Three major developments in July and August 2026 -- a joint Duke University and University of Cape Town report, a GAO technology assessment, and a $75 million DOE pilot-plant programme -- independently arrive at the same conclusion: the United States does not have a mining problem, it has a processing problem. Together, they reveal how academic research, federal audit findings, and industrial grant-making are converging on midstream separation, recycling, and conversion as the decisive terrain in the competition to reduce import dependence on critical minerals.
Introduction
Three institutions with very different mandates published their findings within weeks of each other this summer, and each reached the same uncomfortable conclusion. A joint academic report from Duke University and the University of Cape Town, released August 13, flagged midstream processing as the world's most consequential industrial chokepoint. A GAO technology assessment, published July 22, found the United States has made almost no progress reducing import reliance on the minerals that power batteries and semiconductors. And a $75 million DOE grant programme, announced July 1, placed its entire bet on pilot-scale recovery from coal-derived feedstocks rather than on opening new mines.
The convergence is not coincidental. Each document reflects a growing consensus in research, policy, and industry that the critical minerals challenge has been misframed. The debate has centred on geology: where the deposits are, who controls the reserves, how quickly new mines can open. But the binding constraints lie downstream of the mine gate, in the separation plants, chemical conversion facilities, and recycling infrastructure that transform raw concentrates into the battery-grade chemicals and separated rare earth oxides that manufacturers can actually use.
China understood this years ago. Its dominance is not primarily a function of geological luck. It is the product of deliberate, sustained investment in the processing stages that most other countries were happy to outsource. The three developments examined here represent the clearest signal yet that the United States, and its academic and multilateral partners, are finally internalising that lesson -- even as the gap between understanding and industrial capacity remains formidably wide.
China Controls the Chokepoint, and the Numbers Are Stark
The Duke and UCT report leads with a statistic that frames everything else: China controlled approximately 91% of global magnet rare earth separation in 2024. That figure, drawn from IEA data, is not a measure of mining dominance. It is a measure of processing dominance, and it explains why the expansion of rare earth mining outside China has done relatively little to reduce strategic vulnerability. A tonne of rare earth concentrate shipped from a new Australian or American mine still travels, in most cases, through a Chinese separation facility before it can become a usable oxide or alloy.
The GAO's numbers tell a parallel story for the semiconductor supply chain. The United States is at least 41% import-reliant on each of the eleven battery and semiconductor minerals the agency examined. For five of them -- arsenic, gallium, graphite, indium, and manganese -- domestic supply is effectively zero, with 100% of consumption met by imports. China produces 99% of the world's gallium, a semiconductor-critical element that Beijing targeted with export controls in 2023 and that the DOE's new coal-byproduct programme specifically aims to recover domestically.
The IEA's 2026 Global Critical Minerals Outlook adds a structural dimension to these snapshots. The average market share of the top three refining nations across copper, lithium, nickel, cobalt, graphite, and rare earth elements rose to 86% in 2024, up from around 82% in 2020. Almost all of that supply growth came from the single top supplier in each category, which for nearly every mineral except nickel means China. Concentration in processing is increasing, not decreasing, even as mining diversifies. That divergence is the strategic problem in its most precise quantitative form.
What Academic Research Says About Where Investment Should Go
The Duke and UCT report, described by its authors as a first-of-its-kind mapping exercise, was designed to inform the Council for Critical Minerals Development in the Global South, a collaborative platform created in anticipation of surging commodity demand. But its recommendations carry implications well beyond the Global South. The report identifies five investment priorities for critical mineral processing, all targeting technology readiness levels 5 through 7: capturing value at the processing chokepoint; funding the pilot-to-demonstration stage; moving beyond traditional evaporation ponds for lithium; reducing the energy and carbon cost of lithium conversion; and treating rare earth separation and recycling as co-equal priorities.
Direct lithium extraction deserves particular attention here. DLE technologies can compress lithium recovery time from the twelve to twenty-four months required by conventional evaporation ponds to a matter of hours, while potentially recovering more than 90% of available lithium compared to the 30% to 50% typical of pond-based systems. The DOE has recognised this potential: a May 2026 award to Ohio University specifically funded the development of coal-derived lithium-selective electrodes for DLE from domestic waste streams including produced water and acid mine drainage.
Yet the Duke and UCT report is careful not to oversell the technology. DLE performance varies significantly by brine chemistry, including magnesium content, temperature, and contaminant composition. Hybrid systems combining DLE with conventional methods may be more realistic in the near term than pure DLE deployments. The authors extend this caution to recycled rare earth feedstocks as well, noting that the volume of retired NdFeB magnets from electric vehicles and wind turbines will remain limited until around 2030, when the first generation of EV drivetrains approaches end-of-life at scale. The implication is that some of the most promising recycling pathways are structurally supply-constrained for at least several more years, regardless of processing technology readiness.
The report's most pointed contribution may be its critique of how the commercialisation gap is typically framed. Capital, the authors argue, is necessary but not sufficient to move technologies from TRL 5-7 into commercial operation. Feedstock reliability, customer qualification, and repeatable output are equally essential. A separation facility cannot attract long-term contracts without demonstrating consistent product quality across variable input streams. A battery recycler cannot scale without a reliable collection infrastructure. These are institutional and logistical challenges, not scientific ones, and they require different kinds of intervention than grants and loans alone can provide.
GAO and the Executive Order: When Audit Findings Become Industrial Policy
The GAO technology assessment published July 22 is a sober document. Its central finding on battery recycling is optimistic in relative terms: the technology to extract and refine critical minerals from spent lithium-ion batteries is mature, and the agency assessed that domestic battery recycling could reduce imports of copper, cobalt, lithium, and nickel within two to three years. The obstacle is not the science. The United States lacks the collection infrastructure, the sorting logistics, and the hydrometallurgical processing capacity to act on that technological maturity at meaningful scale.
As I reported in August, the GAO's findings on recycling infrastructure arrived at a moment of particular fragility for the domestic recycling industry, with high-profile bankruptcies at Li-Cycle and Ascend Elements having already removed significant processing capacity from the market. The agency recommended four policy responses: establish domestic manufacturing capacity for viable substitutes; build domestic recycling infrastructure; secure input streams for recycling through collection and sorting programmes; and coordinate across agencies and industry. These are not radical proposals, but they require sustained institutional commitment of a kind that has been historically difficult to sustain across budget cycles and administrations.
Eight days after the GAO report's release, on July 30, President Trump signed a presidential memorandum delegating authority under the Defense Production Act to Commerce Secretary Howard Lutnick to restrict exports of recoverable critical minerals and materials. The covered materials include black mass, end-of-life rare earth permanent magnets, swarf, and tungsten scrap from spent batteries and industrial processes. The Bureau of Industry and Security subsequently issued a Temporary Final Rule imposing a 100% domestic sales requirement on these materials, effective August 27.
The sequencing matters. The GAO identified the infrastructure gap; the executive order created a commercial imperative to fill it by removing the export option that had allowed domestic recyclers to monetise their output without building domestic processing capacity. The policy logic is coherent: if you cannot export the black mass, you have to refine it at home, which creates market pressure for investment in the hydrometallurgical infrastructure the GAO found to be absent. Whether that pressure will translate into buildout at the required speed, given the existing industry distress and the January 2027 deadline for defence contractors to cease purchasing specified minerals from China, Russia, Iran, and North Korea, remains an open question.
Coal as a Critical Minerals Feedstock: A Novel Pathway with Pilot-Scale Ambitions
The DOE's $75 million Mines and Metals Capacity Expansion programme, announced July 1, represents a different kind of federal intervention: not a policy restriction but a direct technology investment targeting a feedstock source that most critical minerals discussions overlook entirely. The five selected recipients -- the University of North Dakota, Valor Metals, CONSOL Innovations, American Resources Corporation, and Peabody Energy -- will develop pilot-scale facilities to recover rare earth elements, germanium, gallium, and aluminium from coal and coal-derived industrial streams including ash and refuse.
The coal-byproduct pathway is strategically interesting for several reasons. It draws on existing industrial sites and workforces, reducing greenfield development risk. It targets germanium and gallium specifically, the two minerals that China weaponised with export controls beginning in 2023 and which the GAO flagged as 100% import-dependent. And it connects the Trump administration's political commitment to the domestic coal sector with the supply-chain security agenda, creating a constituency for the programme that crosses traditional industrial and national-security coalitions.
The awards sit within a broader programmatic context. They follow the DOE's August 2025 announcement of nearly $1 billion for critical minerals technologies across the full supply chain, and form part of a larger $275 million initiative announced in late 2025 specifically for byproduct recovery at domestic industrial facilities. The National Energy Technology Laboratory will oversee all five projects, providing federal scientific backstop to what are fundamentally industry-led deployments.
Perhaps the most technically novel element of the broader programme is not the coal pathway itself but the Argonne National Laboratory biohydrometallurgy project, which targets recovery of critical minerals from black mass using a next-generation bioleaching process based on Argonne's arrested methanogenesis technology, combined with Mycocycle's patented bioaccumulation approach to concentrate cobalt, lithium, and nickel. The project aims to achieve greater than 80% key mineral recovery from recycled batteries at continuous pilot scale of 14 litres. Biological processing pathways have received considerably less commercial attention than hydrometallurgical or pyrometallurgical routes, but they offer potential advantages in energy consumption and reagent costs that could matter significantly at commercial scale. Idaho National Laboratory has separately demonstrated an electrochemical membrane reactor achieving more than 95% recovery of nickel and cobalt from spent battery leachates using only water, air, and electricity, a process that produces reusable acid as a co-product.
The Coordination Problem: Why Capital, Technology, and Policy Are Not Enough on Their Own
The three developments examined here share a subtext that deserves explicit attention: the gap between knowing what needs to be done and having the institutional machinery to do it. The Duke and UCT report calls explicitly for an international public registry of critical mineral processing technologies, including technology readiness level evidence and pilot results, to reduce duplication of effort and direct capital toward the most advanced projects. The authors note that no such registry currently exists, and that the absence of shared evaluation frameworks is causing significant wasted investment as countries and companies independently re-learn lessons that others have already paid to discover.
The GAO echoes this coordination concern from a domestic perspective. Despite the maturity of battery recycling technology, the U.S. lacks not only processing infrastructure but also the collection and logistics networks that would feed material to processing facilities. Infrastructure for sorting, transporting, and storing end-of-life batteries and manufacturing scrap is a prerequisite for any recycling programme, and it requires coordination among municipal waste systems, manufacturers, retailers, and federal agencies that has not yet been achieved at scale.
The patent landscape adds another layer of complexity. In rare earth magnet production, grain boundary diffusion technology, which enhances permanent magnet performance and reduces heavy rare earth content, is highly patented and serviced by only one equipment supplier outside China. That supplier reportedly charges more than ten times the price of Chinese-manufactured equivalents, with longer lead times. This is not a technology readiness problem; the science is known. It is an industrial ecosystem problem, of the kind that the IEA projects will leave a gap between planned rare earth refining capacity and actual downstream magnet manufacturing capacity of roughly 22,000 tonnes by 2035.
Building on my analysis of the black mass export ban in August, the immediate commercial pressure created by that policy directive now intersects with these structural gaps in a way that creates both urgency and risk. Urgency, because the export restriction removes the path of least resistance for domestic recyclers and forces investment decisions. Risk, because if hydrometallurgical processing capacity cannot be built fast enough to absorb the restricted material flows, the policy could damage recyclers without generating the domestic supply it was designed to create. The race between policy ambition and industrial buildout timelines is the central tension in the American critical minerals story heading into 2027.
Conclusion: Three Documents, One Diagnosis
What makes the convergence of the Duke/UCT report, the GAO assessment, and the DOE coal-byproduct programme significant is not that any one of them breaks entirely new ground. Analysts have argued for years that processing, not mining, is the strategic bottleneck. What is new is the simultaneity and specificity. Within six weeks, the United States government audited its import dependence and found it essentially unchanged since 2021, committed $75 million to pilot-scale processing from a novel feedstock category, and signed an executive order turning recycling into a national security obligation rather than a voluntary commercial activity. An academic collaboration spanning two continents published a framework for thinking about midstream investment that translates the same diagnosis into actionable priorities for both policy and capital allocation.
The IEA's projection that the announced pipeline of rare earth refining outside China totals nearly 50,000 tonnes of capacity by 2035, while planned downstream magnet manufacturing stands at only 18,000 tonnes, illustrates the structural imbalance that persists even in an optimistic buildout scenario. More separation capacity without more magnet manufacturing is not a supply chain; it is a longer version of the same dependency. The Duke and UCT recommendation to treat separation and recycling as co-equal priorities, rather than sequential steps, is a recognition of this systemic risk.
The near-term watchlist that emerges from these three documents is concrete. Will the Council for Critical Minerals Development establish a public technology-readiness registry with comparable pilot data by the end of 2026? Will DLE developers publish sustained recovery and reagent consumption data on commercial brines in 2027? Will Argonne's biohydrometallurgy process achieve its greater-than-80% recovery target at continuous pilot scale? And will the Defense Production Act restrictions on black mass exports generate the domestic processing investment they were designed to incentivise, or will they instead accelerate the consolidation and attrition of an already-stressed recycling industry?
The answers will determine whether the summer of 2026 is remembered as the moment that American and allied critical minerals strategy turned decisively toward processing self-sufficiency, or as a period of policy activation that outpaced the industrial capacity to act on it. The science is increasingly ready. The question, as it has been throughout this story, is whether the infrastructure can follow fast enough to matter.