Mercury Hg · 80

The only metal that is liquid at room temperature, and one of the most tightly regulated substances in the world.

Pourquoi est-ce important ?

Mercury is the clearest case of a material being deliberately designed out: the Minamata Convention is phasing it out of products and processes worldwide.

Where it is in the Earth

Mercury is a dense, silver-coloured liquid at ordinary temperatures, with a melting point of −38.8 °C, which means it is never solid in any warm environment. In the Earth's crust it is genuinely rare, and it becomes mineable only where specific geological conditions have concentrated it into a mineral called cinnabar — mercury sulfide — over long periods of hydrothermal activity. Hydrothermal simply means hot water moving through rock: water heated by magmatic or geothermal sources dissolves mercury from surrounding rock, carries it upward through fractures, and deposits cinnabar when the water cools or encounters chemically different rock near the surface. Because this process depends on relatively shallow, low-temperature circulation, mercury deposits tend to sit in the upper few kilometres of the crust, often close to ancient volcanic arcs or fault zones.

The classic geological setting is what geologists call an epithermal deposit — formed at low temperatures and shallow depths — typically hosted in sedimentary rocks such as limestone, chert, or shale that have been cut through by faults. The Almadén district in Spain, historically the world's largest source, sits in Palaeozoic sediments deformed by faulting. The Idria deposit in what is now Slovenia, the Huancavelica deposit in Peru, and various deposits across central Asia share the same broad character: fault-controlled conduits, chemically reactive host rock, and a tectonic history involving compression or extension that opened pathways for heated fluids. China's producing areas, concentrated in Guizhou province, follow the same pattern — fault systems cutting through carbonate sequences.

Because the geological conditions are specific and not especially common, the world's significant mercury deposits have always been few in number and geographically clustered. Most of the historically large deposits are now exhausted or deliberately idled under international agreements. What remains in active production is, by comparison with the industry of the mid-twentieth century, very small.

Getting it out

Cinnabar, the principal ore of mercury, typically occurs in veins or as disseminations through the host rock rather than in massive, uniform bodies. Historically this meant underground mining, following the veins by hand and later by machine. Some deposits were amenable to open-pit methods where ore was close to the surface and concentrated enough to justify the earthmoving involved. The choice between open-pit and underground working comes down to the geometry of the ore body and the grade — that is, how much mercury metal can be recovered from each tonne of rock mined. Because mercury deposits are generally small and irregular compared with, say, a large copper porphyry, the volumes of rock moved are modest by modern mining standards.

Grade matters enormously because mercury is expensive to process and the metal itself has a declining market. An ore body that might have been economic at higher mercury prices a generation ago may not justify the cost of development today. In practice, much of the world's current supply does not come from primary mercury mining at all. A significant portion is recovered as a by-product of mining other metals — particularly gold, where mercury occurs alongside gold mineralisation in some deposits — or from the retorting of mercury-bearing waste and tailings left by earlier operations. By-product recovery changes the economics entirely: the cost of producing the mercury is largely absorbed by the primary metal, and the mercury is captured because releasing it to the environment is no longer acceptable under national and international regulations rather than because it commands a price that rewards recovery on its own terms.

The direction of primary mercury mining is unambiguously downward. The Minamata Convention on Mercury, which came into force in 2017, commits its signatories to phasing out primary mercury mining. The figures in the production table reflect this: global mine output is a fraction of what it was in earlier decades, and the countries still reporting primary output are few. Norway's figure relates to legacy material management rather than active ore extraction in the conventional sense.

What pulls on it

Mercury's uses have contracted sharply over the past half century, and that contraction has been deliberate. The clearest driver is regulation: many countries have banned or heavily restricted mercury in products ranging from thermometers and blood-pressure monitors to fluorescent lamps, batteries, and dental amalgam. The Minamata Convention extends this pressure globally, requiring signatories to phase out whole categories of mercury-added products by agreed deadlines. The result is that demand from what were once large markets — medical devices, electrical switches, measuring instruments — has largely disappeared in countries with effective regulation, and is declining in those still implementing it.

The uses that remain significant fall into two broad categories. Artisanal and small-scale gold mining, known as ASGM, is by some measures now the largest single use. Miners add liquid mercury to gold-bearing alluvial sediment; the mercury amalgamates with gold particles, separating them from the surrounding material. The amalgam is then heated to drive off the mercury and recover the gold. This technique is old, simple, and effective for the type of deposit worked, which is why it persists in parts of sub-Saharan Africa, South America, and South-East Asia despite well-documented harm to human health and river systems. Chlor-alkali production — the industrial process that makes chlorine and caustic soda by passing electricity through brine — historically used mercury-cell technology and consumed mercury continuously as cells aged. Most modern chlor-alkali capacity has converted to membrane-cell technology that uses no mercury, but some older plant remains in operation in certain countries.

For demand to change sharply in either direction, the conditions would need to change at the level of enforcement and international compliance. If ASGM-using countries implement the formalisation and alternative-technology programmes that the Minamata Convention envisages, that use would decline. If enforcement weakens or illegal supply chains expand, it could persist or grow. There are no significant new commercial applications on the horizon that would generate large additional demand; mercury's chemical and physical properties that made it attractive — density, electrical conductivity, liquid state — can be replicated or approximated by other means in almost every modern application where the toxicity is unacceptable.

Turning ore into product Niveau 3

The conversion of cinnabar ore into metallic mercury is among the simpler pyrometallurgical processes in the industry. Cinnabar (HgS) is thermally unstable at relatively modest temperatures: roasting the ore in a furnace — heating it in the presence of air — decomposes the mineral, releasing mercury vapour and sulfur dioxide. The mercury vapour passes out of the furnace with the flue gases and is then condensed back to liquid metal by cooling. Because mercury is already a liquid at room temperature, collection is straightforward once the vapour is cooled; the metal simply runs out of the condenser. This sequence of roasting followed by condensation is called retorting, and it has been the standard approach for centuries, with modern equipment adding better temperature control and gas-handling systems to capture the sulfur dioxide and prevent atmospheric mercury emissions.

Recovery efficiency depends on ore grade and mineralogy. Where cinnabar is the dominant mercury phase and the ore is not too fine-grained or locked within dense silicate minerals, recoveries can be high. In lower-grade ores, or where mercury is present partly as native metal inclusions or as organic complexes rather than as pure cinnabar, recovery is lower and the economics deteriorate. The sulfur dioxide produced during roasting requires treatment — it cannot simply be vented — and managing this gas stream adds to operating cost. Refining the condensed mercury to remove impurities such as base metals or sulfur compounds is a further step, typically involving filtration and acid washing or vacuum distillation, depending on the end use and the purity specification required. The traded form, the 34.5 kg flask, is a historic unit that dates from the period when Almadén dominated global trade; the flask remains the standard commercial unit even though primary production is now very small.

For mercury recovered as a by-product — for instance, from gold roasting circuits — the processing sequence is similar in principle but integrated into a larger plant. The mercury condenses in the off-gas train designed primarily to handle other volatiles, and is then collected and packaged. The capital cost of recovery is largely shared with the primary operation, but the operator must still meet strict handling and storage standards. In jurisdictions that have ratified the Minamata Convention, long-term storage of surplus mercury in a form that prevents re-entry into commerce is a regulatory requirement, which means some mercury recovered as a by-product is not sold but stockpiled under defined containment conditions.

Substitution and recycling Niveau 3

In the uses that have already been regulated out of existence in much of the world, substitution has been demonstrated and is essentially complete. Digital thermometers replace mercury-in-glass instruments with no meaningful performance loss for most clinical and laboratory applications. Alcohol-filled thermometers serve in temperature ranges where a liquid-expansion device is still preferred. Electronic pressure sensors replace mercury manometers and sphygmomanometers. Reed switches and solid-state relays replace mercury tilt switches and wetted contacts. In fluorescent lighting, the move toward LED technology eliminates mercury entirely rather than substituting one mercury-containing product for another. In each of these cases the substitute is now cheaper, often more convenient, and the performance difference is irrelevant to the application. The substitution has happened not because the economics were obviously compelling from the start but because regulation forced manufacturers to solve the engineering problem, and once solved, the new approach became standard.

In chlor-alkali production, membrane-cell and diaphragm-cell technologies are mature and well-established substitutes for mercury-cell plant. The capital cost of converting an existing mercury-cell facility is significant, which is why some older plant has remained in operation longer than anticipated, but the technology risk is negligible. In dental amalgam, composite resin materials have improved sufficiently that many dentists now use them for most restorations; the phase-down of amalgam under the Minamata Convention is therefore supported by a credible alternative, though some practitioners maintain that amalgam remains superior for certain posterior restorations in terms of durability and ease of placement under wet conditions.

The case where substitution is genuinely difficult is ASGM. Mercury amalgamation is cheap, requires no infrastructure, is understood by the miner, and works in the field without electricity or chemicals that need to be sourced through formal supply chains. Alternatives exist — gravity concentration, direct smelting, and in some settings cyanide leaching — but each requires capital, skill, or infrastructure that artisanal miners may not have access to. Recycling of mercury from ASGM settings is minimal; the quantities lost to rivers and soils, or volatilised when amalgam is burned in the open, do not re-enter any formal material cycle. In industrial settings, mercury recovered from decommissioned chlor-alkali plant and from lamp processing is collected and, under the Minamata framework, is supposed to move into long-term storage rather than back into commerce. The recycling rate in the sense of mercury returned to use is therefore falling by design, not by technical failure.

Where the chain is fragile Niveau 4

The supply picture for mercury is unusual because the primary risk is not scarcity but over-supply of a substance that parties to an international treaty are committed to removing from circulation. Primary mine production is now heavily concentrated: the 2025 data show China accounting for the overwhelming majority of reported global output, with Kyrgyzstan, Tajikistan, Morocco, and Norway contributing the remainder. The United States and Peru report no primary mine output. This concentration in a single country is a conventional supply-chain risk factor, but the practical significance is complicated by the fact that demand is also shrinking and that several large legacy stockpiles — from decommissioned chlor-alkali plant, from historically mined material in Europe, and from US government holdings — represent potential supply that could, if released to commerce, overwhelm primary mine production. The Minamata Convention specifically addresses the management of these stockpiles, requiring that they not undermine the phase-out, but the governance of stockpile disposal is imperfect.

Reporting quality introduces genuine uncertainty. The USGS figures on which the production table is based acknowledge that Peru's output is not available. Several significant historical producers no longer report data in a form that allows verification. Mercury reporting is subject to the same difficulties as any controlled substance: informal trade, undeclared by-product recovery in gold-mining districts, and the deliberate opacity of some producing operations make published totals uncertain in ways that are hard to quantify from outside. The unit basis — gross weight of mercury metal — is consistent across the data presented here, but comparisons with older figures require care because historical statistics sometimes included flask-weight rather than net mercury weight, or combined primary and secondary sources in different ways.

The price series (average unit value of imports, in dollars per kilogram) shows significant year-to-year variation between 2021 and 2024, with values of 29, 33, 22, and 50 dollars per kilogram respectively. Import unit values are a proxy for price rather than a market benchmark; they reflect the mix of grades, origins, and contractual arrangements in a given year's import flow, and a small total volume means that a single large transaction can move the annual average substantially. The thinness of the market is itself a structural feature: as primary production contracts and demand falls, the number of market participants decreases, and standard price-discovery mechanisms become less reliable. Researchers using these figures should treat them as indicative of order of magnitude and directional movement rather than as precise market prices.