In July 2023, China’s Ministry of Commerce announced export controls on gallium and germanium—two metals most people have never heard of, both of which are essential to semiconductor manufacturing, fiber optics, infrared optics, solar cells, and military hardware. Exporters were required to apply for licenses, disclose end-use information, and identify the final destination of every shipment. The result was immediate: Chinese gallium exports dropped from 6,876 kilograms in July 2023 to 227 kilograms in October 2023. Germanium fell from 7,965 kilograms to 590 kilograms in the same period. European prices for both metals nearly doubled within a year. By May 2025, the Rotterdam price of gallium had hit $687 per kilogram—an increase of over 150 percent from pre-control levels. Meanwhile, gallium prices inside China fell, because domestic oversupply had nowhere to go. Beijing was sitting on cheap material it refused to sell, watching the rest of the world scramble.

In December 2024, China escalated to an outright ban on gallium and germanium exports to the United States, along with antimony and superhard materials—a direct retaliation for the Biden administration adding 140 Chinese semiconductor companies to the Entity List. The ban was suspended in November 2025 as part of bilateral trade negotiations, with general licenses issued through November 2026. But the legal framework remains intact. The controls can be reactivated at any time. The message was delivered: China controls 98 percent of global gallium production and 60 percent of germanium, and it’s willing to use that leverage the same way OPEC uses oil—as a strategic instrument with a valve.

What gallium and germanium actually do

These aren’t rare earth elements—they’re critical minerals with their own supply chain vulnerabilities and their own reasons for mattering.

Gallium’s primary semiconductor application is gallium nitride (GaN), a wide-bandgap material that handles higher voltages, operates at higher temperatures, and switches faster than silicon. GaN-based chips are more efficient and more durable than their silicon equivalents, which is why they’re displacing silicon in power electronics, fast chargers, 5G base stations, radar systems, and military communications hardware. Gallium arsenide (GaAs) is the backbone of radio-frequency chips in smartphones—the components that connect your phone to a cell tower use gallium, not silicon. Every 5G phone on earth contains gallium-based semiconductors. LED lighting runs on gallium compounds. The photovoltaic industry uses gallium in high-efficiency multijunction solar cells for spacecraft and concentrated solar installations.

Germanium’s niche is narrower but equally non-substitutable. Its high electron mobility makes it essential for high-speed transistors. It’s the material of choice for infrared optical components—night vision goggles, thermal imaging cameras, missile guidance systems, satellite sensors. Fiber-optic cables use germanium-doped silica to minimize signal loss over long distances, which means the physical infrastructure of the internet—the glass cables that carry data between continents—depends on a material that one country dominates. An F-35 fighter jet’s infrared targeting system, the fiber-optic backbone connecting data centers, and the night vision goggles worn by infantry all share a supply chain vulnerability that runs through Beijing.

How China got here

Gallium doesn’t occur in nature as a primary ore. It’s a byproduct of aluminum smelting—extracted from bauxite processing residues at concentrations so low that recovery is only economical if you’re already running an aluminum smelter at scale. China produces more aluminum than any other country on earth, which means it generates more gallium-bearing waste streams, which means it dominates gallium production not because it set out to corner the market but because it cornered the upstream industry that gallium falls out of. The same pattern: whoever processes the most bauxite gets the most gallium, and China processes the most bauxite.

Germanium is slightly more distributed—China controls 60 percent rather than 98 percent—but the refining infrastructure is similarly concentrated. Global annual demand for gallium is below 700 metric tons, a fraction of markets like copper (25.9 million tons) or nickel (3.1 million tons). The small market size is itself a strategic advantage for Beijing: it’s easier to manipulate a 700-ton market than a 25-million-ton market. Small disruptions in supply produce large price swings, which gives China leverage that’s disproportionate to the tonnage involved.

The controls weren’t random. They were calibrated responses to specific American actions. The August 2023 licensing requirement answered the initial rounds of U.S. chip export controls. The December 2024 ban answered the Entity List expansion. The November 2025 suspension was part of a broader negotiated pause. Each escalation was timed, proportional, and reversible—designed to demonstrate capability without triggering a full decoupling. China has been explicit that the controls are not permanent policy. They’re a deterrent. The message: if you restrict our access to advanced chips and lithography equipment, we restrict your access to the materials those chips are made from.

The rerouting problem

The ban is leakier than it looks. Stimson Center analysis of Chinese customs data found that in 2024, the quantity of germanium exported to the United States fell by approximately 5,900 kilograms—almost exactly the amount by which germanium exports to Belgium increased (6,150 kilograms). The combined total to both countries was essentially flat across 2023 and 2024. The material appears to be flowing through third-country intermediaries that reimport it to the United States without Chinese end-use restrictions applying.

For gallium, the picture is more complicated because Canada and Germany have secondary gallium production from their own aluminum smelting operations, making it harder to distinguish genuine non-Chinese supply from rerouted Chinese material. The U.S. Census Bureau records imports by the country that produced the material unless it underwent “substantial transformation” in a third country—a classification that creates ambiguity about whether Belgian-processed germanium originally sourced from China counts as Belgian germanium.

The rerouting doesn’t eliminate the vulnerability. It adds cost, uncertainty, and transit time. It creates a supply chain that depends on Beijing’s tolerance of the workaround, which can be withdrawn. And it doesn’t address the fundamental concentration: if China decided to enforce end-use controls across all destinations—not just the United States—the third-country channels would close.

What the West is building

The response has been faster than for rare earths but still measured in years rather than months.

MTM Critical Metals is building a facility in Texas to extract gallium from industrial scrap, scheduled to begin operations in early 2026—an unusually fast timeline for critical mineral projects. The company is reportedly negotiating binding agreements with Indium Corporation that include minimum price floors designed specifically to protect against Chinese market manipulation. Canada’s 5N Plus and Germany’s PPM Pure Metals have secondary production from domestic aluminum operations. Japan has invested in recycling infrastructure to reduce import dependence.

The EU’s Critical Raw Materials Act targets reducing dependency on single-source suppliers. The CHIPS Act allocated funding for domestic semiconductor material infrastructure. But the structural problem is the same one that affects rare earth diversification: building new supply takes years, the markets are small enough that Chinese pricing can undercut new entrants at will, and the byproduct economics mean you can’t produce gallium at scale without producing aluminum at scale, which means diversifying gallium supply requires diversifying an entire upstream industry.

Gallium prices inside China are lower than international prices because the domestic surplus can’t be exported. If China eventually lifts all controls, the price crash could make every Western diversification project uneconomic overnight—the same dynamic that has killed rare earth mining ventures outside China for two decades. Beijing doesn’t need to maintain the export ban permanently. It just needs the threat of reimposing it, combined with the ability to flood the market with cheap material if Western alternatives get too close to viability. The weapon isn’t the embargo. It’s the optionality.

What it tells you about the next decade

Gallium and germanium are test cases for a broader pattern. China identified that its dominance of bauxite processing gave it accidental control of a small but critical material, weaponized that control in response to American technology restrictions, calibrated the escalation to demonstrate capability without provoking full decoupling, and then suspended the controls as a negotiating chip—while keeping the legal framework active for reimposition. Every element in the critical minerals portfolio—antimony, graphite, rare earth processing technology, medium and heavy rare earths—has been subject to the same playbook in sequence since 2023.

The progression: rare earth processing dominance (established over decades) → gallium and germanium controls (2023) → antimony controls (2024) → rare earth processing equipment and technology controls (October 2025, suspended November 2025). Each step expands the scope. Each suspension is temporary and conditional. The architecture for comprehensive export controls across the entire critical minerals supply chain is built. It’s just not fully activated—yet.

We cover gallium and germanium alongside the helium shortage, rare earth recycling, and the full landscape of critical materials that underpin modern technology across our Rare Earth Elements course—including why the most strategically important metals in the semiconductor supply chain are ones most people can’t name, produced as byproducts of industries most people don’t think about, and controlled by a country that knows exactly what it has.

Less than 1 percent of rare earth magnets currently come from recycled sources. In the United States, the figure is under 1 percent. Almost all spent neodymium-iron-boron magnets—the permanent magnets inside electric vehicle motors, wind turbines, hard drives, headphones, MRI machines, and F-35 fighter jets—end up in landfills or low-grade scrap. Every one of those magnets contains neodymium, praseodymium, and often dysprosium, mined at enormous environmental cost, refined predominantly in China, and then buried in the ground a second time when the product they powered reaches end of life. The circular economy for rare earths is, in 2026, essentially a concept with a handful of pilot plants attached to it. The technology to recycle rare earths exists. The economics, logistics, and collection infrastructure to do it at scale do not.

This matters more than it used to. Global demand for neodymium-iron-boron magnets is increasing at over 15 percent annually, driven by the energy transition—electric vehicles use up to 4 kilograms of rare earths per motor, and a single large offshore wind turbine can contain 200 kilograms. China controls 60 to 90 percent of global rare earth mining and refining. The EU’s Critical Raw Materials Act requires 25 percent of critical raw materials to come from recycling by 2030. The gap between that target and the current 1 percent recycling rate is not a gap that incremental improvement will close. It’s a structural problem with structural causes.

Why recycling rare earths is hard

Traditional mining produces up to 2,000 tons of toxic waste per ton of rare earth elements extracted. You’d think that alone would make recycling the obvious alternative. The reason it isn’t comes down to three problems that compound each other.

The first is physical access. Neodymium magnets are embedded deep inside products—glued into electric motors, bonded into hard drive assemblies, sealed inside speaker housings, integrated into sensor systems. Extracting them requires disassembly of the product, which is labor-intensive, sometimes destructive, and rarely designed for. A car manufacturer optimizes an electric motor for performance and cost, not for magnet recovery 15 years later. The magnets are small relative to the product that contains them, which means the labor cost of extraction can exceed the value of the recovered material. And neodymium magnets are strongly magnetized, which makes handling them in bulk—particularly from large EV motors—a safety hazard requiring specialized equipment.

The second is chemical complexity. Recovered magnets are contaminated with coatings, adhesives, and other metals that must be removed before the rare earth elements can be reprocessed. Different products use different magnet compositions—the ratio of neodymium to dysprosium varies by application, complicating standardized recycling processes. Neodymium magnets are also sensitive to oxidation; if their protective coatings are damaged during extraction, the material quality degrades, and oxidized rare earth elements are harder to refine back to usable purity.

The third is economic competition with virgin material. China’s dominance of rare earth mining and refining means that primary rare earth oxides are available at prices that recycled material struggles to undercut, particularly when the collection, disassembly, and reprocessing costs of recycling are factored in. In Europe, recycling is currently more expensive than importing raw material from China. The economic case for recycling depends on either the price of virgin material rising (which China can manipulate through export controls) or the cost of recycling falling (which requires scale that doesn’t yet exist). Strategic necessity—reducing dependence on a single supplier—is driving investment, but strategic necessity doesn’t automatically translate into competitive unit economics.

What actually works

The recycling technologies exist, and some of them work well at laboratory and pilot scale.

Hydrogen decrepitation—the HPMS process—injects hydrogen gas into sintered neodymium magnets, cracking them into powder without harsh chemicals. The process preserves the alloy composition, allowing the powder to be re-sintered directly into new magnets. HyProMag, a UK company expanding into the United States, uses this method and reports that its hydrogen-processed powder matches new-magnet grades while using 90 percent less energy than manufacturing from virgin material. Hydrometallurgical methods dissolve magnets in acid solutions to separate individual rare earth elements, which can then be refined to high purity. The SEEE process developed by Kyoto University has achieved 96 percent recovery for neodymium and 91 percent for dysprosium at purities above 90 percent.

A 2025 paper in PNAS described flash Joule heating combined with chlorination—a single-step process that achieves greater than 90 percent purity and greater than 90 percent yield while reducing energy consumption by 87 percent, greenhouse gas emissions by 84 percent, and operating costs by 54 percent compared to traditional hydrometallurgy. The process eliminates water and acid use entirely. REEcycle, a Texas-based company, has developed an electrochemical separation process claiming 99.8 percent recovery efficiency. Phoenix Tailings uses acid-free leaching and molten salt electrolysis to recover rare earths from mining waste at pilot scale, targeting thousands of tonnes per year. Canada’s Cyclic Materials, backed by investment from BMW and Jaguar Land Rover, achieves over 90 percent rare earth recovery from EV motors and electronics.

In Italy, startup RarEarth raised €2.6 million to build the country’s first neodymium magnet factory using recycled e-motor waste. The UK’s CREEM consortium—£11 million, led by Ionic Technologies, with participants including Ford, Bentley, and Wrightbus—aims to build scalable recovery loops for end-of-life EV magnets. Apple has invested $500 million in expanding recycling infrastructure that includes rare earth recovery from consumer electronics. The REE4EU project has produced magnets containing over 99 percent recycled material.

The technology portfolio is genuine: hydrogen processing, hydrometallurgy, pyrometallurgy, flash Joule heating, electrochemical separation, bio-adsorption, ion chromatography. Multiple methods achieve recovery rates above 90 percent at purities sufficient for remanufacturing. The problem isn’t that recycling can’t be done. It’s that it can’t yet be done at the scale, cost, and collection efficiency required to make a meaningful dent in the 1 percent recycling rate.

The collection problem beneath the technology problem

Even if every recycling technology worked perfectly at industrial scale tomorrow, the system would still face a bottleneck that no amount of chemistry can solve: getting the magnets out of the products and into the recycling plants.

An electric vehicle sold in 2025 won’t reach end of life for 10 to 15 years. The wind turbines being installed now have operational lifespans of 20 to 25 years. The rare earth magnets inside these products are, from a recycling perspective, locked in a time capsule that won’t open until the 2035–2050 timeframe. The feedstock available today comes primarily from manufacturing scrap (the dust and shavings produced during magnet shaping—called swarf), end-of-life consumer electronics (hard drives, speakers), and decommissioned industrial equipment (MRI machines, factory motors). These are real sources, but they’re diffuse, low-volume relative to the magnets that will eventually come from the EV and wind turbine fleets, and require collection logistics that don’t yet exist at scale.

IDTechEx predicts that rare earth magnet recycling will increase 6.5 times over the next decade and could represent up to 10 percent of global supply by 2036. Ten percent by 2036. Not 25 percent. Not 50 percent. The EU’s target of 25 percent recycled critical raw materials by 2030 is, by independent industry analysis, aspirational rather than achievable on the current trajectory. The honest timeline: recycling will become a meaningful supplement to primary mining within the decade, and a significant supply source by the mid-2030s when the first wave of end-of-life EVs and wind turbines begins generating large-volume magnet feedstock. It will not replace mining. It will reduce the rate of growth in mining demand, which—given that mining produces 2,000 tons of toxic waste per ton of extracted rare earths—is worth doing even if the circular economy remains incomplete.

The rare earth recycling problem is, at bottom, a timing problem. The technology is arriving before the feedstock. The products that contain the largest volumes of rare earth magnets haven’t reached end of life yet. The circular economy for critical minerals is being built during the interval between when the products were sold and when they’ll be discarded—an interval measured in decades, during which the world’s dependence on Chinese mining continues, the environmental cost of extraction accumulates, and the collection infrastructure that will eventually be needed is either built now or scrambled together later.

We cover rare earth recycling alongside neodymium supply chains, the helium shortage, and the full landscape of critical materials that underpin modern technology across our Rare Earth Elements course—including why the circular economy for the most important magnets on earth is stuck at 1 percent, and what has to change before it isn’t.