How Does Fire Purify Gold? The Science of Refining

Fire purifies gold by exploiting a fundamental chemical difference between gold and nearly everything else: gold resists oxidation at high temperatures, while most impurities do not. When a gold-bearing material is heated in a furnace, base metals like copper, lead, zinc, and iron react with oxygen or chemical agents and are drawn off as slag or gas, leaving increasingly pure gold behind. This principle has driven gold refining for thousands of years, and some fire-based techniques remain essentially unchanged in modern practice.

Why Gold Survives the Furnace

Gold is what chemists call a “noble” metal, meaning it has very little desire to react with oxygen. Copper tarnishes, iron rusts, and lead forms a powdery oxide crust, but gold can sit in open air indefinitely without corroding. This stubbornness extends to furnace temperatures. When you heat a mixture of gold and base metals past roughly 1,000°C, the base metals eagerly combine with oxygen in the air or with added chemical agents, transforming into oxides, chlorides, or sulfides that can be skimmed, absorbed, or vented away. Gold just sits there, molten and unbothered.

Gold’s melting point of about 1,064°C also plays a role. It melts at a temperature that is high enough to liquify the metal for physical separation but low enough to be reached in a simple charcoal or gas-fired furnace. Its density, roughly 19.3 grams per cubic centimeter, means that in a molten bath, gold sinks to the bottom while lighter slag floats to the top. Gravity does some of the purification work for free.

Cupellation and the Ancient Approach

The oldest fire-refining method still in regular use is cupellation, a process that dates back more than two millennia. You place impure gold (often mixed with lead) in a shallow, porous cup called a cupel, made from bone ash or magnite, and heat the whole thing in a blast of air. The lead and other base metals oxidize and are absorbed into the porous cupel material, while the gold and any silver present remain as a bright bead on the surface. Cupellation is the foundation of fire assay, the standard way to test the gold content of ore samples, and the technique has barely changed over the centuries.

Fire assay remains the gold standard (literally) for measuring how much gold is in a rock sample. A weighed portion of crushed ore is mixed with lead oxide and fluxes, melted in a furnace, and the resulting lead button collects the gold and silver. That button is then cupelled, the lead is absorbed away, and the remaining bead of precious metal is weighed. Gold recovery from platinum group metal ores through fire assay accounts for an estimated five tonnes per year globally, and in some respects the extraction techniques have not changed for centuries.¹

Miller Chlorination, the Industrial Workhorse

For large-scale refining, cupellation is too slow and leaves silver mixed in with the gold. The method that dominates modern gold refining is Miller chlorination, developed in the 1870s by Francis Bowyer Miller at the Sydney Mint. The concept is straightforward: you melt impure gold in a furnace, then bubble chlorine gas through the molten metal. Chlorine reacts aggressively with silver, copper, zinc, iron, and most other contaminants, converting them to metal chlorides that either boil off as gas or float to the surface as a slag. Gold, characteristically, refuses to react with chlorine under these conditions and stays behind in the melt.

The result is gold that typically reaches about 99.5% purity, which is good enough to pour into standard trading bars. The process is fast, too. At South Africa’s Rand Refinery, one of the world’s largest gold refiners, Miller chlorination was chosen as the primary refining method because it produces molten gold of sufficient purity for saleable bars within about two hours.² That speed and throughput make it the go-to method for processing large volumes of doré, the semi-pure gold bars that come from mining operations.

When Chlorine Is Not Enough

Miller chlorination handles most impurities, but it has a blind spot: platinum group metals. Platinum, palladium, rhodium, and their relatives are noble metals in their own right, meaning they also resist chlorine gas at refining temperatures. If your gold feed contains significant amounts of these metals, chlorination alone will not remove them, and the resulting bar will fall short of the highest purity grades.

This is where electrolytic refining steps in. The Wohlwill process, named after the German chemist Emil Wohlwill, uses an electric current passed through a gold chloride solution to plate ultra-pure gold onto a cathode while impurities, including platinum group metals, fall away as anode sludge. Rand Refinery applied Wohlwill electrolysis specifically to deposits with known high platinum group metal content, producing gold of 99.99% purity, the so-called “four nines” standard required for investment-grade bullion.² Electrolytic refining is slower and more expensive than Miller chlorination, so refiners typically reserve it for material where fire alone cannot reach the required purity.

Separating Gold from Silver

One of the trickiest problems in gold refining is separating gold from silver, because both metals are noble enough to survive most fire-based processes. After cupellation, you often end up with a gold-silver alloy bead rather than pure gold. Historically, the solution was a chemical process called parting: dissolving the silver out of the alloy with acid, leaving gold behind.

Nitric acid has been the traditional choice for parting because it dissolves silver readily but barely touches gold. The alloy is granulated into thin pieces to maximize surface area, then soaked in hot nitric acid. The silver dissolves into solution as silver nitrate, and the gold remains as a dark, spongy residue that can be melted down. This process has been studied extensively, covering both the industrial scale and the laboratory fabrication of nanoporous gold materials from the same dealloying principle.³ In modern refineries, Miller chlorination has largely replaced acid parting for bulk separation, since the chlorine reacts with silver first and pulls it out as silver chloride slag. But acid parting remains important for small batches and analytical work.

How Smelting Chemistry Squeezes Out More Gold

Fire refining does not always mean simply melting ore and waiting for gravity. In many cases, the chemistry of the furnace charge is carefully engineered to coax gold out of stubborn mineral hosts. Refractory gold ores, where gold is locked inside sulfide minerals like pyrite and arsenopyrite, are a particular challenge because the gold particles are too small and too embedded to be freed by grinding alone. These ores typically need to be roasted or smelted first to break down the sulfide matrix.

During iron matte smelting of refractory gold concentrates, certain elements act as collectors that help gold transfer from the slag into the metal phase. Research on antimony and arsenic as additives has shown that both improve gold recovery rates during smelting. As the antimony or arsenic content in the charge increased, more gold reported to the matte rather than being lost to the slag. Under optimized conditions combining both elements, gold recovery reached about 97%, while gold left behind in the slag dropped to less than two grams per tonne.⁴ This kind of fine-tuning represents the modern engineering side of fire refining: the furnace itself has not changed much, but the recipes fed into it keep improving.

Health Hazards in Gold Workshops

Fire refining sounds clean in theory, but in practice it releases a cocktail of toxic byproducts. At the industrial scale, pyrometallurgical processing of sulfide ores can release mercury, sulfur dioxide, and fine particulate matter. Smelters that process lead and zinc alongside gold have historically been significant sources of airborne mercury, with flue gas treatment systems recovering hundreds of tonnes of mercury from zinc smelters alone.⁵

The hazards are even more acute in small-scale and artisanal settings. In handmade gold jewelry workshops, where goldsmiths still use coal fires, candle-flame blowing, and acid treatments reminiscent of ancient practice, workplace contamination with toxic metals is a serious concern. Analysis of ash and dust from such workshops in Bangladesh found elevated concentrations of cadmium, chromium, lead, and arsenic, all metals with well-documented health effects including kidney damage, cancer risk, and neurological harm.⁶ Workers in these settings typically lack adequate ventilation or protective equipment, making chronic exposure the norm rather than the exception.

The Carbon Footprint of Fire Refining

Beyond direct toxicity, the energy demands of fire-based gold processing carry a substantial climate cost. A life cycle assessment of a large-scale gold production facility processing refractory ore in North America found that off-gas treatment after roasting accounted for over half of the operation’s total greenhouse gas emissions, while grinding contributed another third.⁷ The study found that a 10% reduction in electricity consumption could cut emissions from crushing and grinding meaningfully, but had little effect on the off-gas treatment stage, which is driven more by chemical processes than by electricity use. This means that even switching to renewable power would only partially address the problem for operations processing refractory ores.

These findings put fire refining’s environmental cost in perspective. The furnace itself is only part of the picture; the real emissions burden often comes from the steps surrounding it, particularly the treatment of gases released when sulfide minerals are broken down at high temperature.

Mercury-Free Borax Smelting for Small-Scale Miners

Millions of artisanal and small-scale gold miners around the world still use mercury amalgamation to extract gold. The miner grinds ore, mixes it with liquid mercury (which binds to gold particles to form an amalgam), then burns off the mercury to leave crude gold behind. The process is cheap and requires minimal equipment, but it poisons waterways, contaminates food chains, and exposes miners and their families to mercury vapor.

A simpler fire-based alternative exists: borax smelting. Instead of mercury, the miner mixes concentrated gold-bearing material with borax, a common household mineral, and heats it in a crucible. The borax lowers the melting point of the slag, allowing impurities to separate from the gold at temperatures reachable with a simple propane torch or charcoal furnace. Tests at a Bolivian gold mine compared the two methods head-to-head, splitting roughly 50 kilograms of milled ore from the same source into two batches. The mercury-free borax method recovered more gold in both trials and was almost 10% faster than mercury amalgamation.⁸

Despite these advantages, adoption has been slow. Mercury is deeply embedded in artisanal mining culture, and many miners distrust the borax method because the gold is less visible during processing. With mercury, you can see and feel the amalgam; with borax, you are trusting the furnace to do its work out of sight. Training programs in Bolivia, the Philippines, Indonesia, and Tanzania have demonstrated the technique successfully, but changing entrenched practice takes more than a good field trial.

Recovering Gold from Electronic Waste

Fire refining is not limited to mined ore. Circuit boards, connectors, and other electronic components contain small but economically meaningful amounts of gold, and pyrometallurgy is one of the main routes for recovering it. The process typically involves shredding the e-waste, feeding it into a smelter along with copper-bearing material, and recovering gold from the resulting copper matte through further refining stages.

A comparative life cycle assessment of four different methods for recovering copper and gold from waste circuit boards found that pyrometallurgy had the lowest overall environmental impact when electricity came from fossil fuels. Even when the electricity grid shifted to carbon-free sources, pyrometallurgy still outperformed chemical and biological leaching methods in nearly every impact category, with the exception of global warming potential, where a hybrid bioleaching-chemical leaching process edged ahead.⁹ The advantage of smelting comes from its ability to process mixed, contaminated feedstocks without extensive pretreatment, and from the fact that the high temperatures destroy organic compounds that would complicate wet-chemical processes.

This makes fire refining an increasingly important part of the circular economy for precious metals. As electronic devices proliferate and ore grades at traditional mines continue to decline, the relative importance of secondary sources, meaning recycled material, is growing steadily.

What Fire Refining Leaves Behind in the Historical Record

An unexpected application of understanding fire refining comes from archaeology. Ancient smelting and cupellation processes leave characteristic trace-element and isotopic signatures in the gold they produce, because fire refining does not remove every last atom of every impurity. In particular, lead isotope ratios in refined gold can sometimes be traced back to the ore body the gold came from, giving archaeologists a chemical fingerprint for ancient trade routes and mining activity.

This approach has been used to provenance metals including lead, silver, and bronze for decades, though it has significant limitations. Overlap in isotope composition between different ore sources remains a persistent challenge, and the refining process itself can alter trace-element profiles in ways that complicate interpretation.¹⁰ Still, the fact that fire refining leaves a readable chemical record in the gold means that ancient objects can sometimes tell us not just that they were refined by fire, but where the ore was dug out of the ground and, by extension, which trade networks carried it to its final destination.