There is a category of critical mineral whose supply cannot be increased by building more mines, raising the price, or passing legislation — because the mineral is not what anyone is mining for. Tellurium is a by-product of copper refining. Indium is a by-product of zinc smelting. Neither has a primary mine anywhere on Earth. When the world needs more tellurium, it needs more copper refineries to recover tellurium from the anode slime left over after electrolytic copper purification. When the world needs more indium, it needs more zinc smelters processing indium-bearing zinc ores and choosing to install the additional recovery circuits required to capture the indium rather than letting it wash into the waste stream. In both cases, the decision to produce more of the by-product is made by an operator whose business model is the primary metal — and whose investment decisions, expansion timelines, and operational priorities are governed by the economics of copper or zinc, not the economics of indium or tellurium. The by-product supply ceiling is the hardest constraint in the critical minerals supply chain to solve, because it doesn’t respond to any of the normal market signals. You can’t mine what doesn’t have a mine.
This matters now because the two fastest-growing applications for indium and tellurium are both critical to the energy transition — and both are scaling into a supply structure that physically cannot scale with them.
Indium: the element on every screen
Global indium production was approximately 900 metric tons in 2023, with China producing roughly 60%, South Korea 12%, and Japan 8%. Indium’s dominant application — over 50% of global demand — is indium tin oxide, a transparent conductive film deposited on glass or plastic substrates. ITO coats the touchscreen on every smartphone, every tablet, every laptop, every ATM, every airline check-in kiosk, every Tesla center console, and every flat-panel display manufactured in the last two decades. When you swipe your finger across a screen, you are interacting with a layer of indium tin oxide roughly 200 nanometers thick. The combination of electrical conductivity and optical transparency that ITO provides has no commercially viable substitute at the price and performance level the display industry requires.
The second major application is CIGS — copper indium gallium selenide — thin-film solar cells. CIGS panels are lighter, more flexible, and perform better in low-light and high-temperature conditions than conventional crystalline silicon panels, making them attractive for building-integrated photovoltaics, curved surfaces, and military applications. The CIGS market is small relative to silicon — less than 2% of global indium demand currently goes to solar — but the technology is one of three thin-film architectures (alongside CdTe and perovskites) competing for market share in applications where rigid silicon panels don’t fit. If CIGS deployment scales with the broader solar buildout, indium demand from that sector alone could double or triple within a decade.
The by-product constraint is the ceiling. Indium is recovered from the residues of zinc smelting — specifically from the zinc sulfide ores that happen to contain indium at concentrations of 1 to 100 parts per million. Not all zinc ores contain indium. Not all zinc smelters have installed recovery circuits. The decision to install a recovery circuit — which requires additional capital expenditure and process complexity — is made by a zinc smelter operator whose primary revenue comes from zinc. If zinc prices are low and indium prices are high, the smelter might invest in recovery. If zinc prices are high, the smelter has no incentive to complicate its operation for a by-product that contributes a small fraction of total revenue. The production ceiling for indium is set by the intersection of zinc ore geology, zinc smelter economics, and the discretionary decision of zinc operators to capture a by-product they are not required to capture. That’s not a supply chain. That’s a side hustle with a geological asterisk.
China’s 60% share of global indium refining creates the same concentration risk the Rare Earth Elements course has documented for gallium, graphite, antimony, and lithium refining. The gallium post already covered the mechanism: China’s 2023 export controls on gallium and germanium demonstrated that by-product metals refined primarily in China can be restricted through the same licensing regime Beijing has applied to every subsequent mineral. Indium has not been restricted. The infrastructure for restricting it is identical to the infrastructure that restricted gallium.
Tellurium: the element inside America’s solar bet
Tellurium’s story is simpler and stranger. Global production in 2023 was approximately 640 metric tons — roughly two-thirds the volume of indium and measured in quantities so small that industry analysts note the actual available supply may be “significantly higher” than official figures because copper refiners don’t always track or report tellurium recovery. Tellurium is extracted from the anode slime that accumulates during electrolytic copper refining — a residue that also contains selenium, gold, silver, and platinum group metals. Over 78% of tellurium extraction occurs in three countries: China, Japan, and Canada. The United States has two electrolytic copper refineries — one in Texas, one in Utah — that produce copper telluride from tellurium-bearing anode slimes. That’s it. Two facilities, producing a fraction of domestic demand.
Tellurium’s critical application is cadmium telluride solar cells — CdTe panels — manufactured almost exclusively by one company: First Solar, headquartered in Tempe, Arizona. First Solar is the largest solar manufacturer in the Western hemisphere and the only major solar company that does not use crystalline silicon. CdTe panels currently represent 21% of the U.S. solar market and 4% of the global market. First Solar’s domestic production capacity is set to reach 14 gigawatts by 2026, driven by the Inflation Reduction Act’s manufacturing incentives. The company is building or expanding factories in Ohio, Alabama, and Louisiana. The entire strategic bet — reshoring American solar manufacturing away from Chinese silicon supply chains — depends on a thin-film technology whose active ingredient is a by-product of copper refining produced at roughly 640 metric tons per year globally.
The Department of Energy’s perspective paper on CdTe photovoltaics states the constraint plainly: tellurium’s “constrained availability may place a practical limit on the maximum size of the CdTe PV supply chain.” The limit hasn’t been reached yet. But the industry roadmap targeting 100 gigawatts of CdTe manufacturing by 2030 requires tellurium supply to roughly triple from current levels, and the only way to triple tellurium supply is to either triple the number of copper refineries recovering tellurium from anode slime — an investment decision made by copper companies for copper reasons — or to dramatically improve the recovery rate at existing refineries, most of which were not designed to optimize tellurium extraction because tellurium was never what they were built to produce.
The copper shortage documented elsewhere in this cluster creates a compounding irony: the energy transition needs more copper for wiring, grid infrastructure, and EV motors. It also needs the tellurium that comes out of copper refining as a by-product. But copper mines are not copper refineries. New copper mines in Chile and Peru ship concentrate to smelters in China — and Chinese smelters are not obligated to supply tellurium to American CdTe solar manufacturers. The supply chain for America’s silicon-free solar bet runs through the same Chinese refining infrastructure that the policy was designed to avoid.
The by-product problem across the course
Indium and tellurium are not the only by-product metals in the Rare Earth Elements course with this structural constraint. Platinum group metals include iridium — produced at 7-8 tonnes per year exclusively as a by-product of platinum mining — whose potential scarcity could bottleneck the entire hydrogen electrolyzer industry. Germanium, covered in the gallium/germanium post, is a by-product of zinc smelting, just like indium, and faces the same structural ceiling. Cobalt is predominantly a by-product of copper and nickel mining — the conflict minerals post documented the DRC’s artisanal mining as the exception to by-product dependency, not the rule.
The pattern is consistent: by-product metals cannot be produced independently of their host metals. Their supply responds to the economics of copper, zinc, nickel, or platinum — not to the economics of indium, tellurium, germanium, cobalt, or iridium. When the energy transition creates demand for these by-products that outpaces the growth rate of their host metal industries, the result is a supply ceiling that no amount of investment in the by-product itself can raise. The ceiling can only be raised by expanding host-metal production — which requires mines that take 7-10 years to permit and build, serving a market (copper, zinc, nickel) whose economics may not justify the expansion at any given moment. The lithium supply chain at least has the theoretical advantage that lithium is the primary product: if you need more lithium, you build a lithium mine. If you need more tellurium, you build a copper refinery and hope the anode slime contains enough tellurium to matter.
Why they share a lecture
Indium and tellurium are paired in the Rare Earth Elements course because they are the clearest illustration of the by-product supply ceiling — the constraint that separates critical minerals whose production can theoretically respond to demand from critical minerals whose production cannot. The vanadium supply chain is concentrated but primary — if vanadium prices rise, someone can open a vanadium mine. The antimony supply chain is concentrated and primary. The semiconductor supply chain is concentrated but can, in theory, be duplicated through fab construction. Indium and tellurium cannot be duplicated. They can only be recovered from processes designed to produce something else, at rates determined by someone else’s business model, from geological deposits whose by-product content was never the reason the deposit was developed.
This is the kind of supply chain our Rare Earth Elements course was built to map — where the touchscreen industry runs on a layer of indium 200 nanometers thick recovered from zinc smelter residues, America’s silicon-free solar strategy depends on 640 metric tons of tellurium recovered from copper refinery anode slime, and neither supply chain can be expanded by investing in the thing you actually need more of.