The Challenge
To obtain pure gold, free from contaminates, in order to make a souvenir. This must include the little nugget we were given, the alluvial gold, and the gold extracted from the mountain rock.

Smelting
Smelting is a way of obtaining metals from a mineral-bearing ore and dates back to pre-historic times. It usually involves the reduction (reduction reactions are those which take electrons from elements, reduction being the opposite of oxidizing reactions) of metal oxides (the ore) to metal and the formation of non-metal oxide waste known as slag. It is a chemical process and does not simply involve melting the gold. It sounds like alchemy, and it almost is.

Why do we need a furnace?
We decided to use a furnace. This is how metal has traditionally been smelted. It is the easy option because we are following plans that people have used for centuries to smelt iron. But why do we need a furnace? Couldn't we just use a bonfire?

Metal ores need to be heated to very high temperatures to produce pure metals - far higher than open fires can usually produce. Iron has a higher melting temperature than gold - iron melts at about 1800ºC and gold melts at 1062ºC - so we hope that, even with our resources, we should be able to get a furnace originally used to smelt iron to work with gold.

A furnace also helps maintain certain chemical conditions during smelting, for example, in iron smelting a reducing atmosphere is beneficial. We used a furnace design that has been used throughout known history. It was a rough variation of a historic design known as a Bloomery Shaft Furnace which was used to smelt iron many centuries ago.

Furnace features
Insulation helps prevent heat loss. We used a double skin design and clay-lined brick walls in order to make sure that the high temperatures created weren't lost to the atmosphere. Even though we expected over one thousand degrees Celsius within the furnace the outside shell never felt warm to the touch. In addition, we built the furnace on foundations of dry sand to prevent heat loss to the ground. Water or wet sand would be good conductors, allowing heat to escape from the furnace. Air, or dry sand are good insulators that prevent heat loss.

Fuel is important. We chose to use charcoal rather than wood. This is because it is more or less pure carbon, which provides the correct chemical environment, and because it reaches high temperatures when it is burned. The temperature at which charcoal burns depends on what wood it is made of. If it has been prepared from good hard wood, 600ºC to 1500ºC should be reasonably easy to achieve. Charcoal is a labour intensive fuel to make, however, and it takes 10kg of hardwood to make only 1kg of charcoal.

Plenty of oxygen is needed but in the right place. Chemically speaking, a reducing atmosphere is needed to smelt most metals successfully. In our furnace this was dependent on the presence of carbon monoxide. We can tell that carbon monoxide is present by the colour of the flame at the top of the furnace. We need to look for a blue flame. This shows that we have supplied enough oxygen to allow a hot enough flame, but not so much that we prevent the production of carbon monoxide (with a surplus of oxygen only carbon dioxide will be formed).

A balance is necessary: high temperatures need the fuel to burn well and require oxygen, but too much oxygen and there will be no reducing atmosphere. We can't monitor this exactly, and as a shortage of carbon monoxide means a little waste left in with the gold, but a shortage of oxygen means low temperatures and no smelting at all, we have to strike a compromise.

Why are bellows so important?
The key to supplying oxygen in roughly the correct quantity is in using bellows. But the job is not as straightforward as it first seems. We need the bellows to suck air from the atmosphere and blow it into the furnace. If air is blown in then sucked out of the furnace, no oxygen would be added to the fire. Worse still, we would suck hot air into the bellows and melt or possibly set fire to the bellows. To solve this problem, we made made a one-way valve.

Did we have problems?
Yes. The first problem was when the valve that was fitted to one of the bellows got stuck and a hot coal was sucked into it, setting fire to the leather. The bellows also got very hot. We got around this by cooling the pipe joining the bellows to the furnace with water. By and large, however, the bellows held together very well, especially considering the length of time that they were used.

How hot did we have to go?
We needed a high enough temperature to melt the gold plus a little more. Metals melt at different temperatures. For example, lead melts at 328ºC. If the gold is pure we need over one thousand degrees Celsius (1062ºC). That is a pretty high temperature. We will have to monitor it somehow to see if we can get there with the limited technology (and knowledge) available.

We made a simple thermometer by placing small pieces of lead, aluminium, brass and copper on a piece of brick. These metals melt at increasing temperatures. By taking the thermometer out and looking at the metals to see if they had melted we would know the temperature. However, the thermometer was bulky and couldn't be positioned in the centre of the furnace next to the gold, so Jonathan double checked the temperature by lowering a copper olive (a component used by plumbers to join pipes etc.) into the centre of the furnace using a piece of steel wire. The copper melted; we were in business.

So how did we do? Well, we definitely reached a high enough temperature, and we did manage to smelt the gold, but sadly it was very difficult to make just one nugget, which was what we wanted.

Creating a carbon arc
Jonathan decided to create a carbon arc. This is a method in which an electric current is forced to jump from one electrode to another in order to create an arc. When this happens very high temperatures (up to 3,000ºC) are generated which will be more than enough to melt the now pure gold. We used a car battery to supply the electric current and pinched some carbon rods from a big torch battery (the old fashioned flashlight batteries with two springs as terminals contain carbon rods, not AA type batteries) with which to make the electrodes. Jonathan then held the electrodes each side of our gold, struck an arc and melted it into one odd shaped nugget.

Was this the end of our problems?
Well nearly. During the carbon arc procedure, pits were created on the surface of the gold and the crucible was broken. I wanted to cheat and borrow some gas welding equipment from a farm workshop in order to melt the gold and smooth out the pits. Another option was to fire up the furnace again, but there really wasn't time, so instead Jonathan opted to polish the gold.

And the result?
A pendant, which looks a bit like a whitebait - food that the west coasters are passionate about. Or is it a shrimp? Or is it a spermatozoa? We'll leave it to your imagination!

pendant Copyrighted image Credit: Used with permission