Critical minerals and sovereign materials
The Materials Bottleneck Behind Sovereign Technology
Why processing, not mining, is the real constraint - and who needs to meet whom to solve it.
Atomic answer
The global critical minerals supply chain has three choke points. Mining gets the headlines but processing is the real constraint. China controls a significant share of rare earth separation, graphite refining, and cobalt processing. The U.S. has limited domestic processing capacity. Building alternative supply chains requires coordinated public capital, offtake commitments, permitting reform, and buyers who will pay the security premium. This briefing maps the bottleneck - who controls what, who needs whom, and what signals to watch this quarter.
Why materials are the first front
Sovereign technology - drones, reactors, batteries, AI data centers, undersea systems, advanced manufacturing - all depend on materials capacity. Before strategy becomes capacity, someone has to mine, process, separate, refine, permit, finance, and buy. The materials layer is where policy, engineering, procurement, and capital converge in one visible bottleneck.
Drones need rare earth magnets for flight controllers and electric propulsion. Reactors need uranium fuel processing, zirconium alloys, and high-purity nickel. Batteries need lithium at 99.5% purity, graphite that has been spheronized, cobalt that has been refined. AI data centers need copper, rare earths for cooling systems, and semiconductor-grade materials. Undersea systems need specialty steels and acoustic materials. Each of these supply chains passes through a processing gate - and in too many cases, that gate sits in a single country.
A founder building an autonomous system or an energy storage company cannot fix the materials supply chain. But every founder can map where their bill of materials touches a single-country bottleneck. That map is the first step toward building a supply chain that can survive the next export control, trade restriction, or commodity price shock. The companies that do this work before it becomes a board-level question will have more resilient products, more bankable supply chains, and more defensible market positions.
Where is the real bottleneck: mining or what happens after?
Mining capacity exists. Australia, Chile, the Democratic Republic of Congo, and the United States all have mineral reserves. The bottleneck is what happens after extraction.
Separation: Rare earth elements must be separated into individual oxides. The fifteen lanthanides plus scandium and yttrium are chemically similar, making separation a solvent-intensive, multi-stage process that generates significant waste streams. China remains dominant in global separation capacity. U.S. rare earth production capacity exists at Mountain Pass in California, but domestic separation capacity remains limited. Building domestic separation capacity is not a science problem. It is a capital, permitting, and offtake problem.
Refining: Battery-grade lithium requires high purity. Graphite must be spheronized, and China remains dominant in synthetic and natural graphite processing. Cobalt refining is concentrated in China even though the majority of cobalt is mined in the DRC. The refining gap is not theoretical: it determines whether a battery factory in Georgia or a gigafactory in Nevada can source materials without routing through a single jurisdiction.
Timeline and capital intensity: Processing capacity takes years and substantial capex to build. These are not venture-scale assets. They require project finance, offtake, and public capital coordination.
How do processing, separation, refining, offtake, and permitting fit together?
Five bottlenecks run in sequence. Understanding their order is the difference between a project that reaches commercial operation and one that becomes a press release.
Table 1 - Bottleneck Map:
| Stage | What it is | Who controls it now | Constraint severity |
|---|---|---|---|
| Mining | Ore extraction | Australia, Chile, DRC, US | Medium |
| Processing | Ore → usable material | China dominant | High |
| Separation | Individual rare earths | China near-monopoly | Critical |
| Refining | Battery/defense-grade | China, Japan, Korea | High |
| Offtake | Purchase commitments | Few long-term contracts | High |
| Permitting | Regulatory approval | US/Allied governments | High |
| Finance | Project capital | Public + private | Medium |
Processing: Transforming ore into usable material - crushing, grinding, flotation, leaching. This is the industrial-scale step that determines whether concentrate becomes feedstock.
Separation: Isolating individual rare earth elements. The fourteen stable lanthanides plus scandium and yttrium each have specific end-uses. Neodymium and praseodymium go into magnets. Europium and terbium go into phosphors and defense applications. You cannot separate one without separating them all - the economics require full-stream processing.
Refining: Achieving battery-grade or defense-grade purity. The difference between 99% and 99.5% lithium purity is the difference between a battery that meets specification and one that does not. Defense applications often require higher purity than commercial ones.
Offtake: Purchase commitments that make processing facilities bankable. Public capital can start a plant; offtake determines whether it survives. Without multi-year contracts from battery manufacturers, defense primes, or energy developers, a processing facility cannot secure project finance. This is the bridge most government programs do not build.
Permitting: New U.S. mines and processing facilities can face multi-year review timelines depending on environmental review, local opposition, and agency coordination. Permitting reform is one of the highest-leverage policy changes available and one of the hardest to pass.
Table 2 - Who Needs Whom:
| Actor | Controls | Needs | Why it matters |
|---|---|---|---|
| Mining operators | Ore reserves | Processing contracts, offtake | Without processing partners, ore stays in the ground |
| Processors | Refining capacity | Permits, offtake, project finance | Multi-hundred-million capex needs bankable demand |
| Strategic buyers | Purchase authority | Secure supply, price stability | Defense and energy security depend on supply certainty |
| Government programs | DPA, LPO, OSC authority | Qualified applicants, bankable projects | Public capital can start capacity; offtake determines survival |
| CVCs | Strategic investment capital | Technical co-evaluation, policy risk assessment | Corporate venture needs to diligence both the technology and the regulatory surface |
| Sovereign/allied capital | Patient capital, mandate | Technical due diligence, project pipeline | Allocators know they cannot evaluate these projects independently |
Who controls each stage, company by company?
The stage table above names countries. This section names the companies and state groups that operate the gates, and what changed in 2025-2026.
| Stage | Controller | What the record shows | Source confidence |
|---|---|---|---|
| Rare earth separation and refining | China Rare Earth Group and China Northern Rare Earth Group | By 2025, China's mining and smelting-separation quotas were allocated almost exclusively to these two state champions. China Rare Earth Group was formed in December 2021 by merging the rare earth assets of CHALCO, China Minmetals, and Ganzhou Rare Earth Group, consolidating most heavy rare earth capacity. China Northern Rare Earth Group's parent holds exclusive mining rights at Bayan Obo. | Mixed (RAND analysis of the quota system, citing Reuters; the underlying MIIT quota document was not opened this pass) |
| Global rare earth refining and magnets | China (national share) | China accounted for 91% of refined rare earth output in 2024 and 94% of sintered permanent magnet production, per the IEA. In 2025 its share of magnet rare earth separation was 85%, down from 90% in 2023, with new US and Malaysian capacity the cause; planned projects could take that share to about 70% by 2035. Reported cost: a 2023 era "about 60% separation" figure understates the refining and magnet gates. | Mixed (IEA Global Critical Minerals Outlook 2026 and IEA Rare Earth Elements report; iea.org returned HTTP 403 to automated access on 2026-08-02, and that access check must be re-run before publish) |
| Lithium refining | China (national share); Ganfeng Lithium and Tianqi Lithium are the largest listed Chinese refiners | China accounted for 70% of lithium refining in 2025, per the IEA. Ganfeng reported 182,400 tonnes LCE of lithium chemical output in 2025 and states that its lithium compound capacity ranks first in China and third worldwide; Tianqi is the other large listed refiner, with upstream exposure through Greenbushes. | Primary for the IEA and Ganfeng figures (company disclosure); Analytical for the "two largest listed" framing |
| Cobalt refining | China (national share); Zhejiang Huayou Cobalt via Congo Dongfang International Mining (CDM) | China held 79% of refined cobalt output in 2025, per the Cobalt Institute. Huayou's CDM unit is a major DRC copper-cobalt operator; the DRC suspended CDM operations in November 2025 after a spill. | Primary (Cobalt Institute; Reuters) |
| Western integrated lithium | Albemarle | Albemarle operates a 5,500-tonne conversion site at Kings Mountain, NC. DOE awarded about $150M in 2022 toward a commercial-scale processing plant and DoD awarded $90M in 2023 for mine equipment. The Kings Mountain project completed federal environmental review and permitting on 2026-03-26 and was awaiting state and local permits as of 2026-09-14. Its Chester County, SC "Mega-Flex" hydroxide plant was announced in 2023 at $1.3B for 50,000 t/yr expandable to 100,000 t; construction timing has slipped since announcement. | Primary (Albemarle; permits.performance.gov; DOE NEPA; WFAE) for awards and permits; the SC plant's current construction status is a hold (see Holds) |
Two corrections for readers of older versions of this map. First, China Minmetals is no longer a separate rare earth controller: its rare earth assets were merged into China Rare Earth Group in December 2021. Second, the older "60% of separation" number is a mining-era figure; the refining and magnet shares are higher (91% and 94% in 2024) and are only now beginning to decline. Source confidence: Mixed.
What did the first wave of materials awards actually buy?
The signal list in this article says to watch DPA Title III awards. Here is what the record shows for three companies this map tracks, and what it implies for anyone underwriting a processing project.
- MP Materials. A $35M DoD award in February 2022 funded heavy rare earth separation at Mountain Pass, on top of a $9.6M DPA Title III award in December 2020 for light rare earth processing. In July 2025, DoD signed a package that goes well beyond grant funding: $400M in convertible preferred equity, a 10-year $110/kg NdPr price floor, a 100% offtake commitment for the expanded 7,000 t/yr magnet capacity over 10 years, and a $150M loan for heavy rare earth separation. MP's total US magnet capacity target is about 10,000 t, with the second facility expected to begin commissioning in 2028. Source confidence: Primary (defense.gov; MP Materials; SEC filings).
- Lynas. Awarded about $258M in 2023 for a heavy rare earth separation facility at Seadrift, Texas. The facility will no longer proceed: Lynas's FY2026 report reallocated the funding to a binding letter of intent signed in March 2026 for about $96M of light and heavy rare earth oxides over four years, and the company now produces separated dysprosium, terbium, and samarium in Malaysia. Source confidence: Primary (Lynas FY2026 annual report; WSJ).
- Talon Metals. The $20.6M DoD award in September 2023 was for exploration and mineral resource definition at Tamarack, not for the processing plant. The North Dakota processing facility is a different, larger award: $114.8M from DOE under the Bipartisan Infrastructure Law, with the Beulah site secured in 2025. Source confidence: Primary (defense.gov; USAspending; Senate release; Talon).
The actionable paragraph is the DPA clock. Most DPA authorities, including Title III, now expire on December 11, 2026 (extended by P.L. 119-103, signed September 2, 2026); the DPA Modernization Act of 2026 (H.R. 7688) would extend them to 2031. A reader planning around a Title III award cycle should treat December 11 as a live variable, not background. Source confidence: Primary (CRS R49178).
What three signals should operators watch this quarter?
1. Permitting reform legislation. Senate Energy and Natural Resources Committee activity on critical minerals permitting is the highest-leverage signal to watch. FAST-41 and the BUILDER Act stress tests could compress timelines - or confirm that the current pace is baked into the system. Any bipartisan permitting bill that clears committee with critical minerals provisions is a market-moving event for processing project timelines.
2. DOE Battery Materials Processing grants. Round 2 awards have been announced. Round 3 is expected late 2026. Track USASpending.gov for award patterns: which technologies are receiving funding, which companies are receiving repeat awards, and whether awards are clustering around specific processing stages (separation vs. refining vs. recycling). These award patterns signal where DOE sees the most acute domestic capacity gaps.
3. DPA Title III materials awards. The annual DoD cycle typically runs summer. Watch for rare earth separation and graphite processing awards specifically. A Title III award for domestic rare earth separation would be a demand signal for private capital because it confirms that the Department of Defense has identified a specific processing bottleneck as a national security risk. Most DPA authorities now expire on December 11, 2026 (extended by P.L. 119-103, signed September 2, 2026); the DPA Modernization Act of 2026 (H.R. 7688) would extend them to 2031 (CRS R49178). See 'What did the first wave of materials awards actually buy?' below.
Editorial note: Specific legislative timelines and award dates should be verified against primary sources. This is editorial signal, not procurement advice. Legislation and award cycles shift. Use this framework to track movement; verify specifics before making allocation or application decisions.
What should founders building in autonomy, energy, or advanced manufacturing do differently?
Stop treating critical minerals as a policy problem. It is a market-structure problem with visible capital flows, offtake gaps, and technology bottlenecks.
If you build in autonomy, energy storage, advanced manufacturing, or AI infrastructure: map your bill of materials against the bottleneck table above. Where does your supply chain touch processing or separation capacity? If your magnets come from a supplier whose rare earth oxides pass through a single-country separation gate, that is a strategic risk your investors will diligence regardless of whether you do.
The procurement opportunity is not in mining - it is in processing technology, supply-chain traceability, recycling and recovery, and compliance infrastructure. Companies building solvent extraction technology, electrochemical separation, urban mining, or blockchain-based materials traceability are solving the real constraint. The founders who position their technology against the bottleneck map, not just the periodic table, will find the buyer conversations easier to navigate.
What should CVCs and strategic investors diligence before the term sheet?
Ask: does this company's supply chain pass through a processing bottleneck in a single country? If the answer is yes, the follow-up is whether that bottleneck has a visible alternative in development - and whether the company's cost structure can absorb the security premium if sourcing shifts.
The security premium is the opportunity. Buyers who pay for supply-chain diversity create bankable offtake that unlocks project finance. CVCs who identify companies whose technology reduces processing dependency - through new separation chemistry, recycling yields, or substitution - are positioning ahead of the next supply shock. That is the strategic investment thesis.
Policy risk is no longer a footnote. Export controls, CFIUS, and allied industrial policy are reshaping which supply chains are investable. Map the regulatory surface before the term sheet. A company whose entire supply chain is contingent on a single-country processing relationship may become uninvestable overnight if trade policy shifts - and trade policy is shifting.
What should government buyers ask before funding processing capacity?
Three questions separate projects that will reach commercial operation from those that will become grant-cycle artifacts:
First: Does this project have an offtake pathway, or is it depending on public capital alone? A government grant or loan guarantee can start construction. It cannot generate revenue. If there is no buyer at the end of the processing line - no battery manufacturer, no defense prime, no energy developer - the capacity will not survive the gap between commissioning and commercial operation.
Second: Has the applicant demonstrated processing technology at pilot scale, or is this a feasibility study? Pilot-scale data - not bench-scale, not desktop modeling - is the threshold between a credible project and a science experiment. The difference matters because processing facilities are not software. You cannot iterate after the concrete is poured.
Third: Is there a strategic buyer in the room before the award closes? The strongest signal that a processing project will succeed is a buyer who has already committed - or at minimum expressed binding interest - before the government award is finalized. Programs that fund capacity without confirming demand create stranded assets. Programs that require buyer engagement before closing create industrial capacity that lasts.
Source list
- IEA Critical Minerals Market Review
- USGS Mineral Commodity Summaries
- DOE Loan Programs Office portfolio
- DPA Title III Presidential Determinations
- Federal Register - critical minerals and permitting
- FAST-41 Permitting Dashboard
- Company filings (MP Materials, Lithium Americas, and comparable)
Sources are provided for reference and further reading. Stack & State uses primary government databases, public company filings, and public institutional sources where available. Readers should verify time-sensitive details before acting.
Stack & State is an editorial and ecosystem-intelligence publication. Nothing here is legal, investment, procurement, or compliance advice. Program details change; verify requirements with primary sources and qualified advisors.
Verified sources
- DoD awards $35M to MP Materials for heavy rare earth separation (2022-02-22)
- DoD rare earth element awards, $9.6M DPA Title III to MP Materials (2020-11)
- MP Materials-DoD public-private partnership (2025-07-10)
- MP Materials 8-K and transaction exhibits (2025-07-09)
- Lynas Rare Earths FY2026 annual report and briefings
- Talon DPA Title III $20.6M agreement (DoD release)
- Talon award record (USAspending)
- CRS R49178, The DPA Modernization Act of 2026
- IEA Global Critical Minerals Outlook 2026
- IEA Rare Earth Elements report
- Cobalt Institute, Cobalt Market Report 2025
- FAST-41 permitting record, Kings Mountain
- RAND PEA4756-1, China's rare earth policy
Last verified