Rare earth elements (REEs) - 17 metallic elements with unique magnetic, catalytic and luminescent properties sit at the heart of the energy transition and modern defense technology. Neodymium-iron-boron (NdFeB) magnets made from REEs power EV traction motors and wind turbine generators, while heavy rare earths such as terbium and dysprosium are critical to precision-guided weapons, radar and aerospace. REEs also underpin consumer electronics, robotics and the motors and sensors embedded in AI infrastructure.
China's dominance across this chain has become one of the most consequential strategic issues in commodity markets. China controls roughly 90% of global separation and processing capacity and an even larger share of magnet manufacturing, and in 2025 expanded export controls requiring licenses for rare earths, processing technologies, and even foreign-made products using Chinese-origin materials or technology. These restrictions, used as leverage during 2025 trade disputes, sent a stark signal to Western governments about the fragility of existing supply chains.
Brazil has emerged as a focal point of the diversification push. Home to the world's second-largest rare earth reserves, its Serra Verde mine became the first operating ionic clay rare earth deposit in the Western Hemisphere, and Brazilian exports tripled in the first half of 2025. Yet mining is only the first link in a long chain the central question is whether Brazil can move into separation, refining and magnet manufacturing, or remain a resource supplier feeding China's processing base. This article examines the value chain, China's position, Brazil's opportunity, the refining bottleneck, and the pathway to a more complete Brazilian supply chain.
The value chain comprises three broad stages. Upstream exploration, mining and concentrate production is capital-intensive and geologically dependent, but produces a relatively low-value, commoditized output. Midstream separation of the 17 individual elements and refining into high-purity oxides requires complex, multi-stage solvent extraction and represents the largest technical and environmental barrier in the chain. Downstream conversion of oxides into metals, alloys and NdFeB permanent magnets captures the highest value but depends on a reliable, qualified supply of separated oxides and metals from midstream.
|
Stage |
Activities |
Value Creation |
Barriers to Entry |
|---|---|---|---|
|
Upstream |
Exploration, mining, concentrate production |
Low-moderate; commoditized ore/concentrate pricing |
Capital, permitting, geological risk |
|
Midstream |
Separation, refining, oxide production |
High; technical complexity sets price for purified oxides |
Process technology, environmental compliance, scale |
|
Downstream |
Metal-making, NdFeB magnet production, advanced manufacturing |
Highest; embedded in high-value end products |
Metallurgical know-how, magnet IP, customer qualification |
Separation and refining create the majority of strategic value not because the chemistry is exotic, but because achieving magnet-grade purity at commercial scale demands decades of process know-how, dedicated infrastructure, and tolerance for handling radioactive by-products (notably thorium and uranium) a combination that has concentrated this stage almost entirely in China.
Chart 2: Rare Earth Value Chain Estimated Value Capture by Stage (Illustrative)
China holds the world's largest reserve base at an estimated 44 million tonnes REO roughly 48% of the global total anchored by deposits including Bayan Obo in Inner Mongolia. In 2025, China produced approximately 270,000 tonnes, around 69% of global mine output of 390,000 tonnes, a base no other country can match in the near term.
China's processing leadership extends well beyond resources: an estimated 88% or more of global separation and refining capacity sits in China, built through sustained state investment, integrated industrial clusters, and permitting regimes that have allowed processing at a scale few jurisdictions match. This base underpins China's dominance in NdFeB magnet manufacturing, supplying the vast majority of magnets in global EV and wind turbine supply chains.
China's October 2025 export control expansion required licenses not only for rare earth materials but for mining/processing technologies and magnet manufacturing extending to foreign-made products using Chinese-origin materials or technology, regardless of where production occurs. The objective: preserve leverage over midstream and downstream stages even as mining diversifies, ensuring new supply from Brazil, Australia and others keeps flowing to Chinese processors absent alternatives. The impact has been to accelerate Western financing for non-Chinese processing and magnet projects, though displacing China's dominance will take years.
Chart 3: Global Rare Earth Separation & Processing Capacity (Illustrative)
Source: Estimates based on USGS and Mineral processing capacity data.
Brazil holds an estimated 21 million tonnes REO about 23% of the global total and second only to China, the only other country above 10 million tonnes. Key deposits include Serra Verde's Pela Ema mine in Goiás, the first operating ionic clay deposit in the Western Hemisphere (production began early 2024 after over $1 billion invested); the Araá niobium-rare earth project in Minas Gerais; the long-studied Poços de Caldas complex; and a growing pipeline of earlier-stage ionic clay and carbonatite projects.
Brazil's carbonatite and ionic clay deposits, especially in Goiás, naturally concentrate heavy rare earths such as terbium and dysprosium elements of greatest strategic concern for defense and high-performance magnets giving Brazilian output a favorable composition versus many light-rare-earth-dominated deposits. Serra Verde's mine life is projected at 25 years (production into the 2050s), targeting roughly 6,400 tonnes annually by 2027, with over 50% heavy rare earth content.
Brazil's advantages extend beyond geology: an established mining sector with relevant expertise, abundant renewable hydropower for energy-intensive processing, and a political environment aligned with Western diversification evidenced by a $565 million U.S. DFC commitment to Pela Ema and its subsequent $2.8 billion acquisition by USA Rare Earths, framed as building a vertically integrated Western Hemisphere supply chain. These factors explain why investors view Brazil as the most credible non-Chinese mining growth story in rare earths.
Chart 1: Global Rare Earth Reserves by Country, 2025 (Million Tonnes REO)
Source: USGS Mineral Commodity Summaries, January 2025.
The gap between Brazil's resource base and processing footprint illustrates the industry's central challenge: separation is technically demanding, capital-intensive, and subject to stringent environmental requirements around radioactive by-product management, wastewater treatment and tailings storage. Building a commercial-scale plant typically requires years of process optimization specific to each ore's mineralogy, since circuits aren't easily transferable between deposits.
China's processing infrastructure is mature, integrated and delivers significant cost advantages; Australia has made the furthest progress among Western-aligned nations, with Lynas operating separation in Malaysia and developing US capacity; the US has prioritized rebuilding metal and magnet-making capacity via the DFARS rule effective January 2027, barring the Department of War from sourcing NdFeB and samarium-cobalt magnets from China and other covered countries; Brazil, despite its resource scale, has minimal domestic separation capacity, with Serra Verde historically committing output to Chinese processors under offtakes recently shortened by nearly eight years as Western alternatives develop.
Workforce and technology gaps compound the capital challenge: engineers and metallurgists with hands-on separation experience are scarce outside China, and technology transfer is constrained both by proprietary know-how and by export controls on processing technology. This explains why, despite deposits existing on every continent, commercially viable separation capacity remains overwhelmingly concentrated in one country.
Image 2: Rare Earth Separation / Refining Facility (Illustrative Schematic)
Source: Illustrative schematic of a generic solvent-extraction separation cascade.
Brazil's path beyond concentrate exports starts with separation investment. The renegotiation of Serra Verde's offtake terms with Chinese processors and its acquisition by USA Rare Earths with explicit vertical-integration ambitions signal intent to develop Western Hemisphere separation capacity rather than remain a pure concentrate exporter, though as of early 2026 actual Brazilian separation capacity remains nascent.
Magnet manufacturing potential would depend on first establishing reliable oxide and metal supply; the logic would link Brazilian heavy rare earth output to EV and wind turbine magnet demand in partner markets, and potentially defense applications given Brazil's favorable heavy rare earth mix. Realistically, magnet capacity is more likely to emerge first in the US or allied countries using Brazilian-sourced oxides, rather than in Brazil itself, at least near-term.
Strategic partnerships are the critical enabler. U.S. EXIM financing part of a nearly $4 billion wave of letters of intent across the rare earth supply chain alongside the Serra Verde transaction demonstrate active US engagement. Europe, Japan and South Korea, heavily exposed to Chinese magnet supply, represent additional partners for offtake and technology co-investment. For Brazil to move beyond concentrates, it needs sustained foreign capital and technology transfer for separation facilities, streamlined permitting for processing operations, and binding offtake agreements with Western magnet makers none guaranteed, but all now actively pursued.
Chart 5: Brazil Rare Earth Supply Chain Development Pathway (Illustrative)
Source: Analysis based on company disclosures and CSIS critical minerals tracking.
EV adoption remains the largest single driver of incremental demand, with each traction motor requiring NdFeB magnets containing neodymium, praseodymium and, for higher-performance motors, dysprosium and terbium for thermal stability. Wind energy adds a second major demand pool, particularly direct-drive offshore turbines that use substantially more magnet material per megawatt than geared onshore designs making offshore wind growth disproportionately rare-earth-intensive.
Defense applications, smaller in volume, are disproportionately important strategically: precision-guided munitions, radar and aerospace propulsion rely on heavy rare earth magnets with few viable substitutes why the DFARS 2027 restrictions target this segment specifically. Emerging demand from AI infrastructure and robotics motors, actuators and sensors for data center cooling, humanoid robots and automated manufacturing represents a newer, rapidly growing vector layered on EV, wind and electronics demand.
Collectively, energy transition policy, defense reshoring, and AI/robotics build-out point toward sustained, multi-decade demand growth across nearly every end-use reinforcing the urgency of diversifying supply away from a single-country processing base.
Chart 4: Forecast Rare Earth Demand by Sector, 2025-2035 (Illustrative)
Source: Estimates based on demand projections.
Potential winners: Brazilian miners and developers with heavy-rare-earth-rich deposits, processing companies establishing Western Hemisphere separation capacity (in Brazil or partner countries using Brazilian feedstock), and magnet manufacturers positioned to qualify non-Chinese supply chains ahead of the 2027 DFARS deadline. Potential losers: high-cost producers without comparable heavy-rare-earth content or government backing, and supply chains dependent on single-source Chinese processing as licensing tightens.
Investors should weigh resource quality particularly heavy rare earth content, commanding disproportionate strategic and price premiums against execution risk, given long lead times. Processing economics remain the largest uncertainty: separation costs, yields and compliance outside China are unproven at scale. Geopolitical support government financing, offtake guarantees, and frameworks like DFARS creating guaranteed demand is increasingly a primary determinant of viability. Key indicators: Brazilian separation facility announcements, EXIM/DFC disbursement pace, offtake terms, and realized output versus the gap between Brazil's ~2,000 tonnes (2025) and China's 270,000 tonnes.
Image 5: Global Rare Earth Supply Chain Map (Illustrative)
Source: Illustrative schematic based on USGS, CSIS and company-disclosed trade and investment relationships.
Chart 6: China vs Brazil Competitive Positioning (Illustrative, 0-10 scale)
Source: Analysis based on USGS and company disclosures.
Brazil becomes a significant, growing mining supplier, with Serra Verde expanding to roughly 6,400 tonnes by 2027 and other projects advancing. China remains overwhelmingly dominant in separation, refining and magnets, processing a substantial share of global concentrate including, initially, much of Brazil's output even as offtake terms shift toward shorter durations and Western buyers.
Brazil, backed by US and allied financing, develops meaningful Western Hemisphere separation capacity within this decade; partnerships with the US, Europe, Japan and South Korea accelerate downstream investment; and an early-stage magnet ecosystem emerges in partner countries using Brazilian heavy rare earth oxides, denting China's share specifically for heavy rare earths.
Mining expands as planned, but separation investment proves slower and costlier than anticipated; Brazil remains primarily a concentrate exporter, with value capture accruing to Chinese (or other established) processors; China's export controls adapt to maintain leverage over midstream and downstream stages regardless of where mining occurs.
The core answer: Brazil is highly likely to become the most important non-Chinese mining jurisdiction over the next decade, particularly for heavy rare earths but unlikely to challenge China's processing dominance alone. The realistic pathway runs through Brazilian mining feeding Western Hemisphere separation and magnet capacity built with US, European or allied capital and technology making Brazil a critical enabler of diversification rather than a standalone competitor. The strategic takeaway: the binding constraint is midstream processing capacity, and the speed of capital and offtake commitments into separation and magnet-making not mining output alone will determine how much of China's dominance can be challenged by 2035.