power GaN amplifier (gallium nitride, 3.4 eV bandgap, 200W output).
• US ammunition: >600 tons of antimony annually (1–5% by weight in lead-antimony alloys for hardness/penetration).

The "byproduct trap" constrains gallium and germanium specifically: neither is mined directly. Gallium is a byproduct of alumina (Bayer process — bauxite contains ~50 ppm Ga), germanium a byproduct of zinc smelting. Output cannot scale independently of the host metal, so you cannot simply build a gallium mine. Historically only ~1.4% of theoretically available gallium and ~0.7% of germanium were actually extracted. This is why waste-stream recovery (below) matters disproportionately for these two.

Supplier–customer relationships and choke points

The Western chain is built relationship-by-relationship, with each handoff cemented by an offtake agreement, framework, or financial instrument. The physical and contractual flows that already function:

• Australian feedstock → US processing. Australian monazite ships as ~90% concentrate, kiln-dried, in 2,000 kg bulk bags on pallets, out of WA ports (Geraldton, Bunbury) — the same infrastructure used for zircon and rutile. Chemours FL/GA monazite is trucked directly to White Mesa Mill (~2,500 t/yr today, scaling toward 15,000+ t/yr). Donald Project (VIC) will ship concentrate containing 129 t Sm, 16 t Tb, 92 t Dy/yr to White Mesa; Energy Fuels holds 100% monazite/xenotime offtake (up to 10,000 t/yr).
• Lynas corridor (the only chain that runs today end-to-end): Mt Weld (WA) → Fremantle → Kuantan LAMP (Malaysia) separates → Sojitz/JARE imports to Japan under priority offtake (Lynas supplies up to 7,200 tpa NdPr on priority = ~90% of Japan's LRE and ~one-third of total Japanese RE supply; JARE holds rights to 65% of Lynas HRE Dy/Tb) → Shin-Etsu/Proterial/TDK make magnets for Toyota/Honda. LAMP nameplate ~10,500 t/yr NdPr (upgraded from 7,200), targeting 12,000; HRE circuit 1,500 t/yr expanding to 5,000 t/yr (A$180M). Mt Weld ore grade is exceptional at 14.8% vs ~8% in Chinese deposits. See LYC.AX.
• The proof-of-concept handoff: US monazite → US separation (Energy Fuels) → Korean magnet qualification (POSCO). In Sept 2025 POSCO qualified Energy Fuels NdPr oxide for EV/hybrid magnets (1.2 MT NdPr → ~3.0 MT magnets → ~1,500 EVs); Dec 2025 a major Korean automaker qualified Energy Fuels 99.9% Dy oxide. This proved the complete non-China pathway works mechanically. See UUUU.

Choke points / single points of failure:

• Heavy rare earth separation (Dy, Tb). The acute structural gap. China at ~99–100%. Only two cracks: Lynas LAMP (first non-China commercial Dy/Tb, May 2025) and Energy Fuels (99.9% Dy oxide, Aug 2025). Japan's dependence on Chinese HREE approaches 100%.
• Malaysia / LAMP license. The largest non-China separation facility runs on a Malaysian operating license expiring March 2, 2026, with no public renewal decision and a Permanent Disposal Facility ~72% complete (10% behind schedule). A non-renewal devastates non-China supply with no near-term replacement. Steve frames Malaysia as a risk to route around, not an asset; his trilateral thesis (Australia feedstock → US refining → Japan/Korea magnets) deliberately bypasses it — yet the Japanese leg physically runs through Kuantan today.
• Magnet IP and qualification cycles. Hitachi/Proterial patents plus multi-year OEM qualification cycles gate entry even where capital exists.
• The byproduct trap for gallium/germanium (above).

The four parallel Western pathways

The architecture emerging 2025–26 is four redundant feedstock-to-magnet paths, two of which converge on Vulcan Elements:

• Path A — Energy Fuels (monazite): Chemours (FL/GA) + Donald (Australia) + Tronox/Iluka monazite → White Mesa Mill, UT (crack, leach, SX separation) → NdPr/Dy/Tb/Sm oxides → Vulcan Elements + POSCO (magnets) → defense/auto/tech.
• Path B — ReElement (mixed carbonate + recycling): mixed RE carbonates (allied miners / C5 countries) + recycled magnets (ERI) + e-waste → Noblesville/Marion, IN (LAD chromatography) → NdPr/Dy/Tb oxides → Vulcan Elements → same end markets. ReElement's moat is feedstock flexibility — it accepts mined concentrate, mixed carbonate, recycled magnets, e-waste, and magnet manufacturing scrap, positioning it to take feedstock from C5 and emerging miners that produce mixed carbonate but lack separation. See AREC.
• Path C — MP Materials (bastnäsite): Mountain Pass (CA) → on-site separation (Phase II) + Fort Worth, TX metals + magnets (Phase III) → defense/auto. Still ships concentrate to China for separation until Phase II delivers.
• Path D — Ucore (RapidSX): allied feedstocks + Metallium Flash Joule Heating preprocessing → Alexandria, LA (RapidSX column separation) → separated REOs → magnet makers TBD. See UCU.V.

Steve's key structural insight: Energy Fuels and ReElement are complementary, not competitive — they process different feedstocks (monazite vs mixed carbonate/recycled). Two independent processing pathways with different feedstock sources equals resilience: if monazite supply is constrained, ReElement runs on recycled magnets/mixed carbonate; if recycling feedstock is short, Energy Fuels has the mining-to-oxide route. Both chains converge at Vulcan Elements — making Vulcan the critical downstream node.

Competing separation/processing technologies

The midstream is also where technology disruption sits, because conventional solvent extraction (SX) is the slow, dirty, footprint-heavy chokepoint China perfected:

• Conventional SX: 100–300+ (up to thousands of) mixer-settler stages; the incumbent and the moat.
• Ligand-Assisted Displacement (LAD) chromatography (ReElement): resin-filled column; claims 100x efficiency, 80% less waste, far smaller footprint, 99.5%+ purity; feedstock-flexible.
• RapidSX (Ucore): column-based, claimed 3x faster than conventional SX with reduced footprint.
• Flash Joule Heating (Metallium, licensed from Rice University): not a separation tech per se but a pre-treatment that unlocks refractory feedstock. Passes DC through resistive material to >3,000°C in milliseconds; combined with chlorination (FJH-Cl2) achieves single-step >90% purity / >90% yield in seconds vs hours. On red mud it lifts extraction efficiency from 40–70% to over 80% with dilute 0.1 M HCl, with documented gains of +72% Nd, +103% Sc, +514% Ti recovery. Versus hydrometallurgy: 87% less energy, 84% fewer GHG, 54% lower opex, 100% water elimination, ~87% less acid. FJH sits "at the critical juncture between feedstock and extraction." Metallium feeds Ucore's RapidSX (chloride intermediates) and pre-treats ElementUS's red mud.

Recycling / urban mining as a secondary feedstock layer

Recycling is becoming strategic, not merely environmental — it creates feedstock that bypasses Chinese ion-adsorption-clay HREE entirely. Market value of recycled battery metals grew ~11x between 2015 and 2023; IEA projects recycling cuts new-mine needs by 40% for Cu/Co and 25% for Li/Ni by 2050. EU Battery Regulation (by end-2031): 80% Li, 95% Co/Cu/Ni recycled.

Li-ion battery recycling, three routes:

Route	Temperature	Recovers	Li recovery	Advantage
Pyrometallurgical	800–1,200°C	Co, Ni, Cu as mixed alloys	<50% (Li lost to slag)	Low capex, handles mixed feedstock
Hydrometallurgical	Ambient–100°C	Individual metals via SX/precipitation	80–95%	Higher purity, recovers Li
Direct recycling	Moderate	Preserves cathode crystal structure	>95% potential	Lowest energy, fastest, cheapest

NdFeB magnet recycling (HPMS/HDDR) is magnet-to-magnet, dramatically shortening the production route: expose magnets to H2 gas at room temperature → Nd-rich grain boundary absorbs H2 (NdH2, ~3x volume expansion) decrepitating the magnet to powder → heat to ~800°C under H2 (disproportionation) → remove H2 under vacuum (recombination) to reform Nd2Fe14B grains. ReElement + ERI target this pathway for e-waste and end-of-life magnets. Commercial recyclers: Redwood Materials (>95% recovery of Ni/Co/Cu/Al/Li/graphite, $3.5B expansion), Li-Cycle (Spoke & Hub, 95% recovery, 70% lower emissions), Ascend Elements (Hydro-to-Cathode direct), Umicore.

Waste-stream recovery as a feedstock source (brownfield-first, the DOE DE-FOA-0003583 thesis — converts a permitting liability into a 3–7 year path vs 10–15 years for a new hard-rock mine):

• Red mud (Bayer-process bauxite residue): ~2.5–4 t per tonne of aluminum; >4B t global stockpile; REE enrichment ~2x the input ore (500–1,700 ppm REE, scandium holds 95% of the economic value). ElementUS controls 30M+ dry tons at Gramercy, LA.
• Coal ash: Appalachian Basin averages 591 mg/kg REE (highest in the US); ~11M tonnes REE in US coal ash, ~8x known national reserves; heavy REEs preferentially associate with organic matter. Ramaco's Brook Mine (WY) is the first new US rare earth mine in 70 years and potentially the world's only primary source of Ga/Ge/Sc (those three ~70% of revenue). See METC.
• Acid mine drainage: treatment sludge contains ~2,600x more REE than raw drainage; Appalachia could yield ~800 t REE/yr — roughly the entire US defense requirement.
• Copper/zinc smelter byproducts: zinc smelting residues ~0.5 wt% Ge and 0.4 wt% Ga; Rio Tinto Kennecott (one of two US copper smelters) recovers ten products including tellurium (~20 t/yr, ~3% of global supply) and is developing Ga/Ge recovery. See RIO.

Pricing power and the price-floor overlay

Pricing power in this chain has historically belonged to China, exercised through predatory pricing that drove Western processors and magnet makers out of business (China's magnet share rose from ~50% to 94% partly this way). Two forces are now being engineered to give Western producers durable pricing power:

• Price floors / contracts-for-difference. The DoD–MP Materials deal set a $110/kg NdPr price floor (contract-for-difference) plus 100% offtake of 7,000 MT/yr for 10 years — the template. FORGE (54–56 nations, Korea-chaired) and the Section 232 process (180-day deadline July 13, 2026) aim to establish coordinated price floors and/or minimum import prices across the allied bloc. Either outcome creates a guaranteed floor for domestic producers. Australia is building its own A$1.2B stockpile and examining EFA-administered floors.
• The IRA FEOC structural premium. Under IRA Section 30D an eligible EV may contain no critical minerals extracted/processed/recycled by a foreign entity of concern (China, Russia, Iran, North Korea). Because ~90% of Australian lithium is currently refined in China, that lithium loses the entire $7,500 credit even though Australia is an FTA partner. This makes IRA-eligible, FEOC-free, non-China material structurally worth more — POSCO's qualification of Energy Fuels oxides created the first real market price for non-Chinese REE in a Western auto chain, and that premium should persist as volume scales.

Proposed 25% tariffs on imported rare earth chlorides would add ~$18/kg to Lynas's Malaysia-to-Texas input cost — a reminder that the same tariff wall protecting domestic processors penalizes allied-but-offshore midstream, reinforcing the onshore-first preference.

Players

The single organizing insight across all the research: the processing bottleneck matters more than the mining bottleneck. The world has enough ore — it does not have enough non-Chinese separation, refining, and magnet manufacturing. China's grip tightens as you move downstream: ~60-70% of rare earth mining, ~90% of separation, ~94% of sintered NdFeB magnets, and ~99-100% of heavy rare earth (Dy/Tb) refining. Companies building midstream capacity (Energy Fuels, ReElement, Lynas) are therefore more strategically valuable than pure miners. The winners cluster around three traits: vertical integration toward mine-to-magnet, government backing (DoD equity, EXIM/EFA financing, price floors), and IRA/FEOC-compliant feedstock. The losers are pure miners with no downstream and projects without sovereign backing.

The DoD-MP Materials deal is the template the whole sector is being repriced against: $400M equity + $150M loan + a $110/kg NdPr price floor (contract-for-difference) + 100% offtake of 7,000 MT/yr from the 10X magnet facility for 10 years.

The two (and four) parallel Western REE chains

Steve Zissouu's (The Meridian Report) framing — repeated as the spine of the deep-dive — is that the Western chain is being built relationship-by-relationship, and that Energy Fuels and ReElement are complementary, not competitive: they process different feedstocks (monazite vs. mixed carbonate/recycled), and both chains converge on Vulcan Elements for magnet manufacturing. The deep-dive maps four parallel feedstock-to-magnet paths:

• Path A — Energy Fuels (monazite): Chemours (FL/GA) + Donald concentrate (Australia) + Tronox/Iluka → White Mesa Mill, UT (crack, leach, SX) → NdPr/Dy/Tb/Sm oxides → Vulcan + POSCO magnets.
• Path B — ReElement (mixed carbonate + recycling): mixed RE carbonates (allied miners/C5) + recycled magnets (ERI) + e-waste → Noblesville/Marion, IN (LAD chromatography) → NdPr/Dy/Tb oxides → Vulcan magnets.
• Path C — MP Materials (bastnäsite): Mountain Pass (CA) → on-site Phase II separation + Fort Worth, TX (metals + magnets, Phase III).
• Path D — Ucore (RapidSX): allied feedstocks + Metallium FJH preprocessing → Alexandria, LA (RapidSX) → separated REOs → magnet makers (TBD).

Rare earth supply chain — industry map

Company	Ticker	Role	Revenue	Market Cap	Moat
China Northern RE	600111.SS	Mining + processing + separation	~$5B	~$12B	State-backed, ~40% global, Bayan Obo (~83% of China's light REE)
Lynas	LYC.AX	Mining + separation	~A$700M	~A$8B	Largest non-Chinese, integrated; only non-China HREE separator at scale
MP Materials	MP	Mining + separation (building)	~$200M	~$3.5B	Only active US REE mine
Energy Fuels	UAMY-adjacent / UUUU	Monazite processing → separated oxides	~$120M	~$2.5B	Only US monazite-to-oxide; dual U+REE
ReElement Tech	AREC	Chromatographic separation + recycling	Pre-rev → $100M target	~$500M	Novel LAD tech, recycling, $1.4B Vulcan deal
Vulcan Elements	Private	Magnet manufacturing	Pre-revenue	N/A	10,000-tonne US magnet facility
Ucore	UCU.V	RapidSX separation	Pre-revenue	~$80M	Novel SX technology
Iluka Resources	ILU.AX	Mineral sands → REE refinery	~A$1.2B	~A$4B	Eneabba refinery development
Tronox	TROX	Mineral sands → REE pivot	~$2.8B	~$3B	$600M EXIM/EFA support

(Note: UUUU is the trading ticker for Energy Fuels; the UAMY wikilink above is to United States Antimony, a separate antimony player covered below. Energy Fuels does not yet have its own ticker page — keep its detail there once created.)

Lithium, copper & diversified — industry map

Company	Ticker	Focus	Revenue	Market Cap
Albemarle	ALB	Lithium (#1 global)	~$5.5B	~$12B
SQM	SQM	Lithium (Atacama brine)	~$4B	~$13B
Pilbara Minerals	PLS.AX	Spodumene lithium	~A$3B	~A$12B
Lithium Americas	LAC	Thacker Pass (40ktpa target)	Pre-revenue	~$2.5B
Freeport-McMoRan	FCX	Copper (Grasberg)	~$25B	~$70B
Glencore	GLEN.L	Cobalt (#1), copper, nickel	~$220B	~$60B

Rare earth players — positioning

Energy Fuels (UUUU) — the Western chain linchpin. America's leading uranium producer AND the only US company commercially producing separated rare earth oxides (NdPr, Dy, Tb) from monazite at the White Mesa Mill, Utah — the only conventional uranium mill in the US licensed to handle radioactive thorium, which is the same capability that enables monazite processing (a genuine dual-use infrastructure moat). Dual revenue: uranium provides cash flow, REE is the growth story. It is the only company physically moving Australian feedstock into US separation (Donald Project JV with Astron, 1.81B-tonne resource, 58+ yr mine life). The investment angle: it touches all three legs of the Australia → USA → Japan/Korea trilateral — US-AU (Donald feedstock), US-KR (POSCO qualified its NdPr oxide for EV magnets, Sep 2025; Dy oxide at 99.9% purity qualified by a major Korean automaker, Dec 2025), and US-JP (Vulcan-to-magnet path). The bear case: REE revenue still small vs. total; processing scale tiny vs. China (38,000 kg NdPr in 2024 vs. China's millions); monazite feedstock constrained until Donald delivers (2027+). Premium valuation (~50x P/S, ~6.2x P/B) reflects growth optionality, not current profit; debt-free until Oct 2025's $700M convertible notes. Keep production-milestone and timeline detail on the Energy Fuels ticker page.

ReElement Technologies / American Resources (AREC) — see AREC. The separation disruptor: uses Ligand Assisted Displacement (LAD) chromatography instead of solvent extraction — claims 100x greater efficiency, 80% less waste, far smaller footprint, consistently 99.5%+ purity oxides. The moat is feedstock flexibility — it can process mined concentrates, mixed RE carbonates, recycled magnets, e-waste, and magnet manufacturing waste, which positions it to accept feedstock from C5 countries and emerging miners that produce mixed carbonates but lack separation. Investment angle: the $1.4B partnership with Vulcan Elements + USG (Nov 2025; $80M OSC loan + $2M earlier DoD), Noblesville IN scaling to 250+ MT/yr, the Marion IN "Supersite," weekly shipments underway, a 2026 order book estimated at $75-100M, plus the ERI e-waste partnership and a POSCO MOU (Sep 2025) for a fully integrated US separation-refining-magnet-recycling complex. High-risk, pre-revenue. Deep company detail lives on AREC.

MP Materials (MP) — Mountain Pass. Only active US rare earth mine. Phase II separation + Phase III magnets (Fort Worth, TX) underway, ramping toward 10,000 MT magnet capacity. The transformational DoD partnership ($400M equity making DoD potentially the largest shareholder + $150M OSC loan + $110/kg NdPr price floor + 10-yr offtake) is the sector's template deal. Also the only US company with a Gulf-state processing JV — the MP-Maaden-DoW JV (Nov 2025; DoW finances the 49% US stake, Maaden holds 51%) targeting Saudi's Jabal Sayid deposit (552,000T HREE + 355,000T LREE, 4th-largest globally). Key risk: still ships concentrate to China for separation until Phase II delivers.

Lynas Rare Earths (LYC.AX) — the proven non-Chinese champion. The only non-Chinese company separating REEs at commercial scale. Mt Weld (WA) at an exceptional 14.8% ore grade (vs. ~8% Chinese deposits), producing ~14,000-19,000+ tonnes REO/yr. Its LAMP plant in Kuantan, Malaysia is the world's largest rare earth separation facility outside China (NdPr capacity upgraded to ~10,500 tpa, "Towards 2030" target 12,000 tpa). May 2025 breakthrough: first-ever commercial production of separated heavy rare earths (Dy oxide, Tb to follow) outside China — A$180M LAMP HRE expansion to 5,000 tpa feedstock. Supplies ~90% of Japan's NdPr and one-third of Japan's total RE supply via the JARE (Sojitz/JOGMEC) channel, which holds rights to 65% of Lynas' HREE. Has a DoD-backed Texas (Seadrift) HREE facility — but CEO Amanda Lacaze said in Aug 2025 it "might not proceed" without acceptable offtake (FEED 78% complete, wastewater permitting unresolved). The two-edged investment angle: strategically irreplaceable, but Steve Zissouu closed his Lynas position citing dependence on Malaysia's political climate ("a risk that I wish they hadn't needed to mingle themselves with") — the LAMP operating license expires March 2, 2026 and is the single biggest near-term risk to the entire non-Chinese supply chain.

Vulcan Elements (private) — the magnet chokepoint. 10,000-tonne US NdFeB magnet facility (Benson/RTP, NC; grand opening Nov 2025). $1.4B partnership stack: DoD $620M + CHIPS Act $50M + ~$550M private. Would be the first Western-scale magnet manufacturer. It is the convergence node where both Energy Fuels' and ReElement's chains terminate, making its scaling the critical variable for the whole thesis (10,000 tonnes ≈ ~1% of Chinese capacity — the open question is whether that is enough).

Ucore (UCU.V) — the RapidSX bet. Column-based REE separation, at least 3x faster than conventional SX with a much smaller footprint. $22.4M US Army Contracting Command agreement to build its first commercial facility in Alexandria, Louisiana; DPAS DO-B8 defense-priority rating (July 2025); US Government demo targeted H2 2026. Bokan Mountain (Alaska) HREE resource in portfolio; CDF in Kingston, Ontario. Strategic tech collaboration with Metallium Limited (ASX: MTM; Sep 2025) — Metallium's Flash Joule Heating converts RE-bearing materials into chloride intermediates that feed RapidSX. Very high risk; thesis is "if validated, adds Western separation capacity." Steve's open question: "How does UCORE figure into the chain?"

Iluka Resources (ILU.AX) & Tronox (TROX) — the mineral-sands optionality. Iluka is building the Eneabba refinery (WA, commissioning ~2027, 23,000 tpa REO target) — would become the second non-Chinese integrated refiner if it delivers. Tronox is pivoting WA mineral sands toward a cracking/leaching facility, backed by $600M EXIM/EFA co-financing (stock jumped 32% on the announcement). Both are watchlist names pending feasibility and financing conversion.

Neo Performance Materials — Europe's separator + magnet maker. REE oxide separation at Sillamäe/Narva, Estonia (one of only ~4 sizable non-Asian REO refining plants globally; "Europe's only commercial facility"). Opened a new European Permanent Magnet facility (Sep 2025), on track for 2,000 MT/yr scaling to 5,000+, with a multi-year MOU with Bosch and a heavy-REE pilot line commissioning early 2026. Canada-domiciled; the under-appreciated G7 processing node.

Australian feedstock pipeline (the upstream that everyone needs). Australia is the feedstock lynchpin appearing in 7+ sovereign-agency financing columns. Beyond Lynas and Iluka:

• Arafura Rare Earths (ARU.AX) — Nolans (NT), ~$1.35B financing assembled, FID target Q1 2026, 4,440 tpa NdPr (~5% of global supply). The clearest "multilateral validation" case: financed by 7+ sovereign agencies across 5+ countries (NAIF A$200M, NRF A$200M, Critical Minerals Facility US$533M, EFA US$100M equity, EXIM up to $300M LOI, JBIC, KEXIM $150M, EDC $300M, German $115M guarantees) plus binding Hyundai/Kia 7-yr offtake.
• Northern Minerals (NTU.AX) — Browns Range HREE; part of the EXIM $2.2B Australian package.
• Energy Fuels/Astron Donald (VIC) — Phase 1 ~7,000-8,000 MT REE concentrate/yr (129t Sm = 250% of US annual needs, 16t Tb = 23%, 92t Dy = 34%), shipping to White Mesa.
• EXIM $2.2B LOI cohort also covers Graphinex (graphite, private), La Trobe Magnesium (LMG.AX), VHM (VHM.AX) (titanium/REE, Goschen, up to $200M), RZ Resources (RZR.AX) (scandium, Copi), Sunrise Energy Metals (SRL.AX) (Ni/Co/Sc, $67M).

China-aligned producers (the incumbents to displace)

Company	Control	Key Asset	Production
China Northern Rare Earth	State-owned	Bayan Obo (world's largest)	~83% of China's light REE
China Rare Earth Group	State-owned	Multiple Southern China	~60-70% of China's heavy REE
Shenghe Resources	SOE-linked	Processing/trading	Major separator

Global magnet manufacturers — the ultimate chokepoint

Magnet manufacturing is the highest-barrier downstream activity: a modern sintered NdFeB plant costs $300-500M, with complex IP (Hitachi patents), decades of process know-how, and multi-year customer qualification cycles. China produces ~300,000+ tonnes/yr of NdFeB; the entire non-Chinese world produces under 15,000 tonnes — a ~20:1 ratio. Even Chinese leaders have seen margins compress (JL MAG from ~10% in mid-2022 to ~3% in 2024) on overcapacity — a structural opening for Western producers with government support and captive defense demand.

Chinese (dominant):

Company	Capacity	Notes
JL MAG Rare-Earth	38,000 MT/yr (targeting 60,000)	Tesla, BYD supplier
Beijing Zhong Ke San Huan	12,000+ MT/yr	Technology leader
Ningbo Yunsheng	15,000+ MT/yr	Vertically integrated

Western / allied (emerging) — magnet manufacturing nodes:

Location	Operator	Capacity Target	Status
Benson/RTP, NC (US)	Vulcan Elements	10,000 MT/yr	Grand opening Nov 2025; scaling
Fort Worth, TX (US)	MP Materials	10,000 MT/yr target	Phase III; 1,000 MT ramping
Stillwater, OK (US)	USA Rare Earth (USAR)	10,000 MT target	310,000 sq ft facility; commissioning H1 2026
Sumter, SC (US)	VAC Vacuumschmelze	~2,000 MT	Late 2025
Narva/Estonia	Neo Performance	2,000 → 5,000+ MT/yr	Opened Sep 2025; Bosch MOU
Kuantan (Malaysia)	Lynas-JS Link JV	3,000 MT/yr	RM600M (~US$143-145M); SE Asia's first
Yesan (South Korea)	JS Link	Operational scale	Plant nearing completion
South Korea	Kangwon Energy / Noveon	2,000 MT/yr	Production target 2027; Noveon EcoFlux
Japan	Proterial (fmr. Hitachi Metals)	2,500-3,000 tpa	Supplies Toyota, Lexus, Honda, Nissan, Mazda
Japan (6 locations)	Shin-Etsu Chemical	~3,400 tpa	Supplies Toyota, Subaru, Suzuki
Japan	TDK Corporation	~800 tpa domestic	Established

Japan totals ~4,500 tpa domestic — the bulk of non-Chinese supply, meeting ~60% of its own 7,500-tonne need. USA Rare Earth (NASDAQ: USAR) is a notable downstream-integrator: CHIPS Act funding, acquired Less Common Metals (UK) — the only scaled Western rare earth metal producer — and is pursuing mine-to-magnet from Round Top, TX.

Lithium, copper & diversified — positioning

Freeport-McMoRan (FCX) — copper structural deficit. Grasberg + US operations; the best pure-play on the "metal of electrification." Copper faces a ~30% supply deficit by 2035 with demand up ~30% to 2040; supply covers only ~80% of 2030 needs. The most favorable supply-demand of any critical mineral, and lower execution risk than the REE developers.

Albemarle (ALB) — lithium cycle recovery. World's largest lithium producer; Tier-1 assets in Chile, Australia, US. $150M DOE + $90M DoD. Margins expanding back toward 25% EBITDA after $450M cost cuts. The cleanest large-cap lithium recovery play.

Pilbara Minerals (PLS.AX) — lowest-cost spodumene. Lowest-cost hard-rock lithium; margins expand disproportionately in a price recovery. SQM (Atacama brine) and Lithium Americas (LAC) (Thacker Pass — largest US lithium deposit, 40ktpa target, mechanical completion late 2027) round out the lithium set. Glencore (GLEN.L) is the diversified cobalt (#1)/copper/nickel major, also flagged for potential mega-merger with Rio Tinto.

Antimony, tungsten & gallium/germanium specialists — see UAMY

The us-critical-minerals-strategy file profiles the "seven priority minerals" specialists (samarium, dysprosium, terbium, gallium, germanium, antimony, tungsten) — many of these are distinct from the REE/lithium names above:

• United States Antimony (UAMY) — see UAMY. Operates the only antimony smelter in the US (Thompson Falls, Montana) + a Madero, Mexico facility. The investment angle is the Sept 2025 $245M IDIQ sole-source DLA contract to supply antimony metal ingots for the National Defense Stockpile — ~17x the company's 2024 revenue of ~$14.9M. Capacity expanding from ~160 to 300+ tonnes/month; resumed mining at Stibnite Hill (Oct 2025); 35,000+ acres of Alaska claims. The bear case is extreme valuation vs. fundamentals (~22x P/S, ~11x P/B against ~$32M TTM revenue and ongoing losses; the 52-week gain ran 417-512%). Full company detail lives on UAMY.
• Perpetua Resources (PPTA) — Stibnite Gold Project, Idaho; holds America's only identified antimony reserve (148M lbs proven/probable) + 4.8M oz gold. Final Record of Decision Jan 2025; raw antimony sulfide targeted late 2028. $75M+ DoD; $1.8-2B EXIM loan indication pending Spring 2026. Pure NAV play (~0.5-0.85x NAV; after-tax NPV $3.7B at $2,900 gold).
• Americas Gold & Silver (USAS) — Galena Complex (Idaho), the only producing US antimony mine; >99% extraction breakthrough (Sep 2025).
• American Tungsten (CSE: TUNG) — IMA Mine, Idaho (past-producing 1945-1957); engaged with DoD on non-dilutive financing. The US has had no commercial tungsten production since 2015.
• Perpetua, US Antimony, American Tungsten together address the China-dominated antimony (~48% mine, dominant refining) and tungsten (~80%+) chokepoints.

Brownfield / waste-to-minerals players (the DOE DE-FOA-0003583 cohort)

A distinct cluster the merge previously dropped: companies extracting critical minerals from industrial waste streams (coal ash, red mud, acid mine drainage) — a brownfield-first approach that can cut a decade off permitting (3-7 yr vs. 10-15 yr for new hard-rock mines).

• Ramaco Resources (METC) — metallurgical coal producer ($666M 2024 revenue) whose Brook Mine near Sheridan, Wyoming is "one of the largest unconventional REE deposits in the US." Could be the world's only primary source of gallium, germanium, and scandium (those three ≈ 70% of potential revenue). Fluor July 2025 PEA: $5.1B NPV, >150% IRR at the 5Mt/yr scenario. Broke ground July 2025 — first new US rare earth mine in 70 years. Buy consensus, $38.69 target, but 23.5% short interest signals skepticism.
• Metallium (ASX: MTM; OTCQX: MTMCF) — Australian-listed licensor of Rice University's Flash Joule Heating (FJH) (Dr. James Tour's lab; patent US 12,054,391 B2). FJH restructures the refractory mineral matrix before acid touches it, lifting red-mud extraction from ~40-70% to >80%, with documented 87% energy reduction, 84% lower GHG, 54% lower opex, zero water. US operations via Flash Metals USA (Gator Point Technology Campus, Chambers County, TX; commissioning Q1 2026). Also the Ucore RapidSX feedstock partner. The "why Australian" logic: ASX capital depth + Australia's DPA "domestic source" status lets it tap both Australian and US government funding.
• ElementUS Minerals (private) — JV of DADA Holdings (owns Noranda Alumina) + Enervoxa; exclusive rights to 30+ million dry tons of red mud at Gramercy, Louisiana (the last operating US alumina refinery; ~30 yrs of feedstock at 1Mt/yr). Targets gallium, scandium, yttrium, neodymium (3,000-4,000 ppm total REE, ~25x crustal). $29.9M DoD DPA Title III grant (Nov 2025) + $850M Louisiana investment commitment + Part II of a $400M DOE LPO application. Binding Dec 2025 LOI with Metallium ($10.1M non-dilutive, $15-40/tonne processing royalty) positions it as the potential first US commercial producer of both gallium and scandium. Mirrors the Ramaco-NETL national-lab partnership model.
• Rio Tinto (RIO) — Kennecott (Utah) is one of only two US copper smelters; recovers ten products including tellurium (~20 t/yr, ~3% of global supply since 2022); gallium/germanium recovery in research with Missouri S&T. Diversified-major exposure (~6.6x EV/EBITDA, ~4.1% yield); potential $200B+ merger with Glencore.
• WVU Water Research Institute / Core Natural Resources — REE from acid mine drainage (Mount Storm pilot, ~2 t/yr); Appalachia could yield ~800 t REE/yr (≈ the defense industry's total requirement).
• Mosaic Company + Rainbow Rare Earths — Uberaba (Brazil) phosphogypsum, 24.5% NdPr content.
• NioCorp — scandium; $10M DPA award (Jan 2026). Graphite One (GPHOF) — graphite (AK + OH), $2.07B EXIM LOI. Ares Strategic Mining (ARS.V) — fluorspar, $168.9M DLA IDIQ (Dec 2025). Critical Metals Corp (CRML) — Tanbreez (Greenland) REE, $120M EXIM LOI + $1.5B Saudi TQB JV term sheet.

Comparative financial profiles (us-critical-minerals-strategy)

Company	Market Cap	TTM Revenue	Profitable?	EV/EBITDA	P/Book	Net Cash/(Debt)
Ramaco (METC)	$1.1-1.8B	$580-666M	No	32.4x	1.95x	+$57M cash
Rio Tinto (RIO)	$131-147B	$53.7B	Yes	6.6x	2.27x	-$5-14B debt
US Antimony (UAMY)	$0.9-1.5B	$32M	No	N/A	11.1x	+$19M cash
Perpetua (PPTA)	$3.0-4.2B	$0	Pre-revenue	N/A	6.06x	+$446M cash
Energy Fuels (UUUU)	$4.0-5.2B	$79M	No	N/A	6.23x	Debt-free

The Australia-aligned upstream miners (china-coercion)

Company	Country	Key Asset	Status
MP Materials	USA	Mountain Pass, CA	Operating; ~42,400 MT
Lynas Rare Earths	Australia	Mount Weld	Operating; ~14,000 MT
Iluka Resources	Australia	Eneabba Refinery	Under construction
Energy Fuels	USA	White Mesa Mill	Operating; uranium + REE
Northern Minerals	Australia	Browns Range	Development

Companies at the trilateral intersection (the highest-conviction set)

The deep-dive's "sit at the intersection of all three legs" table — the names that matter most because they touch multiple bilateral relationships:

Company	Ticker	Intersections	Why it matters
Energy Fuels	UUUU	US-AU (Donald), US-KR (POSCO qual), US-JP (Vulcan-to-magnets)	Only US company physically receiving Australian feedstock AND supplying Korean magnet qualification
Lynas	LYC.AX	AU-JP (JOGMEC, Sojitz), AU-MY (LAMP), AU-US (DoD Texas)	Proven non-Chinese model; ~90% of Japan's NdPr; only non-China HREE separator at scale
Arafura	ARU.AX	AU-JP (JBIC), AU-KR (KEXIM, Hyundai), AU-US (EXIM), AU-CA (EDC), AU-DE	Most multilaterally financed single project — 7+ sovereign agencies
POSCO International	003490.KS	KR-US (EFUELS MOU), KR-AU (graphite via Black Rock), KR-TZ (Mahenge)	The bridge entity that proved the pathway works
MP Materials	MP	US-KSA (Maaden JV), US-DoD ($400M equity)	Only US company with a Gulf-state processing JV; diversifies HREE beyond Australia

Ranked investment opportunities (deep-dive)

Rank	Company	Ticker	Thesis	Risk	Timeframe
1	Energy Fuels	UUUU	Only US monazite-to-oxide; dual U+REE; HREE; Western chain linchpin	Medium	2-5 yr
2	Freeport-McMoRan	FCX	Copper structural deficit; irreplaceable Grasberg	Medium-Low	3-10 yr
3	Lynas	LYC.AX	Largest non-Chinese REE; integrated; indispensable	Medium	3-7 yr
4	Albemarle	ALB	#1 lithium; Tier-1 assets; cycle recovery	Medium	2-5 yr
5	MP Materials	MP	Only US REE mine; Phase II/III integration	Medium-High	3-5 yr
6	SETM ETF	SETM	Broad critical minerals (+105% in 2025)	Low-Medium	1-5 yr
7	Pilbara Minerals	PLS.AX	Lowest-cost spodumene; cycle recovery	Medium	2-4 yr
8	Lithium Americas	LAC	Thacker Pass — largest US lithium deposit	High	3-7 yr
9	American Resources (ReElement)	AREC	ReElement separation + recycling; Vulcan partnership	High	2-5 yr
10	Ucore	UCU.V	RapidSX technology; if validated, adds Western capacity	Very High	3-7 yr

Tiered framework (deep-dive)

• Tier 1 (core holdings): Energy Fuels (UUUU) — most strategically positioned, dual U+REE; Freeport-McMoRan (FCX) — best supply-demand of any mineral; Lynas (LYC.AX) — indispensable non-Chinese REE.
• Tier 2 (tactical): Albemarle (ALB) — lithium recovery; MP Materials (MP) — Mountain Pass integration (real execution risk); Pilbara (PLS.AX) — lowest-cost lithium; SETM (Sprott Critical Materials ETF) — broad exposure.
• Tier 3 (watchlist, with promotion triggers): American Resources/ReElement (AREC) → Tier 2 if $100M revenue AND commercial-scale validation; Lithium Americas (LAC) → if Thacker Pass on timeline AND Li recovers to $15+/kg; Ucore (UCU.V) → if RapidSX validated at scale; Iluka (ILU.AX) → if Eneabba comes online; Tronox (TROX) → if feasibility positive AND $600M financing converts.
• Avoid: Indonesian nickel pure-plays (structural HPAL oversupply); early-stage DLE without DOE backing; cobalt pure-plays (cathodes are designing cobalt out).

Recycling / urban-mining leaders

A secondary supply chain the merge dropped. Market value of recycled battery metals grew ~11x between 2015 and 2023; the EU Battery Regulation mandates 80% Li / 95% Co/Cu/Ni recycled by end-2031.

• Redwood Materials (private) — largest US battery recycler; >95% recovery of Ni/Co/Cu/Al/Li/graphite; $3.5B expansion.
• Li-Cycle — patented Spoke & Hub; 95% recovery; 70% lower emissions.
• Ascend Elements (private) — Hydro-to-Cathode direct recycling; DOE-backed.
• Umicore — European recycling leader.
• ReElement + ERI (AREC) — magnet-to-magnet (HPMS/HDDR) recycling from e-waste and end-of-life magnets, creating Chinese-clay-free HREE feedstock.

DLE / lithium disruptors

• EnergySource Minerals (private) — DLE LiOH from California geothermal brine; $1.4B DOE conditional commitment.
• Standard Lithium (SLI) — DLE in Arkansas with Equinor; $225M DOE grant.
• American Battery Technology (ABAT) — Nevada lithium; $900M EXIM LOI.

Head-to-head / comparison sources to reference

There is no dedicated _compare/ head-to-head file among the sources read, but several documents carry explicit comparative or showdown framing that should be cross-referenced rather than re-hosted here:

• Steve Zissouu / The Meridian Report — "Envisioning the Western Supply Chain" (Nov 2025) and "USA, Australia and Japan: Pondering the Developed Critical Minerals Chain" (Oct 27, 2025): the trilateral Australia → USA → Japan/Korea framework and the Energy-Fuels-vs-ReElement "complementary not competitive" thesis. Also "The Red Zone: Surviving the Defense Mineral Embargo" (Jan 24, 2026) on the DOE $275M grant pivot from resilience to mobilization.
• us-critical-minerals-strategy.md — Part 5 "Financial profiles reveal different risk-return profiles" is a five-way comparison (Ramaco / Rio Tinto / US Antimony / Perpetua / Energy Fuels) across market cap, revenue, profitability, EV/EBITDA, P/Book, net cash.
• rice-flash-joule-heating-doe-grant.md — a head-to-head of FJH vs. traditional hydrometallurgy, and the ElementUS-Metallium-Rice consortium vs. the Ramaco-NETL precedent.
• australia-malaysia-rare-earth.md — frames Lynas/Malaysia LAMP against the all-American midstream Steve prefers (the "route around Malaysia" comparison).
• deep-dive §16 Investment Opportunities and §17 Tiered Framework — the canonical 10-name ranking and Tier 1/2/3 sort reproduced above.

The buy-list-2026.md file lists UUUU under "Uranium" (unchecked) — confirming Energy Fuels' dual U+REE classification in Pink's own watchlist; the broader commodity buy-list (copper ETF, copper miner capex play, a Canadian gold miner) is upstream-only and not REE-specific.

Monitor

This is the living monitor for the critical-minerals sector: a rolling, dated log of policy moves, procurement actions, export-control escalations, earnings/production milestones, and the standing watch-items that anchor each scan. It consolidates the recurring framing from the dated briefings of Feb 23, Feb 28, and Mar 1, 2026, the daily scan checklist, and the dated specifics scattered across the deep-dive, China-coercion, ministerial, Australia-Malaysia, US-strategy, and Flash Joule Heating sources. Company-specific detail lives on ticker pages — see UUUU, MP, LYC.AX, FCX, ALB, UAMY, PPTA, METC, AREC, MTM, USAR.

How to read the signals

Each item is staged by where it sits on the path to revenue. This is the framework used in every scan — it converts a headline into a timing estimate.

Signal Stage	What It Means	Lead Time to Revenue
RFI	Govt is sizing the market	12–18 months
FOA / NOFO / BAA	Budget committed, seeking applicants	6–12 months
LOI (EXIM)	Project deemed strategically viable	6–12 months
Application deadline	Applicants known soon	3–6 months
Award announcement	Company selected, contract signed	0–6 months
IDIQ contract	Guaranteed buyer at floor price	Revenue imminent
Section 232 investigation	Tariffs likely	6–12 months
Executive order	Policy direction set	Varies

Color convention from the baseline scan: GREEN = approved capital / signed contract; AMBER = budget committed or negotiations ordered, outcome pending; RED = early-stage RFI / proposed rule, sizing only.

---

Dated developments — chronological log

This is the consolidated timeline. Earlier entries persist because they are still live catalysts; nothing here is "old news" until its deadline passes or its outcome resolves.

2023 — the opening shots (China export controls). July 2023: gallium and germanium export licensing (shipments to the US dropped to near-zero through Oct 2024 per Chinese customs; Henkel declared force majeure). Oct 2023: graphite export restrictions. Aug 2023: Camp David trilateral (US-Japan-Korea) institutionalized critical-minerals cooperation; Korea given MSP chairmanship.

Dec 2024. China imposed a complete export ban to the US on gallium, germanium, antimony, and superhard materials. Antimony trioxide prices subsequently surged 228% from Jan 2024, reaching $39,000/metric ton.

Feb 2025. China imposed tungsten export controls; APT prices surged 20%+. Korea's Special Act on National Resource Security took effect.

Apr 4, 2025. China extended export controls to 7 heavy REEs plus all compounds, metals, and magnets (Dy, Tb among them). European NdFeB magnet prices reached 6x Chinese domestic prices during April–May 2025; automakers cut production. F-35 production was temporarily halted in 2022 when a Chinese-origin magnet was found in the supply chain — the precedent for why this matters.

May 2025. Lynas LAMP (Kuantan, Malaysia) commenced the first-ever commercial production of separated dysprosium oxide outside China; terbium followed in June. First crack in China's absolute heavy-REE monopoly. LYC.AX

Jul–Sep 2025. Energy Fuels commenced heavy REE oxide production (Dy, Tb, Sm) at White Mesa Mill — first US commercial HREE (July). Sojitz/JARE began importing Dy and Tb from Lynas LAMP in October — activating the only operational non-Chinese heavy REE supply chain in the world. POSCO International qualified Energy Fuels NdPr oxide for EV/hybrid magnets (Sep) — the single most important validation event for the Western supply-chain thesis. UUUU

Oct 9, 2025. China expanded controls to "parts/components/assemblies" containing Chinese REEs.

Oct 20, 2025. US-Australia Framework for Securing Supply in Critical Minerals signed: each country committed at least US$1B in financing within six months; total Australian project pipeline $8.5B. Same window: US-Japan Framework (Oct 2025, with bilateral Rapid Response Group), US-Malaysia MoU (Oct 26, non-binding), and EXIM's $2.2B in Letters of Interest to seven Australian companies (Arafura, Northern Minerals, Graphinex, La Trobe Magnesium, VHM, RZ Resources, Sunrise Energy Metals).

Nov 2025. China announced a temporary 1-year pause on the Oct 2025 controls (until Nov 2026) — explicitly a "temporary adjustment," framework intact. Vulcan Elements grand opening (RTP/Benson, NC). $1.4B ReElement-Vulcan-USG partnership announced. ElementUS received a $29.9M DPA Title III grant (one of 18 DPA grants totaling $887M).

Dec 1, 2025. China extended controls extraterritorially to "internationally made" products using Chinese REEs. Dec 12, 2025: binding Metallium-ElementUS LOI signed (33 days before the DOE grant deadline). Dec 15 & 28, 2025: Steve Zissouu (Meridian Report) disclosed he closed his Lynas position, citing dependence on the Malaysian political climate. Dec 19, 2025: Energy Fuels HREE (Dy) qualified for magnets by a major South Korean automaker at 99.9% purity. Pax Silica Declaration (Dec 2025): coalition foundation linking critical minerals to AI/semiconductor security; members US, Australia, Japan, with Canada/Taiwan/EU as observers, India expected.

Jan 1, 2026. China imposed new restrictions on unprocessed rare earths.

Jan 5, 2026. DOE $134M REE Supply Chain NOFO (CMEI) applications closed — awards expected Q1–Q2 2026. (Same Jan 15 date is the cross-cutting DE-FOA-0003583 application deadline.)

Jan 6–8, 2026. China's Ministry of Commerce announced a ban on "dual-use" exports to Japan, effective immediately; two days later began halting rare earth and magnet shipments to Japanese firms. The most significant weaponization since the 2010 Senkaku embargo. Trigger: PM Takaichi's Nov 2025 "survival-threatening situation" statement on Taiwan. Nomura: a 3-month restriction costs Japan ~¥660B ($4.2B, -0.11% GDP); a full year ~¥2.6T (-0.43% GDP). Japan's heavy-REE dependence on China approaches 100%.

Jan 14–15, 2026. Section 232 proclamation on processed critical minerals and derivative products (White House). No immediate tariffs; negotiations ordered with a 180-day deadline (report due Jul 13, 2026). Commerce found the US is 100% import-reliant on 12 minerals and 50%+ on 29 others. The Jan 15 EO authorized price-support mechanisms and trade remedies (tariffs) if negotiations fail. (Note: the Jan 20 date appears in the baseline scan; sources cite Jan 14–15 in the later briefings — flagged as a minor cross-source discrepancy.)

Jan 20, 2026. DOD DPA Title III scandium award ($10M, NioCorp). NB

Jan 23, 2026. DOE $275M By-product Minerals Recovery FOA (DE-FOA-0003583, HGEO) applications closed — awards expected Q2 2026.

Jan 29, 2026. BOEM RFI on OCS minerals offshore Alaska (Federal Register).

Feb 2, 2026 — Project Vault. EXIM approved a $10B direct loan, the largest in EXIM history, to establish the US Strategic Critical Minerals Reserve — a physical stockpile of all 60 USGS-listed critical minerals across US facilities. ~$2B additional private capital → $12B total. Commodity traders Hartree Partners, Traxys North America, and Mercuria Energy handle purchases; corporate partners include General Motors, Stellantis, Boeing, Alphabet (Google), Corning, GE Vernova, Western Digital. Manufacturers commit to buying set quantities at fixed prices with repurchase agreements; full drawdown permitted during major supply disruptions. The single largest demand signal in the sector — approved capital, not an RFI. (Note: corporate-partner lists differ slightly across the Feb 4 ministerial source and the Feb 23 baseline scan — both reproduced.)

Feb 4, 2026 — Critical Minerals Ministerial & FORGE launch. State Department hosted 50+ nations (the briefings cite both "54 nations" and "over 50"). VP Vance announced an allied "trading bloc" and price-floor mechanisms; Secretary Rubio launched FORGE (Forum on Resource Geostrategic Engagement) as successor to the Minerals Security Partnership. South Korea to chair through June 2026. The US, EU, and Japan announced plans for critical-minerals price floors. 11 new bilateral MOUs signed/expected (Argentina, Ecuador, Guinea, Morocco, Peru, Philippines, UAE, UK, Uzbekistan, Cook Islands, Paraguay). EU represented by VP Séjourné (proposed US-EU roadmap within 3 months + joint purchasing + strategic stockpiling center). India's Jaishankar met Rubio Feb 3. Korea's Cho Hyun cited a $350B US investment commitment. DRC's Tshisekedi led a delegation (DRC holds 70% of global cobalt reserves), offering 100,000 MT copper exports. Australia's Madeleine King pushed price floors and announced Australia's own A$1.2B stockpile (rare earths, antimony, gallium). The Framework Agreement requires signatories to identify priority projects within six months, commit to expedited permitting, and implement national-security reviews on asset sales. Tension flagged: reports indicated the administration stepped back from guaranteed minimum prices, causing Australian mining stocks to drop; some European countries hesitated to sign over Greenland uncertainty.

Feb 4, 2026 — H.R. 4090 (Critical Mineral Dominance Act) passed the House. Sponsor Rep. Pete Stauber. Directs Interior to fast-track mining on federal lands, suspend burdensome regulations, and review state/local laws impeding development. White House issued a Statement of Administration Policy in support. Now headed to the Senate. Companion track: S. 2860 (Offshore Critical Minerals Dominance Act, Sen. Sheehy — hearing held Feb 12), S. 789 (Critical Minerals Security Act), S. 2550 (Critical Minerals Partnership Act). House Energy & Mineral Resources legislative hearing Feb 24.

Feb 5, 2026. Malaysia's Natural Resources Minister told Parliament the US trade agreement "creates no binding supply obligations or market preferences" and rare earths "were never offered as a concession or quid pro quo."

Feb 9, 2026. Treasury RFI on CFIUS Known Investor Program (Federal Register, open).

Feb 20, 2026. Forest Service proposed rule on locatable minerals on federal lands (Federal Register 2026-03364) — could streamline permitting, the single biggest bottleneck for domestic supply. Comments due Apr 21, 2026.

Feb 24, 2026. BOEM published proposed revisions to OCS hard-mineral regulations (aligns with EO 14285) — opens a virtually untapped offshore frontier. Comment period closes Apr 27, 2026.

Feb 28, 2026 — DLA strategic materials $1B+ push (2025-in-review snapshot). Recent purchases: $500M cobalt, $245M antimony (US Antimony Corp), $100M tantalum, $45M scandium (Rio Tinto + APL Engineered Materials). RFIs open for tungsten, graphite, dysprosium, terbium, samarium. OBBA allocates $2B additional for the National Defense Stockpile, expected to deploy late 2026/early 2027. Industry items same window: Critical Metals Corp / Saudi TQB $1.5B rare earth JV (25% Tanbreez offtake from Greenland); Glencore/Orion Critical Mineral Consortium DRC MOU.

Feb 28 / Mar 1, 2026. QazMoly near close on ~$240M in US funding for a tungsten project in Kazakhstan — friend-shoring of a DOD-critical, China-concentrated mineral. Watch for DFC/EXIM involvement and offtake/allied-processing conditionality. (DFC separately exploring $700M Kazakhstan tungsten.)

---

Standing cumulative trackers

These tables roll forward scan to scan — they are the running ledger, not point-in-time news.

EXIM LOI Tracker (cumulative). Total LOIs issued under current administration: $14.8B+.

Company	Ticker	LOI Amount	Mineral	Country	Date
Graphite One	GPHOF	$2.07B	Graphite	USA (AK + OH)	Dec 2025
American Battery Tech	ABAT	$900M	Lithium	USA (NV)	Apr 2025
American Rare Earths	ARR	$456M	Rare earths	USA	Sep 2024
Arafura Rare Earths	ARU.AX	Part of $2.2B (up to $300M)	Rare earths	Australia	Oct 2025
Northern Minerals	NTU.AX	Part of $2.2B	Rare earths (HREE)	Australia	Oct 2025
Graphinex	Private	Part of $2.2B	Graphite	Australia	Oct 2025
La Trobe Magnesium	LMG.AX	Part of $2.2B	Magnesium	Australia	Oct 2025
VHM	VHM.AX	Part of $2.2B	Titanium, REE	Australia	Oct 2025
RZ Resources	RZR.AX	Part of $2.2B	Scandium	Australia	Oct 2025
Sunrise Energy Metals	SRL.AX	Part of $2.2B	Nickel, cobalt, scandium	Australia	Oct 2025
Anovion Technologies	Private	$400M	Synthetic graphite	USA (GA)	Nov 2024
Critical Metals Corp	CRML	$120M	Rare earths	Greenland	Jun 2025
Amaero Ltd	3DA.AX	$22.8M	Titanium alloys	USA	Feb 2025
Lithium (unnamed)	—	$400M	Lithium	USA (AR)	Recent
Cobalt/Nickel (unnamed)	—	$350M	Cobalt, nickel	Australia	Recent
Tronox (with EFA)	TROX	$600M	Mineral sands → REE	Australia	2025

Watch the EXIM-EFA co-financing mechanism specifically: when an LOI converts to a final commitment under it, both governments have validated the project. Track the LOI → Final Commitment conversion rate for the seven Australian projects.

DOE funding pipeline status. Total pending pipeline ~$1.08B.

Program	Amount	Stage	Next Milestone
REE Supply Chains (CMEI)	$134M	Apps closed Jan 5	Awards Q1–Q2 2026
Mines & Metals Capacity Expansion (DE-FOA-0003583)	$275M	Apps closed Jan 23	Awards Q2 2026
Mine of the Future (HGEO)	$80M	Apps closed Jan 2026	Awards Q2 2026
Battery Materials Processing (MESC)	Up to $500M	NOI issued	NOFO expected Q1 2026
Critical Minerals Accelerator (AMMTO)	$50M	NOI issued	NOFO expected Q1 2026
ARPA-E RECOVER (minerals from wastewater)	$40M	Under selection	Selections 2026
ARPA-E ROCKS (ore characterization)	TBD	NOFO expected	Late Feb/Mar 2026

DE-FOA-0003583 detail: total up to $275M, individual awards $10–75M across ~13 projects; Topic Area 2 allocates $200M for non-coal industrial-byproduct recovery (TRL 4-5 → TRL 7); 20% minimum cost-share; 1:50 commercial-scale requirement. DOE award announcements often move micro/small-cap stocks 20–50% on the day.

DLA / NDS stockpile RFI tracker.

Mineral	RFI Status	Advanced to Procurement?	Notes
Scandium	RFI 2025	Yes → sole-source IDIQ + NioCorp $10M DPA	Active
Indium	RFI 2025	Yes → sole-source IDIQ	Active
Fluorspar	RFP 2025	Yes → $168.9M IDIQ (ceiling $250M), Ares Strategic Mining (ARS.V)	Awarded Dec 2025
Tungsten	RFI 2025	Under evaluation	Watch for RFP
Graphite	RFI 2025	Under evaluation	Watch for RFP
Samarium	RFI 2025	Under evaluation	Watch for RFP
Dysprosium	RFI 2025	Under evaluation	Watch for RFP
Terbium	RFI 2025	Under evaluation	Watch for RFP
Bismuth	RFP 2025	Seeking offers	Active
Vanadium	RFP 2025	Seeking offers	Active

The next IDIQ awards signal which companies have a guaranteed government buyer at a floor price.

---

Upcoming deadlines and dates to watch

The binary catalysts. The two most consequential near-term events are the Section 232 July deadline and the China export-control pause expiration in November.

Date	Event
Mar 2, 2026	Lynas LAMP Full Operating Stage License expires (Malaysia). Non-renewal would devastate non-Chinese REE supply with no near-term replacement. PDF construction ~72% complete, ~10% behind schedule. Govt "considering renewing"; environmental groups (SMSL, Greenpeace) demanding rejection. LYC.AX
Apr 21, 2026	Comment deadline: Locatable Minerals proposed rule (Forest Service)
Apr 27, 2026	Comment deadline: BOEM OCS critical-minerals rule
Q1 2026	DOE MESC $500M battery-materials NOFO + AMMTO $50M accelerator NOFO expected
Q1–Q2 2026	DOE award announcements ($134M REE, $275M mines, $80M mine tech)
Q1 2026	USA Rare Earth (Stillwater, OK) magnet facility commissioning; Arafura Nolans FID target USAR
Apr 2026	Lynas HRE expansion Phase 1 (samarium) production target
Spring 2026	Perpetua Resources $2B EXIM Bank loan — board review pending PPTA
June 2026	FORGE chair rotation (South Korea handoff)
Mid-2026 / H2 2026	Ucore heavy-REE facility (Alexandria, LA) commissioning; Ucore RapidSX demo to US Government UCU.V
Jul 13, 2026	Section 232 PCMDP negotiations — 180-day report to President due. If negotiations fail, tariffs and/or minimum import-price floors likely. Either outcome is a guaranteed price floor for domestic producers. The single biggest near-term catalyst.
Q4 2026	Energy Fuels large-scale Dy/Tb/HREE production capacity target UUUU; Caremag (France, JOGMEC/Iwatani) operational
Nov 2026	China export-control pause expires. Beijing retains full legal authority to reinstate bans on gallium, germanium, antimony, tungsten, and REEs with minimal notice. Resume or permanent?
Late 2026 / early 2027	DLA NDS $2B OBBA funds expected to deploy
Jan 2027	US DoD deadline: American defense contractors must source ALL magnets from non-China manufacturers — large pull-demand for Japanese/allied magnet makers
Late 2028	Perpetua Stibnite raw antimony sulfide production target
Ongoing	DLA RFIs advancing to RFPs/IDIQs (tungsten, graphite, Dy, Tb, Sm); EXIM LOIs converting to final commitments

---

Standing watch-items (the "Track:" list)

These are the recurring questions every scan should re-test. They are framed as conversion events — the trigger that re-rates the sector or a specific node.

• JOGMEC investment announcements predict which projects will produce. Any company receiving JOGMEC investment or a Sojitz offtake is being certified as strategically viable by Japan's government.
• Multi-sovereign financing convergence. When JOGMEC (Japan), EXIM (US), EFA (Australia), KEXIM (Korea), and EDC (Canada) all commit to the same project, project-failure risk drops sharply. Arafura Nolans is the template (7+ agencies, 5+ countries). Track which other projects attract 3+ sovereign financiers — they are being designated Western supply-chain nodes.
• POSCO-Energy Fuels offtake volumes and magnet qualification of additional minerals (Dy, Tb) — watch volume scaling from the 1.2 MT proof-of-concept to 100+ MT commercial supply.
• KEXIM and JBIC co-financing patterns — a joint Japan-Korea procurement initiative or co-investment in Australian projects would doubly validate projects like Arafura.
• EXIM LOI → final-commitment conversion rate for the seven Australian projects (Graphite One $2.07B and ABAT $900M are the largest standalone LOIs).
• Project Vault mineral-specific procurement announcements — minerals to be stockpiled not yet fully specified; the commodity traders (Hartree, Mercuria, Traxys) will source from miners. Watch for offtake announcements.
• Khazanah-Chinese SOE refinery talks (Malaysia) — if a deal materializes, Malaysia hosts both Western and Chinese processing technology, creating acute supply-chain-security concerns.
• Whether the Lynas Texas (Seadrift) facility proceeds — ~$300M DoD-backed; CEO said it "might not proceed" without acceptable offtake; FEED 78% complete, wastewater permitting unresolved. If it stalls, Malaysia's LAMP remains the only game in town for Western HREE.

Open research questions (keep re-testing)

• How fast can Vulcan Elements scale magnet production? 10,000 tonnes ≈ 1% of Chinese capacity — is that enough?
• Will FORGE price floors be high enough for Western mining/processing viability?
• Can ReElement's chromatographic (LAD) separation work at commercial scale? Weekly shipments are promising; 2026 order book estimated $75–100M. AREC
• What happens when China's export-control pause expires Nov 2026?
• How does Canada fit the map? Nickel, lithium, REE potential under-explored.
• Realistic timeline for US magnet production at scale? China spent 30 years building it.

What would change the analysis (thesis triggers)

• Lithium carbonate above $20/kg → confirms recovery; bullish ALB, PLS, LAC.
• China permanently ending the export-control pause → sustained premium; very bullish UUUU, LYC, MP.
• Phase II at Mountain Pass achieving >99% purity → de-risks MP integration. MP
• A major auto OEM signing a long-term offtake with a Western magnet maker → validates the thesis.
• Sodium-ion capturing >10% of EV market → bearish lithium demand.

---

The daily/weekly scan procedure

The standing checklist for self-run scans. Morning scan (5–10 min): Federal Register public inspection (search critical minerals / rare earth / lithium / graphite / cobalt / copper — note open comment periods); SAM.gov new opportunities (last 7 days, DLA/DOD/DOE RFIs/RFPs/BAAs/IDIQs); DOE EERE Exchange open FOAs; Grants.gov (simpler.grants.gov, last 30 days). Weekly scan (15–20 min): Congress committee schedules (Senate Energy & Natural Resources, House Natural Resources, Armed Services — hearings, NDAA markups, Section 232); White House + State (EOs, bilateral agreements, ministerials, fact sheets); DOD/DLA (businessdefense.gov DPA Title III announcements); EXIM Bank news (LOIs, final commitments, "Make More in America"); ARPA-E news.

Monitoring sources — quick links. Federal Register (daily), SAM.gov (daily), DOE EERE Exchange (weekly), Grants.gov (weekly), DOD Business Defense (weekly), EXIM Bank News (weekly), ARPA-E (weekly), Congress Committees (weekly), White House (weekly), State Dept (weekly), The Meridian Report (as published — Western supply-chain thesis + policy monitoring), Inside Govt Contracts (weekly), InvestorNews Funding Tracker (monthly). Supplementary research feeds: IEA Critical Minerals Outlook, CSIS, Benchmark Mineral Intelligence (Li/Co/Ni/graphite pricing), Sprott Insights, company IR pages, FORGE ministerial updates.

Key minerals to track (and why)

Mineral	Why	Primary Signal Sources
Rare earths (NdPr, Dy, Tb)	Defense, EVs, wind. China 91% refining	DOD/DLA, EXIM
Lithium	EVs, storage. Deficit emerging 2026	DOE LPO, EXIM
Graphite	Anodes. China 98% processing	DOE MESC, Section 232
Copper	Electrification. 30% deficit by 2035	Section 232 (50% tariff)
Cobalt	Batteries. DRC 74% mining	DOD/DPA, DLA
Gallium	Radar, semiconductors. China export ban	DOD, ARPA-E
Germanium	Night vision, defense. China export ban	DOD, DLA
Antimony	Munitions, flame retardants. China ban	DLA stockpile
Tungsten	Defense, cutting tools	DOD Title III
Fluorspar	Steel, aluminum, chemicals	DLA (IDIQ awarded)
Scandium	Aerospace alloys	DLA (RFI advanced to IDIQ)

Buy-list cross-reference (Pink's 2026 watchlist)

Open commodity/mineral items on the 2026 buy list that this monitor feeds: uranium (UUUU, URG/UEC, Cameco/UEC, URNM); copper (a copper ETF and a copper-miner capex play); one Canadian gold miner (per Doug — not the one with African assets). These are unchecked as of the last buy-list update (2026-04-18).

---

Scan cadence: the baseline scan (2026-02-23) recommended a 2-week interval. Baseline tallied 17 signals — 3 GREEN (Project Vault, NioCorp DPA, Ares fluorspar IDIQ), 9 AMBER, 5 RED. Monitor SAM.gov, Federal Register, and EXIM.gov weekly. For on-demand automated scans use /monitor [sector].

Sources

Primary author / thesis driver

The spine of this sector's coverage is Steve Zissouu's The Meridian Report (source-meridian-report) — the Western supply chain thesis, the Australia → USA → Japan/Korea trilateral framework, and most of the inter-country relationship analysis trace back to it. Key cited pieces:

• "Envisioning the Western Supply Chain," The Meridian Report, November 26, 2025 — the anchor essay for the trilateral architecture and the two-parallel-chains (Energy Fuels monazite vs. ReElement mixed-carbonate/recycling) framing.
• "USA, Australia and Japan: Pondering the Developed Critical Minerals Chain," October 27, 2025 — the explicit Australia Feedstock → USA Refining → Japan Downstream OEM/Magnets thesis (Malaysia deliberately routed around).
• "Critical Minerals: Four Catalysts" — https://www.themeridianreport.com/p/critical-minerals-four-catalysts
• Trading Terminal post, December 15, 2025 — Zissouu explains closing his Lynas position ("their solvent extractions operations were too dependent on political climate in Malaysia").
• Q&A post, December 28, 2025 — frames Lynas's Malaysia reliance as a risk: "That is a risk that I wish they hadn't needed to mingle themselves with."
• "The Red Zone: Surviving the Defense Mineral Embargo," January 24, 2026 — covers the DOE $275M DE-FOA-0003583 grant program's pivot from "resilience" to "mobilization"; largely paywalled.

Policy & geopolitics analysis

• IEA — Global Critical Minerals Outlook 2025 (https://www.iea.org/reports/global-critical-minerals-outlook-2025); cited as the definitive global supply-demand data source, the ~$325B combined market sizing, the recycling-reduces-new-mine-needs projections (40% Cu/Co, 25% Li/Ni by 2050), and the Malaysia "world's second-largest REE refiner by 2030 / 12% share vs. China 77%" projection.
• CSIS (Center for Strategic and International Studies) — "China's New Rare Earth and Magnet Restrictions Threaten US Defense Supply Chains"; "Ten Years of C5+1 US-Central Asia Minerals Cooperation"; plus the midstream "high-cost endeavor with slim profit margins" assessment. General label: geopolitical supply chain analysis.
• Atlantic Council — "US Critical Minerals Policy Goes Collaborative with FORGE."
• CNBC — FORGE price floors (https://www.cnbc.com/2026/02/05/...forge-china-rare-earths-ai-chips-pax-silicchina-.html); "Taking on China from Russia's border: Inside Neo's rare earths factory" (Dec 4, 2025).
• BPC (Bipartisan Policy Center) — "Project Vault and FORGE: The Administration's Latest Moves to Secure Critical Minerals."
• Chatham House — "China's new restrictions on rare earth exports send stark warning to West" (Oct 2025); Dec 2025 warning that Trump's bilateralism/protectionism may undermine the Camp David consensus.
• CFR (Council on Foreign Relations) — "Leapfrogging China's Critical Minerals Dominance."
• ODI — "Critical Minerals Geopolitics in 2026: Risks, Supply Chains and Global Power Shifts."
• Carnegie Endowment — "Central Asia CRM Offers" (Jan 2025).
• ASPI (Australian Strategic Policy Institute) — assessment that Australia-South Korea cooperation has advanced to "tangible commercial action" versus framework-stuck partnerships.
• Sullivan & Cromwell — legal analysis that US frameworks with Australia/Japan are "more concrete and more ambitious" than the Malaysia/Thailand MOUs.
• Nomura Research Institute — estimate that a three-month Japan REE export restriction costs ~¥660B ($4.2B)/0.11% GDP; a full year ~¥2.6T/0.43% GDP.
• Wall Street Journal — Jan 8, 2026 report that Beijing began halting rare earth and magnet shipments to Japanese firms.
• Reuters — Oct 2025 report on Khazanah Nasional's early talks with a Chinese SOE for a rare earths refinery.

Government frameworks & primary documents

• White House — US-Japan Framework for Critical Minerals (Oct 27-28, 2025, with Rapid Response Group); Section 232 proclamation on processed critical minerals (Jan 14/20, 2026); SAP in support of H.R. 4090; Jan 15, 2026 executive order authorizing price-support mechanisms.
• State Department — 2026 Critical Minerals Ministerial release (Feb 4, 2026) and FORGE launch; the circulated nonbinding Framework Agreement on Cooperation on Critical Minerals Sourcing and Processing.
• EXIM Bank — "EXIM Powers America First with $2.2 Billion in Critical Minerals Commitments"; Project Vault $10B board approval (https://www.exim.gov/news/project-vault).
• METI (Japan) — France heavy rare earth (Caremag) project release (Mar 17, 2025).
• DFC — "DFC Joins $1.8 Billion Consortium" (Orion Critical Mineral Consortium).
• UAE Embassy — UAE-US Framework signing release (Feb 2026).
• USGS — 2025 Final List of 60 critical minerals (Nov 6, 2025); Central Asia REE inventory fact sheet FS2017-3089 (384 Tien Shan occurrences).
• Congress.gov — H.R. 4090 (Critical Mineral Dominance Act, Rep. Stauber); S. 2860 (Offshore Critical Minerals Dominance Act, Sen. Sheehy); S. 789 (Critical Minerals Security Act); S. 2550 (Critical Minerals Partnership Act).
• Federal Register — Forest Service locatable minerals proposed rule (2026-03364, comments by Apr 21, 2026); BOEM OCS hard minerals proposed rule (2026-03690, comments by Apr 27, 2026); Treasury CFIUS Known Investor Program RFI; BOEM offshore Alaska OCS RFI.
• Inside Government Contracts — "Federal Push for Critical Minerals Stockpiling: 2025 in Review and Outlook for 2026."
• GAO — finding that DOD lacks data to project shortfalls for nearly half of critical materials.

Company IR / corporate milestones

• Energy Fuels (UUUU) investor releases — HREE production (Jul 17, 2025); first kilogram 99.9% Dy oxide (Aug 21, 2025); "six of seven REOs under controls" (Apr 17, 2025); US-mined-and-processed REEs into permanent magnets (Sep 9, 2025); HREE qualified for magnets (Dec 19, 2025); EFA Conditional Letter of Support for Donald (Oct 20, 2025); POSCO-EFUELS collaboration (Mar 17, 2025); Energy Fuels-Vulcan MOU (Aug 26, 2025).
• Chemours — Chemours-Energy Fuels strategic alliance press release (Mar 2025).
• Vulcan Elements — $1.4B partnership announcement (vulcanelements.com).
• ReElement Technologies / American Resources (AREC) — via StockTitan: capacity-doubling and weekly-shipments releases; ERI-ReElement strategic processing partnership (eridirect.com, Nov 2025).
• MP Materials (MP) — DoD transformational public-private partnership release ($400M equity, $150M loan, $110/kg NdPr price floor, 10X facility offtake).
• Ucore Rare Metals — DoD initial construction funding (~US$18.4M); DPAS status update; Ucore-Metallium strategic partnership.
• Lynas Rare Earths (LYC.AX) — Malaysia heavy REE coverage via DiscoveryAlert.
• Neo Performance Materials — Q3 2025 results (neomaterials.com); plus the CNBC Estonia factory feature above.
• Noveon Magnetics / Kangwon Energy — 2,000T magnet facility MOU (PR Newswire).
• Lithium Americas (LAC) — Thacker Pass project page.
• Metallium (ASX: MTM / OTCQX: MTMCF) and Flash Metals USA — ASX disclosures on FJH licensing, capital raises, Gator Point Technology Campus.

Scientific / technical foundation (Flash Joule Heating)

• Rice University / Dr. James M. Tour lab — FJH invention; patents owned by William Marsh Rice University, key patent US 12,054,391 B2 (granted Aug 6, 2024).
• Nature — Luong et al., "Gram-scale bottom-up flash graphene synthesis," Nature 577, 647-651 (Jan 2020) — the seminal FJH paper.
• PNAS — Xu et al. 2025, life-cycle assessment quantifying FJH's 87% energy / 84% GHG / 54% opcost reductions.
• Additional peer-reviewed validation cited across Nature Communications (2021), Science Advances (2022), Nature Chemistry (2024), and Advanced Materials (2025).
• NETL (National Energy Technology Laboratory) — Ramaco Brook Mine "one of the largest unconventional deposits" validation; CRADA precedent model.
• Fluor Corporation — July 2025 Preliminary Economic Assessment of Ramaco's Brook Mine ($5.1B NPV, >150% IRR).
• University / lab partners cited: University of Tokyo (Minamitorishima deep-sea REE mud discovery), Missouri S&T (gallium/germanium extraction with Rio Tinto Kennecott), Ames National Laboratory (Critical Materials Innovation Hub), WVU Water Research Institute (AMD-to-REE, Mount Storm pilot).

Specialized industry / market intelligence

• Benchmark Mineral Intelligence — listed as the reference for Li, Co, Ni, graphite pricing.
• Sprott Insights — critical materials ETF perspective (context for the SETM ETF).
• InvestorNews — company funding tracker.
• REEx (Rare Earth Exchanges) — "REEx 2025: The Year Ex-China Rare Earth Supply Chains Hit Critical Mass."
• Columbia (Center on Global Energy Policy) — "How the US-Japan Critical Minerals Partnership Is a Long-Overdue Step Toward Real Supply Chain Security."
• SFA Oxford — "Japan-US Rare Earths Deal" analysis.
• Mining.com — QazMoly ~$240M US funding for Kazakhstan tungsten (Feb 28, 2026).
• StockTitan, DiscoveryAlert, PR Newswire, eridirect.com — press-release aggregation / primary corporate news.

Daily-scan monitoring sources (operational watchlist)

The monitor and checklist pages run off a recurring source set: Federal Register, SAM.gov, DOE EERE Exchange, Grants.gov, DOD Business Defense, EXIM Bank News, ARPA-E, Congress committee schedules, White House, State Dept, The Meridian Report, Inside Government Contracts, and the InvestorNews Funding Tracker.

Attribution caveats

• The two large primer files (critical-minerals.md v1 and critical-minerals-deep-dive.md) are near-identical duplicates carrying the same Meridian Report attribution and the same consolidated source block — treat as one body of work, not two independent sources.
• Several research files are tagged source: claude-chat (synced from claude.ai conversations), meaning the underlying citations are second-hand synthesis rather than directly fetched primary documents.

Consolidated source files

This sector page consolidates the following vault files (do not treat as external citations — they are the inputs being merged):

• critical-minerals — primer v1 (topic page, Meridian Report-informed)
• critical-minerals-deep-dive — /deep-dive output; duplicate of the primer
• critical-minerals-china-coercion — Jan 2026 Japan export-ban analysis + supply-chain primer
• critical-minerals-ministerial — Feb 4, 2026 Critical Minerals Ministerial / Project Vault / FORGE
• australia-malaysia-rare-earth — Australia-Malaysia corridor, Lynas LAMP license risk
• us-critical-minerals-strategy — DE-FOA-0003583, the seven irreplaceable minerals, domestic-producer financial profiles
• rice-flash-joule-heating-doe-grant — Rice / Metallium / ElementUS FJH consortium
• critical-minerals-monitor-2026-02-23 — baseline policy & procurement monitor
• critical-minerals-monitor-2026-02-28 — daily briefing
• critical-minerals-monitor-2026-03-01 — daily briefing (light feed)
• daily-scan-checklist — manual daily/weekly scan procedure
• buy-list-2026 — watchlist (commodity/uranium/copper/gold)

---

Consolidation queue (merged 2026-05-30 — section-scoped rebuild)

Industry-wide content folded in from these source files. They stay live pending Pink's archive confirm.

• [ ] critical-minerals.md
• [ ] critical-minerals-deep-dive.md
• [ ] critical-minerals-china-coercion.md
• [ ] critical-minerals-ministerial.md
• [ ] australia-malaysia-rare-earth.md
• [ ] us-critical-minerals-strategy.md
• [ ] rice-flash-joule-heating-doe-grant.md
• [ ] critical-minerals-monitor-2026-02-23.md
• [ ] critical-minerals-monitor-2026-02-28.md
• [ ] critical-minerals-monitor-2026-03-01.md
• [ ] daily-scan-checklist.md
• [ ] buy-list-2026.md