Selenium

A Byproduct Metal: Selenium Rides on Copper's Refining Output, Not Its Own Mine Plan

1. How anode slime becomes refined selenium

Selenium occurs in trace concentrations within copper sulfide ores, particularly chalcopyrite, and becomes economically recoverable only after copper is electrolytically refined: the metal collects in the residual anode slimes left behind in the refining cells, alongside gold, silver, tellurium, and platinum-group metals. Per USGS Mineral Commodity Summaries 2026, “selenium is recovered principally as a byproduct of the electrolytic refining of primary copper, where it accumulates in the residues of copper anodes,” with lead, nickel, and zinc ores serving as minor secondary sources. Coal also contains meaningful selenium, but USGS states that recovering it from fly ash, while technically feasible, “will likely not be economical in the foreseeable future.” A typical process, as documented at Poland's Głogów smelter, dedusts the Kaldo furnace off-gas into a suspension, leaches it with sodium hydroxide to precipitate lead, then precipitates selenium from the leach solution with sulfur dioxide to yield “technical selenium containing 99.65% Se,” with that single plant producing “80 to 90 Mg/year” (80–90 tonnes annually) (Journal of Chemical Technology and Metallurgy, Głogów precious-metals plant process description).

2. Yield economics: grams of selenium per tonne of copper

Because selenium content in ore and anode slime varies by deposit, refiners typically recover on the order of tens to roughly one hundred grams of selenium per tonne of copper cathode produced, a yield too small to justify standalone extraction but large enough, aggregated across a global copper-refining base of over 20 million tonnes of electrolytic cathode per year, to sustain a multi-thousand-tonne byproduct market. Sweden's Rönnskär smelter, run by Boliden, lists selenium as one of a broad byproduct slate — “Au, Ag, Se, Te, Ni, PGM, OPM, H₂SO₄, SO₂” — recovered from the copper circuit alongside its main copper, zinc-clinker, and lead products (Boliden Supplier Summit 2019, smelter product slate). A Danish Environmental Protection Agency technical review of the same Rönnskär flow describes how “the sludge in the electrolyte tanks contains precious metal impurities and is forwarded for a separate precious metal plant, where gold, silver, platinum, palladium and selenium are recovered,” noting that recovery economics govern how much is actually extracted — “at present, about 80% may be recovered from an economical point of view” (Danish EPA, Recycling Analysis, copper process description).

3. Minor secondary sources: lead, nickel, and precious-metals slimes

Beyond copper, selenium also reports to slimes generated during lead and nickel refining, and to precious-metals slag processing at integrated smelter-refineries. Poland's KGHM operates its Głogów Precious Metals Division specifically to extract “refined silver, gold, palladium-platinum concentrate and selenium” from anode slime generated alongside 99.99%-pure copper cathode production, with that division added to the smelter in 1990–1993 specifically for precious- and minor-metal recovery (KGHM, Głogów Copper Smelter and Refinery). Similarly, Chile's historical smelter base handled selenium as an incidental pollutant/byproduct stream; academic air- and soil-quality research around Codelco's former Ventanas smelter documented “selenium concentrations in rainwater, soils and alfalfa” near the site, reflecting the same anode-slime and flue-dust pathway before that smelter's permanent 2023 closure (Universidad de Chile, Documento de Trabajo SDT 501; Codelco, Ventanas transformation Q&A).

Why it matters: because selenium supply tracks copper smelting decisions made for entirely unrelated reasons (copper price, concentrate availability, smelter maintenance cycles, and even environmental permitting that can close a smelter outright, as at Ventanas), the market cannot respond to a selenium demand shock by opening new capacity — producers can, at best, marginally improve anode-slime selenium recovery rates. Any surge in demand (for example, from photovoltaics or metallurgy) runs into a supply curve set by the global copper industry, not by selenium economics. USGS also reports that reserves and resources of selenium “are generally not reported at the mine or country level and cannot be reliably quantified” (USGS MCS 2026).

Global Refined Production: China's Share Crosses 50% as Copper Refining Capacity Expands

Refinery output and capacity by country, 2024–2025

Source: USGS MCS 2026. U.S. production is withheld as proprietary; historical USGS series also list Mexico, Peru, and Poland as producers, though 2026 disclosed figures for those countries in the salient-statistics table were not separable from rounding in the underlying release. World refinery capacity (~5,100 t) meaningfully exceeds current output (~3,800 t), indicating the industry is running well under nameplate — consistent with a byproduct commodity where refiners process whatever anode-slime volume copper operations generate rather than running selenium circuits at full utilization.

The Chinese capacity build-out and the Kazakhstan entrant

USGS specifically flags that “selenium production in China increased significantly over the past 10 years, corresponding with an increase of nearly 75% in the production capacity of electrolytically refined copper,” while “the production capacity of copper anodes, the feedstock material for electrolytic copper refineries, more than doubled over the same time period” (USGS MCS 2026). Chinese refined-selenium producers include large integrated copper groups such as Jiangxi Copper and Yunnan Copper, whose byproduct precious- and minor-metal output scales directly with their primary copper cathode volumes rather than any independent selenium investment plan. Outside the established base, USGS also notes that in January 2025, the first batch of refined selenium shipped from a newly completed plant in Kazakhstan, expected to produce approximately 75 tonnes per year of 99.5%-purity selenium — a small but notable new entrant diversifying supply beyond the traditional China/Japan/Europe axis (USGS MCS 2026).

Named Japanese, European, and Nordic refiners

Among named Western and Asian refiners, Sumitomo Metal Mining and Mitsubishi Materials process selenium as part of their integrated copper smelting and precious-metals recovery operations in Japan, with Sumitomo's Niihama/Toyo complex among the longest-operating byproduct-selenium circuits in the industry (Sumitomo Metal Mining, selenium product page; Mitsubishi Materials, metals products). In Europe, Aurubis recovers selenium as part of its multi-metal output from copper anode slimes at its Hamburg site and markets high-purity selenium products under its RETORTE brand, while Umicore recovers selenium and related minor metals through its Hoboken, Belgium precious-metals refining operations tied to copper and complex feed processing (Aurubis, Life Cycle Assessment of RETORTE Selenium Products, 2024; Aurubis Consolidated Financial Statements FY2024/25). Sweden's Boliden Rönnskär smelter recovers selenium, tellurium, nickel, and platinum-group metals alongside its main copper, gold, silver, and lead output, describing itself as processing “Boliden's entire production of copper concentrate” through an integrated Kaldo-furnace and precious-metals circuit (Boliden Rönnskär, industrial area expansion filing). Poland's KGHM Głogów smelter/refinery sells “technical selenium, guaranteed specification Se – min. 99.40% (typical content 99.5–99.9%)” as a named commercial product line alongside its refined silver and gold (KGHM, Other Products — Selenium specification).

Why it matters: unlike gallium or germanium, where China's dominance stems from deliberate downstream chemical-processing investment, selenium's China share is a mechanical byproduct of China becoming the world's largest copper smelter base. That means Western de-concentration would require new copper refining capacity outside China — a much larger and slower capital commitment than building a standalone selenium plant.

End Uses: Metallurgy and Glass Absorb 60% of Global Consumption

Source: USGS MCS 2026.

1. Metallurgy: free-machining steel, lead-free brass, and manganese oxidant

Metallurgy is the largest single use. Selenium dioxide is consumed as an oxidant in electrolytic manganese metal production, and metallic selenium is added in small quantities to free-machining steels and lead-free copper alloys (brasses) to restore the chip-breaking machinability historically provided by lead, without lead's toxicity — USGS lists bismuth, lead, and tellurium as substitutes for selenium in free-machining alloys, and bismuth/tellurium as substitutes in lead-free brasses, underscoring how closely selenium's metallurgical niche overlaps with bismuth's (USGS MCS 2026).

2. Glass manufacturing: decolorizer and ruby-red colorant

Glass manufacturing uses selenium compounds both as a decolorizer — offsetting the greenish tint iron impurities impart to container and flat glass — and, at higher loadings, in combination with cadmium sulfide/selenide as a colorant to produce reddish/ruby glass tones, a technique documented in glass-industry technical literature going back decades (U.S. Patent 2,224,791, selenium-containing ruby glass; British Glass, Developments in the Use of Selenium).

3. Agriculture and animal health: a regulated micronutrient additive

Agriculture and animal health reflects selenium's role as an essential micronutrient added to livestock feed. In the United States, the FDA authorizes selenium (as sodium selenite, sodium selenate, or selenium yeast) in complete feed for chickens, swine, turkeys, sheep, cattle, and ducks at up to 0.3 ppm, with tighter per-head daily caps for supplements and free-choice mineral mixes (21 CFR §573.920, U.S. Government Publishing Office). The European Union caps total selenium in complete feed at 0.5 mg/kg, with organic-source selenium capped further at 0.2 mg/kg, under EFSA's ongoing safety review of the margin between nutritional benefit and toxicity (EFSA Journal, Consumer safety of feed additives containing selenium, 27 Jun 2024). New Zealand agronomy guidance for naturally low-selenium pastures recommends specific per-species feed-selenium targets — “dairy cattle 0.4 mg/kg, beef cattle 0.3 mg/kg, sheep 0.15 mg/kg, pigs 0.2 mg/kg, horses 0.1 mg/kg” — delivered via mineral supplements such as Solmin/Selovet because unfertilized New Zealand pasture is naturally deficient (Bioactive Soils, Selenium (Se) technical note).

4. Chemicals, pigments, and legacy electronics

Chemicals and pigments use selenium compounds as catalysts, in cadmium sulfoselenide red pigments for plastics and ceramics, and in specialty organic synthesis. Electronics historically used selenium rectifiers and photoconductive drums in photocopiers, though USGS notes silicon has substituted for selenium in most low- and medium-voltage rectifier applications, shrinking that legacy electronics demand pool over time (USGS MCS 2026). The xerographic photoconductor-drum and selenium-rectifier businesses that once anchored a large share of Western electronics demand have effectively disappeared from current end-use tallies, replaced by organic photoconductors and silicon diodes; USGS's current 10% “electronics and photovoltaics” category is now dominated by the photovoltaic component described in Section 4, not by legacy rectifier or photocopier demand.

Why it matters: selenium's demand base is diversified across five genuinely unrelated end markets (steel/brass machinability, glass color, livestock nutrition, electronics, and solar), which insulates it from a single-sector demand shock but also means no single growth driver can move the whole market — a contrast with metals whose fate is tied to one headline application (e.g., cobalt to EV batteries).

Thin-Film Solar: A Real but Marginal Niche Against Cadmium Telluride and Silicon

1. CIGS absorber chemistry and the competitive landscape

CIGS cells use a copper-indium-gallium-selenide (CuInGaSe₂) absorber layer deposited on a flexible or rigid substrate; USGS's 2026 substitutes discussion states plainly that “amorphous silicon and cadmium telluride are the two principal competitors with CIGS in thin-film photovoltaic cells” (USGS MCS 2026). In practice, CdTe — led at scale almost entirely by First Solar — has won the utility-scale race: cadmium telluride held roughly 59% of the thin-film PV market in 2024, with First Solar's Series 7 modules shipping at 18.1–19.7% efficiency and qualifying for U.S. domestic-content incentives, while CIGS commercial modules ship at a lower 14–17% and rarely undercut CdTe on cost per watt (Mordor Intelligence, Thin Film Solar PV Market Report, Dec 2025; SurgePV, 2026).

2. First Solar's CdTe alloy: selenium as a trace efficiency dopant

First Solar's cadmium telluride cells are not pure CdTe but a cadmium-selenium-telluride alloy; the company's own Series 7 Environmental Product Declaration lists “thin film CdTe semiconductor” at just 0.1% of module mass across both its U.S. and India production lines, within which a selenium alloying addition is used industry-wide to widen the cell's bandgap grading and lift conversion efficiency (EPD Norge, First Solar Series 7 Environmental Product Declaration). First Solar has continued to expand CdTe manufacturing capacity domestically, inaugurating a new 3.5 GW-nameplate, fully vertically integrated facility in Iberia Parish, Louisiana in 2025, taking its U.S. footprint to 14 GW in 2026 and a planned 17.7 GW in 2027 with a South Carolina facility — expansion that, at the margin, is the largest single driver of Western CdTe-linked selenium alloy demand (Semiconductor Today, First Solar Louisiana facility inauguration, 24 Nov 2025).

3. CIGS after Solar Frontier: Idemitsu Kosan's pivot to space-grade cells

CIGS retains a defensible niche in building-integrated photovoltaics (BIPV), flexible and curved-surface installations, and lightweight rooftop applications where crystalline silicon's rigidity and CdTe's glass-glass format are less suitable; current producers cited in market reporting include Avancis, MiaSolé, and Midsummer. Solar Frontier — the Japanese CIGS pioneer that once held the CIGS lab-efficiency record at 23.6% — exited conventional module manufacturing, but its parent, Idemitsu Kosan, has redirected the underlying CIGS technology toward space applications: Idemitsu announced in October 2025 that its CIGS solar cell for space use had reached a milestone, and in November 2025 announced a strategic collaboration with Source Energy Company on CIGS satellite power; by April 2026 it announced plans to “establish a bench manufacturing plant for space-grade” CIGS cells, indicating that Japan's CIGS selenium-consuming capability is continuing in a specialized, low-volume space-power niche rather than ending outright (Idemitsu Kosan, CIGS Solar Cell for Space Application milestone, 17 Oct 2025; Idemitsu Kosan, Idemitsu and Source Energy strategic collaboration, 6 Nov 2025; Idemitsu Kosan, space-grade CIGS bench manufacturing plant announcement, 22 Apr 2026).

USGS's 10% figure for “electronics and photovoltaics” combined in the 2025 end-use breakdown (Section 3) is not broken out further between the two categories, so the CIGS- and CdTe-specific selenium draw is smaller still than the headline 10% figure suggests once legacy electronics demand (rectifiers, photoconductors) is netted out (USGS MCS 2026).

Why it matters: selenium bulls sometimes frame CIGS solar as an energy-transition growth story analogous to lithium in batteries, but the market data does not support that scale: CIGS has been stuck near 1% global PV share for years against a dominant, cheaper, more bankable CdTe incumbent, and total thin-film technology (CdTe + CIGS + amorphous silicon combined) is only 5–7% of the overall solar market (GlobeNewswire, Thin Film Photovoltaics Global Market Report 2025–2035).

Essential Yet Toxic: A Narrow Dose Window Shapes Regulation on Both Ends

1. Drinking-water and dietary intake limits

Selenium is an essential trace nutrient in humans and animals, required for antioxidant enzyme function, but the U.S. National Academies and EFSA both place its tolerable upper intake level only a few multiples above the recommended intake — EFSA's 2022 opinion set the adult tolerable upper intake level at 255 µg/day, down from an earlier 300 µg/day standard, against a US recommended daily value of 55 µg (Nutrients journal, Legal Standards for Selenium Enriched Foods, 28 Oct 2024). The EPA's drinking-water Maximum Contaminant Level (MCL) and Maximum Contaminant Level Goal (MCLG) for selenium have both stood at 0.05 mg/L since the regulation took effect in 1992, set “to protect against damage to the nervous system” (US EPA archived Consumer Factsheet on Selenium). The World Health Organization's guideline value is slightly stricter, at 0.04 mg/L (Water Quality Association technical fact sheet, citing WHO 2011). Natural sources — discharge from petroleum refineries, erosion of natural mineral deposits, and drainage from mining — are EPA's listed causes of selenium contamination in water supplies (US EPA).

2. Livestock feed thresholds and the deficiency-toxicity gap

On the livestock side, regulators must thread the same needle in the opposite direction: the FDA's 21 CFR §573.920 sets a maximum feed inclusion rate of 0.3 ppm for complete feed across most farm species, with tightly bounded per-animal daily intake caps for supplements (0.7 mg/head/day for sheep, 3 mg/head/day for beef cattle) (21 CFR §573.920). Agricultural extension guidance underscores why the ceiling is so tightly drawn: selenium can become chronically toxic to livestock at levels “as low as 3.0 to 5.0 ppm,” only roughly ten to fifteen times the minimum required level — one of the narrowest essential-versus-toxic windows of any trace mineral used in animal nutrition (Michigan State University Extension, Use Selenium Cautiously When Feeding Livestock).

3. Deficient-soil regions and national supplementation programs

Some regions have naturally selenium-poor soils, creating a public-health rationale for deliberate supplementation rather than restriction. Finland ran the world's most comprehensive national program: in 1984, the Finnish Ministry of Agriculture and Forestry began adding sodium selenate to all multi-nutrient fertilizers nationwide because Finnish soils were “particularly poor in selenium” and the population showed measurable deficiency; the National Public Health Institute had been tracking population blood-selenium levels since the 1970s, and after the program raised food-chain and blood selenium levels successfully, the initial higher application rate was reduced in 1990, settling at the current rate of 10 mg of selenium per kilogram of fertilizer — a program described as “unique in the world” for demonstrating “the safety, effectiveness, and cost-efficiency of selenium fertilization to raise selenium levels in a population” (Farming First, Selenium Fortified Fertilizers in Finland). New Zealand faces a parallel, longstanding natural deficiency in unfertilized pasture soils, addressed through selenium-fortified fertilizer products, drinking-water dosing systems (e.g., Solmin/Selovet), and selenium bullets/chips for livestock, with agronomy guidance explicitly warning that “a little too much Se has been severely toxic to horses,” illustrating the same narrow deficiency-toxicity window seen in U.S. and EU feed regulation (Bioactive Soils, Selenium (Se) technical note).

4. Acute toxicity: clinical evidence from supplement mismanufacture

In humans, acute overexposure produces a recognizable syndrome — documented case reports describe garlic-odor breath from exhaled dimethyl selenide, hair and brittle-nail loss, and gastrointestinal symptoms, with a 2010 case series in the Archives of Internal Medicine describing acute selenium toxicity from a mismanufactured dietary supplement that delivered roughly 200 times the labeled dose (Archives of Internal Medicine, Acute Selenium Toxicity Associated With a Dietary Supplement, 8 Feb 2010). Iron-selenium (Fe-Se) combination supplements are also marketed for populations with concurrent deficiencies, reflecting selenium's routine clinical use as a nutritional supplement even as regulators cap upper intake to guard against the toxicity documented in overdose case reports.

Why it matters: selenium is one of the few industrial metals simultaneously regulated as a required nutrient (minimum thresholds) and a contaminant (maximum thresholds) in both human and animal exposure pathways, which keeps it under continuous regulatory review by EFSA, EPA, and FDA even though it carries no export-control or critical-minerals designation in major Western frameworks.

No LME Contract, Steadily Rising Price, and Selenium's Absence From China's Export-Control Lists

1. Price history, 2021–2025

Source: USGS MCS 2026, selenium chapter. Converted to a per-pound basis (roughly $12.70–$13.15/lb at the 2025 estimate), selenium's assessed price remains within the broad $10–40/lb range that professional minor-metals desks have quoted for 99.5%-purity material through the 2020–2025 period, with volatility driven primarily by swings in Chinese industrial demand rather than by any single supply shock.

2. Price-assessment mechanism: Fastmarkets and Argus, no exchange contract

Because there is no exchange-traded futures contract, price discovery runs through independent assessors: Fastmarkets publishes a benchmark assessment for “Selenium 99.5% Se min, in-whs Rotterdam, $/lb” (Fastmarkets, Selenium price assessment page), and Argus Media maintains a parallel minor-metals selenium price and news service (Argus Media, Selenium prices, charts and news). USGS's own annual average figures, drawn from Metal Bulletin-style reporting, remain the most widely cited reference point for U.S. and European selenium pricing in the absence of exchange settlement data. Unlike copper, aluminum, or nickel, selenium has never supported a listed futures or options market, reflecting its small absolute market size (a few thousand tonnes per year globally) relative to LME-listed base metals.

3. Trade flows and U.S. import reliance

The U.S. remains structurally import-dependent regardless of its lack of exposure to export controls: net import reliance as a share of apparent consumption was greater than 50% in every year from 2021 through 2025e, with the Philippines (25%), Mexico (14%), Chile (12%), and Poland (11%) the leading import sources of selenium metal for 2021–24, and the Republic of Korea (78%) dominating imports of selenium dioxide specifically (USGS MCS 2026). This import slate is distinct from most other critical minor metals in that China does not appear among the top U.S. selenium metal suppliers, reflecting selenium's genuinely multi-polar refining base (Japan, Belgium, Germany, Poland, Chile, the Philippines, Peru) rather than the China-concentrated import profiles seen for bismuth, tungsten, or gallium.

4. Selenium's conspicuous absence from MOFCOM's 2024–2026 control actions

China's broader annual Catalogue of Dual-Use Items and Technologies Subject to Import and Export License Administration — last refreshed for 2026 under Announcement No. 91 of 2025, effective 1 January 2026 — added new entries covering tungsten-nickel alloys, samarium compounds, indium compounds, bismuth, and telluride items relative to the prior year's catalogue, but did not add selenium (China Briefing, 2025 vs 2026 Dual-Use Catalogue Comparison; ChemRadar, Announcement No. 91 of 2025 text). USGS's own tariff table for selenium and selenium dioxide lists both as duty-free under U.S. Normal Trade Relations through year-end 2025, with no export-control annotation (USGS MCS 2026).

Why it matters: selenium's absence from every 2024–2026 Chinese export-control action — despite China holding the largest single-country refining share — distinguishes it from bismuth, tellurium, tungsten, gallium, germanium, antimony, and the rare earths that Beijing has explicitly weaponized in the tariff dispute. That makes selenium's 2025 price rise (16–17%) a function of ordinary copper-refining supply tightness and demand growth, not geopolitical restriction — a materially different risk profile than the metals MOFCOM has placed under license.

Recycling: USGS Calls It “Insignificant” — A Dispersive-Use Problem

1. Why selenium's biggest uses are structurally unrecoverable

Selenium's two largest consumption categories — glass manufacturing (20%) and agriculture/ animal feed (20%) — are textbook “dispersive uses” in critical-minerals terminology: once selenium is melted into a glass matrix at parts-per-million concentrations, or fed to livestock as a micronutrient and metabolized/excreted, there is no physical or economic pathway to recover the metal. Metallurgical uses (40% of consumption, the largest category) are similarly dispersive: selenium added to free-machining steel or used as a manganese-refining oxidant is consumed or diluted below recoverable concentrations in the final product or process residue. Only the smaller electronics/photovoltaics and chemicals categories retain any end-of-life recovery potential, and even there, USGS's own assessment is that what little scrap exists is “exported for recovery” rather than processed domestically in most consuming countries (USGS MCS 2026).

2. Copper refining as a “manufactured mineable resource”

Because selenium has essentially no post-consumer recycling loop, its entire practical supply response depends on primary extraction from copper (and minor lead/nickel) anode slime at the point of refining — a dynamic USGS and the broader critical-minerals literature describe as a byproduct metal being effectively a “manufactured mineable resource” whose available tonnage is set by decisions in an entirely different industry (copper smelting capacity, ore grade, and slime-processing investment) rather than by a selenium-specific ore body or a recycling stream that could buffer against supply swings. This reinforces the Section 1/2 point that selenium supply security is inseparable from global copper-refining capacity trends.

3. Limited scrap recovery in electronics and chalcogenide glass

Where recovery is technically possible, it remains a niche activity: producers of selenium-tellurium chalcogenide glass and electronic-grade selenium occasionally recover process scrap in-house during manufacturing, and some end-of-life electronic scrap containing selenium compounds is collected and exported to specialized processors (chiefly in Asia and Europe) for metal recovery, consistent with USGS's note that most such material moves internationally rather than being recycled domestically where it is generated (USGS MCS 2026). No dedicated selenium-recycling industry association or standard (comparable to the Cobalt Institute's battery-recycling initiatives, for instance) exists, reflecting how small and dispersed the technically recoverable pool is.

Why it matters: selenium is one of the clearer examples among industrial minor metals where recycling cannot meaningfully cushion a supply disruption. Any future demand growth (from thin-film photovoltaics or chalcogenide optics) or supply contraction (from copper smelter closures, such as Codelco's 2023 Ventanas shutdown) has to be absorbed entirely by primary byproduct output, since there is no secondary-material buffer to draw down.

Forward Look, 2026–2030: Solar Growth, Chalcogenide Optics, and a Supply Curve That Doesn't Move on Its Own

1. Photovoltaic demand: CdTe capacity additions outpace CIGS

The clearest near-term selenium demand growth vector remains photovoltaics, but concentrated in CdTe rather than CIGS: First Solar's Louisiana facility (3.5 GW nameplate, commissioned ahead of schedule in 2025) and a planned South Carolina facility are set to lift First Solar's U.S. manufacturing footprint to 17.7 GW by 2027, expanding the CdTe alloy volume that uses selenium as a bandgap-grading dopant (Semiconductor Today, 24 Nov 2025). CIGS, by contrast, is pivoting toward specialty and space-power niches rather than utility-scale volume, as shown by Idemitsu Kosan's 2025–2026 space-grade CIGS announcements (Section 4), implying its selenium draw will stay a rounding error against CdTe and silicon even as the technology finds a defensible high-value application.

2. Chalcogenide glass and infrared optics: a small but technical growth pool

Selenium-germanium (and selenium-antimony/arsenic-sulfur) chalcogenide glasses are used to mold infrared lenses and optical windows for thermal-imaging cameras, sensors, and fiber-optic infrared transmission, prized because they transmit in the mid- and far-infrared bands where conventional oxide glass is opaque; academic and industry literature on precision glass-molding of chalcogenide infrared optics documents ongoing materials development in Ge-Sb-Se and related Se-based systems for lens and fiber applications (Micromachines, A Review of the Precision Glass Molding of Chalcogenide Glasses, 2018; American Ceramic Society Bulletin, Chalcogenide Glasses: Engineering in the Infrared Spectrum, 2026). This remains a specialty-optics volume rather than a bulk-tonnage selenium consumer, but it is one of the few genuinely new technical applications for selenium metal identified in current literature, alongside cadmium selenide quantum dots used in display backlighting and biomedical imaging, a market one industry forecast values at “$1.8 billion in 2025” growing toward “$4.2 billion by 2034” on quantum-dot display and next-generation photovoltaic demand (Dataintelo, Cadmium Selenide Semiconductor Market Research Report, 2025 estimate).

3. Supply-side constraint: byproduct economics cap the response

Whatever combination of these demand vectors materializes, the supply side cannot respond independently: as Sections 1 and 2 established, selenium output is capped by the pace of global copper anode-refining capacity growth (fastest in China) and by discrete plant-level events such as Codelco's permanent 2023 closure of the Ventanas smelter in Chile, which removed one historical selenium-bearing byproduct stream from the market (Codelco, Ventanas Q&A). World refinery capacity of roughly 5,100 tonnes already exceeds 2025 output of roughly 3,800 tonnes (Section 2), meaning near-term demand growth could plausibly be met by higher utilization of existing byproduct circuits without requiring new copper-refining investment — but any demand growth beyond that buffer would again run into the multi-year lead times of new copper smelter/refinery capacity.

4. Key risks and what to watch, 2026–2030

The principal risks to the current benign trade-policy environment are: (1) selenium's shared byproduct stream with copper, gold, and silver means any future MOFCOM catalogue action targeting copper-refining byproducts as a class could sweep in selenium even without a selenium-specific rationale; (2) further global copper smelter closures on environmental grounds (as at Ventanas) could tighten the byproduct-slime supply base faster than new Chinese or Kazakh capacity can offset it; and (3) EFSA or EPA regulatory tightening on the deficiency-toxicity margin (Section 5) could constrain the agriculture/feed end use, currently 20% of demand, without materially affecting the metallurgical or solar categories. Absent a specific export-control or stockpile action, selenium's 2026–2030 trajectory is set to remain a function of ordinary industrial supply and demand rather than geopolitics — a genuinely different risk profile from the bismuth, tungsten, gallium, germanium, and antimony narratives running in parallel across the critical-minerals complex (USGS MCS 2026).