There Is No Such Thing as a Gallium Mine

In short
Key takeaways
- Gallium is captured from Bayer process liquor left after alumina refining, so production follows aluminium capacity rather than geology
- The USGS puts US net import reliance on gallium at 100 percent, which means no domestic primary production at all
- China suspended its US-directed ban from 9 November 2025 until 27 November 2026, not permanently
- Chinese unwrought gallium exports fell about 94 percent from 2024 levels and had not recovered as of May 2026
- The older licensing controls on gallium and germanium were never removed, so the expiry may change less than it appears to
Start with the process, because the process is the story. Bauxite is digested in hot sodium hydroxide. Aluminium hydroxide comes out of solution and goes on to become alumina, and then aluminium. What stays in the spent liquor is everything else that was in the ore, including gallium, at concentrations low enough that no mine would ever be dug for it. The gallium is pulled back out of that liquor by solvent extraction and electrolysis, and it exists as a product only because an alumina refinery already exists next to it.
There is no gallium mine. There is no ore body to prospect, no deposit to permit, no seam to open. Any country that wants gallium independence is not shopping for a mine. It is shopping for an aluminium industry, and then for the small, unglamorous circuit that captures a trace element out of a waste stream.
That single fact reorganises the entire export control story, and it explains a number that otherwise looks like an accounting error.
One hundred percent
The USGS Mineral Commodity Summaries put United States net import reliance on gallium at 100 percent. Not 98. Not 95 with a domestic sliver. One hundred.
Read that carefully, because it says more than "the US buys gallium abroad." It says there is no domestic primary production to lean on. And because gallium is only produced by refineries that are already making alumina, the absence of a gallium industry is really the absence of the upstream industry that would carry it. The import reliance figure is not a procurement problem with a procurement solution. It is a statement about which industrial base exists.
A mining permit does not fix this. A strategic reserve does not fix this. A stockpile buys months of inventory against an interruption. It does not create a producer, and a producer cannot be created without an alumina refinery attached to it, built at refinery capital cost, to serve a market whose entire American import bill is measured in tens of millions of dollars.
Two legal layers, often confused
China did not impose one gallium control. It imposed at least two, at different times, for different reasons, and the distinction is where most coverage goes wrong.
The first layer is a licensing regime applied to gallium and germanium exports generally, imposed in 2023 and expanded since. It governs who may ship the material anywhere, on what conditions, and to whom.
The second layer is specific and political. On 3 December 2024, China's Ministry of Commerce issued Notice 2024 No. 46, banning exports to the United States of gallium, germanium, antimony and superhard materials, on the stated grounds of dual military and civilian use.
In October 2025, as part of the truce reached at Busan, Beijing suspended that second layer. The suspension took effect on 9 November 2025 and runs until 27 November 2026. Pillsbury's reading of the notices is that the suspension covered six dual-use announcements issued on 9 October 2025 and part of Notice 2024 No. 46, which is a precise and slightly awkward formulation, and the awkwardness matters. Only part of the December 2024 ban was suspended. Which part was left in place, and therefore which part is scheduled to return, is not something the public record states clearly.
So the cliff edge on 27 November 2026 is real but it is narrower than the headlines will make it sound. Removing the US-directed prohibition does not touch the licensing regime underneath, because that licensing regime was never lifted. China's own announcements did not withdraw it, and analysts note that Beijing never publicly announced suspending the earlier controls at all.
What the suspension actually produced
The test of a suspension is volume, and the volume did not come back.
Chinese exports of unwrought gallium fell roughly 94 percent from 2024 levels, and as of May 2026 had not recovered, according to CSIS. The Council on Foreign Relations assessed the post-Busan period in September 2026 and found that exports of key materials had not rebounded to original levels, that Beijing had approved far fewer general licences than the truce implied, and that the material which did move came with conditions attached.
The licence procedure itself became the instrument. Analysts describe information requirements used to gather intelligence about supply chains and defence industrial networks, categorical rejections for defence companies, and the prospect of future licences used to extract concessions in unrelated policy areas.
The demonstration case is Japan. After Prime Minister Sanae Takaichi said Japan could defend Taiwan if it were attacked, Chinese flows of gallium and several rare earths to Tokyo fell to near zero through 2026, on customs figures reported by Reuters and Kyodo. Gallium shipments specifically went to zero between February and April, per the Sasakawa Peace Foundation's tracking. That is the mechanism in its clearest form: not a tariff, not a blanket embargo, but a licence that can be withheld because of a sentence.
The market is tiny, and that is the trap
US imports of gallium metal run to roughly 15 million dollars a year on USGS summary figures. The material feeds gallium nitride power devices, radio frequency electronics, and gallium arsenide used in photonics and radar. Those downstream applications are worth vastly more than the input that enables them.
This ratio is why export controls on byproducts work so well, and why they are so hard to answer.
No company builds a two billion dollar refinery to serve a fifteen million dollar market. The market is too small to justify the capital and too important to do without. That is not a market failure that resolves when prices rise. It is a permanent structural condition that has to be managed politically, because the normal price signal cannot summon capacity that the market is too small to pay for.
It also means the American exposure is not proportional to the dollar value. A fifteen million dollar import line that gates the radio frequency front end of the defence industrial base is a small number carrying a large consequence.
What to watch after 27 November
The useful thing about a dated event is that it can be checked. Four things, in order.
First, the notice. A new Ministry of Commerce announcement that either extends the suspension, converts it to a permanent removal, or says nothing at all. Silence matters, because under an automatic expiry the ban returns by default.
Second, Chinese customs data for November and December 2026. Watch the US destination line specifically. If it stays near zero through both months, then the licensing layer was always the binding constraint and the expiry was a non-event. If it returns to 2024 levels, the suspension was doing real work.
Third, US import data. If American imports rise while direct Chinese shipments do not, the material is arriving through a third country. The Stimson Center documented exactly that pattern for gallium and germanium in 2025. Any claim that the controls bind has to survive that possibility.
Fourth, price. A sharp move in the gallium benchmark in the first fortnight of December is the market's own verdict on whether anything changed.
Where this could be wrong
The most important gap is the scope of the suspension. The public record does not say which part of Notice 2024 No. 46 was suspended, and that determines what 27 November actually means. A post that reports the ban "returning" without establishing which materials are affected would be repeating the error this piece is trying to correct.
The underlying Chinese instruments were not read in the original. Notice 2024 No. 46 and the November 2025 suspension are described here through CSIS, Pillsbury, Reuters and CSET summaries rather than the Chinese texts. The document numbers and the expiry date should be confirmed against those texts.
The 100 percent reliance figure and the 15 million dollar import value were both checked against the USGS 2026 gallium summary. The rest of that chapter was not read, and the trade quantities in it are estimates.
The 94 percent export decline is attributed to CSIS. Chinese customs data was not queried directly for this piece.
Transshipment is unresolved. If Chinese gallium is reaching the United States through third countries at scale, then effective dependency is lower than the headline figure and the controls are weaker than they appear.
And the artificial intelligence framing needs care. Gallium matters to gallium nitride power devices and radio frequency components, which sit downstream of data centre power conversion and telecommunications. It is not a primary input to AI compute in the way that transformers or turbines are. The buildout is downstream of this supply chain, not the reason it matters.
And on length: this was built from a research dossier rather than fresh reporting, and it runs shorter than the usual 2,000 words for that reason. The gaps above are named rather than filled.
The part that does not expire
Whatever happens on 27 November, the structural fact stays where it is. The United States imports all of its gallium because it does not have the industrial base that produces gallium as a side effect, and no policy aimed at gallium alone will change that. The choke point is not a mine in another country. It is an absence at home.
The date is worth watching. The absence is worth writing about.
On this page
Sources
- Mineral Commodity Summaries 2026, gallium chapterUnited States Geological Survey, 2026
- The U.S.-China Trade Truce Has Not Solved the Gallium ProblemCenter for Strategic and International Studies, 2026
- Dismantling China Supply-Chain WeaponJoris Teer, Council on Foreign Relations, 2026
- China halts ban on gallium, germanium, antimony exports to USReuters, 2025
- China Suspends Export Controls on Certain Critical Minerals and Related ItemsPillsbury Winthrop Shaw Pittman, 2025
- Ministry of Commerce Notice 2024 No. 46China Ministry of Commerce, translation via CSET, 2024
- China's heavy rare earth tap stays closed for Japan in JuneReuters, 2026
- China's weaponization of critical minerals, satellite and trade trackingSasakawa Peace Foundation, 2026
Frequently asked questions
Why can't the United States just mine more gallium?
Because nothing mines gallium. It is recovered as a byproduct from the liquor left after bauxite is refined into alumina, and from zinc residues. Opening a mine would not produce gallium unless the alumina refining and the extraction circuit were built alongside it. That is why a 100 percent import reliance figure is a structural problem rather than a permitting one.
What actually expires on 27 November 2026?
The suspension, not the ban itself. China suspended for one year the US-directed prohibition it imposed in December 2024. The suspension took effect on 9 November 2025 and runs until 27 November 2026. If it is not extended, the December 2024 prohibition is restored by default.
Did the trade truce restore gallium shipments to the US?
Not according to the export data. Chinese unwrought gallium exports fell roughly 94 percent from 2024 levels and had not recovered as of May 2026, and analysts reported fewer general licences approved than the truce implied. The suspension removed one of two legal barriers and left the licensing regime in place.
How much gallium does the US import?
On the order of 15 million dollars of gallium metal a year according to USGS summary figures. That is the whole point of the story: a very small import line gates a much larger downstream capability in gallium nitride power devices and radio frequency electronics.