How to Separate Gold From Other Metals by Melting

Separating gold from other metals by melting relies on a handful of well-established pyrometallurgical techniques, each exploiting the fact that gold behaves differently from base metals and even from other precious metals when exposed to heat, air, or reactive chemicals at high temperatures. The most common approaches include smelting with a chemical flux, cupellation, and chlorine-based refining, and the right choice depends on whether you are starting with raw ore concentrates, scrap jewelry, or electronic waste. None of these methods is as simple as “just melting everything and pouring off the gold,” because gold alloys freely with metals like silver and copper, so you need a chemical or physical trick to drive those other metals away.

Why Melting Alone Does Not Separate Gold

Gold melts at about 1,064 °C, and many of the metals it is commonly found with have similar or lower melting points. Copper melts at roughly 1,085 °C, silver at about 962 °C, and tin at only 232 °C. When you heat a mixed-metal sample past these temperatures, the metals do not sort themselves into separate layers the way oil and water do. Instead, they dissolve into each other and form an alloy. If you melt down a piece of 14-karat gold jewelry, for instance, you end up with a single molten pool of gold, copper, silver, and possibly zinc all mixed together. Cooling it back down just gives you the same alloy in a new shape.

To actually isolate the gold, you need to introduce something at high temperature that reacts with the unwanted metals but leaves the gold untouched. Gold’s famous chemical stubbornness is the key here. It resists oxidation, meaning it does not combine with oxygen the way copper, lead, tin, and zinc do. It also resists attack by most single acids. Every separation method described below takes advantage of that inertness in one way or another.

Smelting With Flux

Smelting is typically the first thermal step for anyone working with ore concentrates, fine placer gold mixed with black sands, or crushed electronic scrap. You combine the material with a flux, which is a mixture of chemicals designed to lower the melting point of the waste rock (called gangue) and help the metals collect together at the bottom of the crucible. A common flux recipe for small-scale gold smelting includes borax, soda ash, and sometimes silica sand or powite. Borax alone can work for very simple jobs like consolidating fine gold dust.

When the crucible reaches roughly 1,100–1,200 °C in a furnace or with a torch, the flux melts the non-metallic material into a glassy slag that floats on top, while the heavier metallic fraction sinks to the bottom. After cooling, you crack the slag away and recover a metal button. That button, however, is not pure gold. It contains whatever other metals were present in the starting material, including silver, copper, lead, and iron. Smelting concentrates the gold by removing the rock and sand, but it does not separate gold from other metals within the metallic fraction. You need a second refining step for that.

Cupellation

Cupellation is one of the oldest methods for purifying gold and silver, and it works beautifully when your metal button contains mostly lead along with the precious metals. The technique has been used in assay offices for centuries and is still the standard fire-assay finish in analytical labs worldwide. The idea is straightforward: lead oxide is absorbed by a porous ceramic cup called a cupel, while gold and silver are not.

In practice, you first alloy your sample with a known excess of lead, then place the resulting lead button in a cupel made from compressed bone ash or magnesia. When heated under oxidizing conditions (with plenty of air flow), the lead melts and its surface oxidizes rapidly. The lead oxide, being a liquid at cupellation temperatures, wicks into the porous cupel like water into a sponge, carrying along oxides of copper, tin, zinc, and other base metals with it. Gold and silver, which do not oxidize under these conditions, remain behind as a small bright bead sitting on top of the cupel.1Nevada Bureau of Mines and Geology. Science of the Comstock – Chemistry

If your goal is pure gold and the bead contains both gold and silver, you still have one more step. The gold-silver bead from cupellation is typically dissolved in nitric acid, which attacks the silver but leaves the gold intact. This wet-chemistry step, called “parting,” is the usual follow-up to cupellation rather than another melt-based process. It highlights a practical reality: melting-based methods often get you most of the way to pure gold, but the final separation from silver usually involves acid.

Inquartation and Parting

When your starting alloy is mostly gold with some silver in it, nitric acid struggles to penetrate and dissolve the silver because the gold matrix shields it. The workaround is inquartation: you deliberately add enough silver (or sometimes copper) so that the alloy becomes roughly three parts silver to one part gold by weight. At that ratio, the alloy is porous enough that nitric acid can reach and dissolve the silver throughout the mass.

You melt the gold and the added silver together in a crucible, pour the alloy into water to granulate it into small pieces (which increases the surface area), and then boil those granules in concentrated nitric acid. The silver dissolves into the acid as silver nitrate, while the gold remains as a dark, spongy residue. After filtering and washing, you melt that residue into a gold button that is typically 99.5% pure or better. The melting step here serves two purposes: creating the properly proportioned alloy at the start, and consolidating the purified gold at the end.

Inquartation and parting is the method most commonly used by small-scale refiners and jewelers because it requires relatively modest equipment: a torch or small furnace, nitric acid, and basic safety gear. It does not scale well to industrial volumes, which is where chlorine-based refining takes over.

The Miller Process

For large-scale refining, the Miller process is the workhorse. Developed in the 1860s, it involves bubbling chlorine gas through molten impure gold. At roughly 1,100 °C, the chlorine reacts with silver, copper, zinc, lead, and most other metals present, turning them into metal chlorides. These chlorides are lighter than the molten gold, so they rise to the surface as a slag that can be skimmed off. Gold itself does not react with chlorine at these temperatures in the presence of excess base metals, so it stays behind in the melt.

The process is fast, usually finishing within an hour or two, and produces gold of about 99.5% purity. That is good enough for many commercial applications, including producing standard gold bars. When higher purity is needed (99.99% or “four nines”), the Miller-process gold is typically sent through an electrolytic refining step (the Wohlwill process), which is not a melting method but an electrochemical one.

One limitation of the Miller process is that silver chloride, formed during the refining, is volatile at the operating temperature and can be lost as fumes. The silver chloride slag, however, is valuable and is collected for silver recovery. Running a Miller process also requires serious infrastructure: fume hoods rated for chlorine gas, acid-resistant crucibles (often silica or magnesia), and trained operators. This is not a backyard operation.

Recovering Gold From Electronic Waste by Smelting

A growing application for pyrometallurgical gold separation is the processing of electronic waste, particularly printed circuit boards. Circuit boards contain small amounts of gold in their connectors and plating, along with copper, tin, lead, nickel, and silver. Industrial recyclers process this material by feeding it into copper smelting furnaces, where the organic material (fiberglass, plastic) burns off or is captured in slag, and the metals collect in a molten copper matte at the bottom.

Research on the behavior of waste circuit board material in copper matte smelting has shown that pieces of circuit board melt completely within two to five minutes at smelting temperatures, after which molten alloy droplets containing copper, lead, tin, nickel, gold, and silver form and descend toward the bottom of the crucible.2Metals. Behavior of Waste Printed Circuit Board (WPCB) Materials in the Copper Matte Smelting Process The gold reports almost entirely to the copper matte phase, meaning it follows the copper rather than ending up in the slag. From there, the copper matte goes through conventional copper refining (converting and electrorefining), and the gold is recovered from the anode slimes produced during electrolysis.

This route is efficient for large recyclers but impractical for individuals. The volumes of circuit boards needed to yield meaningful gold are enormous, the equipment is industrial-scale, and the fumes from burning circuit board material are toxic. Small-scale e-waste gold recovery more commonly uses chemical leaching methods rather than smelting.

Choosing the Right Method for Your Situation

The best melting-based separation method depends on what you are starting with and how pure you need the final product to be.

  • Fine placer gold or black sand concentrates: Smelting with borax flux consolidates tiny particles into a single button. If the button is already high purity (which placer gold often is, depending on the source), you may not need further refining. If it contains significant silver or copper, follow up with inquartation and parting or cupellation.
  • Scrap jewelry: Inquartation and parting is the standard approach for small batches. You know roughly what karat the jewelry is, so you can calculate how much silver to add. Melt together, granulate, treat with nitric acid, and remelt the residue.
  • Ore concentrates with high sulfide content: Smelting with a lead-collecting flux followed by cupellation works well and is essentially the classical fire-assay method scaled up slightly. The lead captures the precious metals, and cupellation removes the lead.
  • Large volumes of impure doré bars (gold-silver alloy from mines): The Miller process handles this efficiently at industrial scale, producing 99.5% gold in a continuous operation.
  • Electronic scrap: Feed into a copper smelting circuit if you have access to one. Otherwise, chemical methods are more accessible for small quantities.

Safety Considerations

Every method discussed here involves serious hazards that you should not underestimate. Smelting and cupellation expose you to temperatures above 1,000 °C, with risks of burns, crucible failure, and molten metal splashes. Lead fumes from cupellation are toxic with chronic exposure, so adequate ventilation or a fume hood is not optional. Inquartation and parting requires boiling concentrated nitric acid, which produces choking nitrogen dioxide fumes that can cause severe lung injury. This step should only be done outdoors or under a proper fume extraction system, never in an enclosed room.

The Miller process multiplies the danger by introducing chlorine gas, which is acutely toxic even at low concentrations. Industrial facilities manage this with sealed reactor vessels, dedicated off-gas treatment systems, and continuous monitoring. Pyrometallurgical off-gas treatment in general is a complex engineering challenge involving gas cooling, particulate capture, and pollutant scrubbing, with the specific design depending on the metals being processed and the scale of the operation.3Nexus Science Review. Cooling and Off-Gas Treatment in Pyrometallurgical Processes: Technologies, Challenges, Industrial Integration, and Future Perspectives—A Critical Review None of this applies to the hobbyist with a propane torch, which is another way of saying the Miller process is not for hobbyists.

Even borax-flux smelting, which is the gentlest method on this list, produces silicate slag that is extremely hot and can shatter if it contacts water. Wear a full face shield, heavy gloves, and leather or flame-resistant clothing. Work on a fireproof surface away from anything combustible, and have a plan for where the molten material will go if the crucible cracks. These precautions sound excessive until you have watched a crucible fail at 1,200 °C.

Common Misconceptions About Melting and Gold Purity

One persistent myth is that you can determine gold purity just by melting a sample and observing its color or behavior. In reality, gold alloys can look extremely similar to pure gold in the molten state, and small amounts of copper or silver do not produce obvious visual changes during melting. A torch test can tell you whether something is gold-bearing (gold melts into a bright, stable bead rather than oxidizing away), but it cannot tell you the purity. For that, you need either an assay or an acid test on the solid metal.

Another misconception is that adding borax to a melt “purifies” gold. Borax is a flux: it dissolves metal oxides and helps them separate from the metallic phase. If you melt dirty placer gold with borax, the borax will clean up the slag and help the gold consolidate, but it will not remove silver, copper, or other noble-ish metals that are alloyed with the gold. Your button will be shinier and better consolidated, but no purer in terms of gold content than it was before. The borax just stripped away the non-metallic junk.

A third misunderstanding involves the idea that gold “floats” or “sinks” away from other metals during melting because of its high density. While gold is indeed very dense (about 19.3 g/cm³), density separation does not work in a fully molten, alloyed state. All the metals are dissolved in each other, forming a single liquid phase. Density-based separation, such as panning or centrifugal concentration, works with solid particles in water, not with molten alloys. In the melt, you need chemistry, not physics, to get the gold to let go of its companions.

What “Four Nines” and “Three Nines” Actually Mean

When refiners talk about purity levels, they use a shorthand based on the number of nines. “Three nines” gold is 99.9% pure (also written 999 fine), and “four nines” is 99.99% (9999 fine). Most melting-based methods top out at roughly 99.5% purity. The Miller process can sometimes push to 99.9% with careful operation, but consistently hitting 99.99% requires electrolytic refining, where gold is dissolved from an impure anode and plated onto a pure cathode in an acid bath. That is not a melting process at all, though the final cathode gold is always melted into bars or grain afterward.

For most practical purposes outside of electronics manufacturing and certain scientific applications, 99.5% gold is perfectly acceptable. Standard “good delivery” bars accepted by major exchanges are required to be at least 99.5% pure. Jewelry alloys are intentionally less pure, with added metals for hardness and color, so a jeweler recovering gold from scrap usually aims for 99.5% and then re-alloys to the desired karat.

When Melting Is Not the Best Path

Melting-based methods are not always the smartest choice, even when they can technically work. If you are dealing with very low-grade material, such as ore with only a few grams of gold per ton, smelting the entire mass is wildly impractical. The amount of flux and fuel needed would dwarf the value of the recovered gold. For low-grade material, chemical leaching (using cyanide solutions or thiosulfate in industrial settings, or less toxic alternatives like ferric chloride or hydrochloric acid/bleach mixtures in smaller operations) extracts the gold into solution first, and then it is precipitated or electroplated out.

Similarly, if your gold is chemically bonded at a microscopic level within sulfide minerals like pyrite or arsenopyrite (called “refractory” gold), simply melting the ore will not free it. The gold particles are too small and too locked within the mineral structure. These ores require either roasting (heating in air to break down the sulfides) or pressure oxidation before the gold can be collected by conventional smelting or leaching. Refractory gold deposits are actually common, and they are the reason many gold mines use multi-stage processing rather than a straightforward smelt.

For anyone working at home or in a small shop, the practical ceiling for melt-based gold separation is a few hundred grams of material at a time. Beyond that, you are looking at custom furnace builds, serious fuel costs, and engineering challenges around fume management that push you toward professional refining services. Many hobbyist prospectors and small jewelers find that the most cost-effective approach is to consolidate and partially purify their gold with a simple flux smelt, then ship the resulting button to a professional refiner who handles the final purification and pays based on assayed gold content.