The United States is spending billions to build critical-minerals processing capacity, but a September 2026 analysis confirms a structural problem: new facilities may not have enough raw material to run. A July Defense Production Act determination, an uneven GAO assessment of recycling alternatives, and a string of recycler bankruptcies together reveal that retaining feedstock and building the capacity to refine it are two very different policy challenges.
Introduction
Imagine building a flour mill before anyone has planted the wheat. That is roughly the situation facing several of America's newest critical-minerals processing facilities. Billions of dollars in federal and private capital have flowed toward plants designed to refine battery materials, separate rare-earth elements, and recover strategic metals from scrap. The problem, as a September 18 analysis by Resource Recycling makes clear, is that investment in processing capacity does not automatically come with a guaranteed supply of raw material to feed those facilities.
This matters to a wide range of people: defense contractors racing toward a January 2027 deadline to stop buying certain critical materials from China, Russia, Iran, and North Korea; battery manufacturers trying to build domestic supply chains for electric vehicles; and policymakers who have spent years arguing that the United States needs to stop shipping its strategic resources overseas for refining. The feedstock gap is the point where those ambitions meet a less comfortable reality.
Three developments in the past two months have brought this tension into sharp focus. A July 22 Government Accountability Office report examined whether recycling and substitution could plug import gaps, and returned an uneven verdict. On July 30, President Trump invoked the Defense Production Act to authorize export controls on recoverable critical materials, including battery black mass and rare-earth swarf. And a steady drumbeat of recycler bankruptcies has illustrated what happens when domestic processors cannot compete with foreign buyers willing to pay more for U.S. feedstock.
The Structural Problem: Mining More Does Not Fix a Refining Gap
The United States updated its critical minerals list in 2025 to include 60 commodities. It was 100 percent reliant on imports for 13 of them, including scandium, tantalum, and yttrium, and more than 50 percent reliant on imports for an additional 20. China supplies more than half of U.S. imports of eight minerals, including antimony, graphite, rare earths, and arsenic.
That dependence is often framed as a mining problem. It is more accurately a chemistry problem. The United States ships more than 95 percent of its rare-earth ore to Asia for processing and imports the finished components back. China controls approximately 90 percent of global rare-earth processing capacity, roughly 80 percent of refined cobalt output, and between 60 and 71 percent of lithium refining. The United States has zero capacity to produce refined cobalt from primary ores. As MIT Professor Elsa Olivetti has put it: "For the majority of these elements, the constraint is not geological scarcity, but rather a lack of processing and manufacturing capacity or know-how."
Building new mines addresses one node in a multi-step chain. The steps that follow, separating ore concentrates, refining them to battery-grade or magnet-grade purity, and converting them into the precursor materials that manufacturers actually buy, remain concentrated in a small number of countries. New domestic mines advance without a domestic home for their output. New domestic processing plants advance without a reliable domestic source of input. The two halves of the equation are being built on separate tracks, and the gap between them is where the feedstock problem lives.
S&P Global puts the average U.S. mine discovery-to-production timeline at nearly 30 years. That means no project currently in early development will close the supply gap before the January 2027 defense procurement prohibitions take effect. Investors also need to add 18 to 24 months to any company's stated production target to account for downstream qualification periods, pushing realistic commercial output from facilities targeting late-2027 completion to 2029 at the earliest.
The Black Mass Paradox: Plenty of Scrap, Not Enough Refiners
Battery black mass is the powdery residue left after spent lithium-ion cells are shredded and processed. It contains lithium, nickel, cobalt, and manganese, exactly the materials that battery manufacturers need and that the United States currently imports in large quantities. In theory, scaling up domestic black mass recycling should reduce that import dependence. In practice, the numbers tell a more complicated story.
U.S. companies are generating approximately 60,000 metric tons of black mass per year, according to S&P Global Energy. Domestic facilities can refine only 25,000 metric tons of that material. The remainder, roughly 35,000 metric tons annually, has had nowhere to go except overseas. U.S. black mass exports reached 100,667 tonnes in 2025, up nearly 49 percent from the prior year. Benchmark Mineral Intelligence estimates that China alone had 1.5 million metric tons of excess black-mass refining capacity in 2025, giving Chinese and other Asian buyers both the appetite and the pricing power to outcompete any domestic refiner trying to secure feedstock.
This created a circular trap. A domestic refiner could not justify the capital expense of a hydrometallurgical plant when feedstock kept leaving the country for higher prices. Without a plant, the feedstock kept leaving. Three companies that tried to break this cycle paid a severe price: Massachusetts-based Ascend Elements filed for Chapter 11 bankruptcy in April 2026; Quebec-based Lithion Technologies entered creditor protection in November 2025; and Li-Cycle, which operated facilities across the United States, filed for bankruptcy in Canada in May 2025 before being acquired by commodity trading house Glencore. All three operated during a period when Asian buyers consistently outbid domestic processors for U.S. feedstock.
The economics are punishing even before competitive pressure enters the picture. Black mass alone can account for roughly 48 percent of hydrometallurgical recycling costs, and the rising market share of lower-energy-density lithium iron phosphate batteries, which contain no cobalt or nickel, makes the financial case harder still.
The July 30 Presidential Determination: Locking the Feed Inside the Country
On July 30, 2026, President Trump signed a determination under Section 101 of the Defense Production Act, a statute originally written for the Korean War, authorizing the Commerce Secretary to establish export controls over what the White House called "recoverable critical minerals and materials." The definition covers battery black mass, end-of-life rare-earth permanent magnets, swarf (the metallic grit generated when cutting tools machine hard materials), and other scrap containing critical minerals. Copper scrap was excluded because it is being addressed through a separate process.
The Bureau of Industry and Security followed six days later with a Temporary Final Rule prohibiting exports of black mass from the initial stage of lithium-ion battery recycling and tungsten waste and scrap without a BIS license or exception. The rule took effect on August 27. It requires U.S. persons selling covered black mass to allocate 100 percent of monthly sales to other U.S. persons. The speed of the rulemaking was notable: there were just two business days between the presidential determination and the pre-publication copy of the rule appearing in the Federal Register. BIS cited "urgent and compelling circumstances" to bypass the Administrative Procedure Act's standard advance-notice-and-comment process.
Tungsten illustrates why the government acted. No commercial tungsten mining has occurred in the United States since 2015, yet an estimated 4,700 tons of contained tungsten were exported in 2025, including waste and scrap. The metal is essential for armor-piercing munitions and cutting tools. Swarf, that grey, oily grit that comes off a machine tool, was until this summer one of the least regulated commodities leaving American ports. It now has its own clause in a Defense Production Act determination.
Electra Battery Materials CEO Trent Mell framed the logic precisely: "The United States has built the front half of the battery recycling processing infrastructure faster than the back half. This rule begins to correct that imbalance and could create the market conditions needed to crowd in private capital for U.S. refining." His company is pursuing a nickel sulfate refinery aimed at filling exactly that downstream gap. But Mell was equally direct about what the rule alone cannot accomplish: "Retaining feedstock is the first step. Financing and building the refining capacity to process it is the necessary second step."
The sequencing concern is real. If domestic hydrometallurgical facilities lack the acid-circuit capacity to absorb all the black mass that can no longer be exported, and if tungsten carbide processors cannot absorb all redirected scrap, the material will need to be stockpiled. That adds warehousing costs and financial pressure to recyclers that are already stressed. The export restriction changes the destination of the feedstock; it does not by itself create the refining capacity needed to use it.
What the GAO Found: Promising in Some Lanes, Years Away in Others
The Government Accountability Office published its assessment, titled "Critical Minerals: Reducing U.S. Import Reliance with Substitution and Recycling Technologies" (GAO-26-108687), on July 22, submitted to the House Committee on Energy and Commerce. The report focused on 11 minerals used in batteries and semiconductors drawn from the broader list of 60 critical minerals. Its findings varied significantly by application.
For batteries, the picture was cautiously optimistic on a short timeline. Established recycling technologies could reduce imports of copper, cobalt, lithium, and nickel within two to three years, the GAO found. Using lithium iron phosphate chemistry for stationary grid storage could reduce cobalt, manganese, and nickel imports in a similar window. The main obstacles are inadequate domestic refining capacity and the continued loss of usable feedstock to landfills and overseas buyers, precisely the problems the July 30 determination is designed to address.
For semiconductors, the picture was much darker. In five of the eleven minerals examined, including gallium, graphite, arsenic, indium, and manganese, the United States imports 100 percent of its supply. Recycling minerals from semiconductors faces a fundamental technical barrier: only a tiny quantity is recoverable from a typical electronic device, the circuits are miniaturized, and the minerals are tightly bonded to other materials. Electronics recycling at scale remains mostly in pilot-stage development. The GAO was equally skeptical about substitution in semiconductor applications, noting that "other materials do not perform as well as these critical minerals across the same conditions" and that substitutes are likely years from commercial maturity.
The report also identified a significant omission in its own scope: rare-earth magnets are excluded from the eleven-mineral analysis, leaving one of the most strategically sensitive parts of the supply chain outside the assessment's frame. The GAO laid out four broad policy options: establishing domestic manufacturing for substitute materials, building domestic recycling capacity, securing recycling inputs, and supporting research and development. What the report did not do was offer a timeline short enough to close gaps before the January 2027 defense procurement deadline.
The Lynas Texas Plant and the Cost of Building It Yourself
The challenge of closing the refining gap through new construction is illustrated most vividly by the planned Lynas Rare Earths processing facility near Seadrift, Texas, about 70 miles north of the site of Tesla's planned lithium refinery. The Defense Department has invested $288 million since 2021 in Lynas, an Australian company, to establish what would be the United States' first processor for heavy rare-earth elements including samarium, which is used in ultra-high-temperature magnets for spacecraft, missiles, stealth aircraft, and electronic warfare systems.
Originally scheduled to begin operations in 2026, the plant has not broken ground. The blocking issue is wastewater. Rare-earth refining consumes large quantities of water and can generate waste containing naturally occurring radioactive material. The Texas facility has been unable to resolve its wastewater disposal permit, and Lynas chairman John Humphrey stated as recently as November 2025 that no progress had been made on the matter. Lynas is simultaneously in talks with the Trump administration about cost escalation driven partly by the administration's own tariff agenda, which has raised the price of imported equipment and materials needed for construction.
The Lynas situation captures the compounding layers of difficulty facing any rare-earth processing project in the United States. Individual facilities can require hundreds of millions of dollars. Refining generates regulated waste streams that require local permitting. Chinese producers can influence global rare-earth prices through their output decisions, compressing the economic case for any competitor. And the absence of deep, liquid markets for most rare-earth products means standard project finance models, which rely on predictable revenue streams from commodity sales, function poorly. Government intervention has become the enabling mechanism rather than the exception, which is why the DOE's Loan Programs Office has assembled roughly $7.3 billion in federal support across major critical minerals projects. Yet an estimated $1 trillion to $1.5 trillion in board-approved private capital remains undeployed, held back primarily by permitting delays.
Building on my earlier analysis of the IEA's 2026 Global Critical Minerals Outlook, which found that the top three refining nations now control 86 percent of key mineral processing and that mining more ore will not fix a bottleneck located in refining and separation, the Lynas case confirms that the challenge is not a lack of intent or even of capital commitment. It is the accumulated friction of technical complexity, regulatory process, and the absence of the competitive infrastructure that makes projects bankable without government backing.
What Comes Next
The regulatory environment is likely to grow more restrictive before it grows more permissive. Less than a month after the BIS black mass and tungsten rule took effect, President Trump signed Executive Order 14420 banning imports of foreign battery energy storage systems, a signal that additional executive actions targeting battery materials are probable. In June 2026, China added MP Materials and USA Rare Earth, two government-backed U.S. rare-earth companies, to its export control list, raising the geopolitical stakes further.
The federal funding pipeline is real but not yet decisive. The DOE has announced a $500 million Battery and Processing notice of funding opportunity targeting U.S. processing capacity, with project selections expected later this year. A separate $134 million program targets commercial-scale rare-earth recovery from unconventional feedstocks including mine tailings and e-waste. The ANSI workshop that I covered last week, examining standardization gaps in U.S. critical minerals traceability and labeling, represents another piece of the framework that processing investors will need before they can confidently qualify products for defense and commercial buyers.
The January 2027 defense procurement deadline is the most immediate forcing function. Defense contractors must stop purchasing certain critical materials from China, Russia, Iran, and North Korea by that date, but the domestic supply alternatives are not yet at commercial scale. Benchmark Mineral Intelligence estimates it could take U.S. black mass refiners until the end of the decade to eliminate the current processing gap. The IEA projects that announced global recycling capacity in 2030 could be seven times available feedstock, but that in Europe and the United States specifically, announced capacity covers only about 30 percent of feedstock availability by 2040, when the supply of end-of-life batteries finally catches up with investment.
The core lesson of the current moment is that supply chain resilience requires both halves of the equation to be built in parallel: the facilities that refine, and the policies and markets that guarantee those facilities a steady feed of material. Export controls address the second problem by regulatory force. Closing the first problem will require sustained capital, permitting reform, and time that the policy calendar is not leaving much room to waste.