Smelting tin

The tin ore

photo by Hans Splinter

All tin during the bronze age was smelted from one type of ore called cassiterite. It's a rare material, which only occurs in a few locations in Europe. One of the locations is Cornwall, UK. During the bronze age, the ore could be collected as heavy, dark tinstone pebbles found in glacial deposits. These are very high in tin concentration, and could be smelted straight. These are practically all gone, so I had to work with mined cassiterite, which has a much lower tin %. I got about 2kg of this ore, of which I don't know the tin %-age. Here you can see an example of mined tin ore with a very high cassiterite contents.
In order to be able to smelt this into tin, the tin contents first has to be increased. Whether this process was done during the bronze age I don't know, but it could have been done when high purity tin ores started running out. The process I went through may therefore not be typical for the bronze age.

The chemical principles of smelting tin ore

The chemical formula of cassiterite is SnO2. This means that it's an oxide, where the tin is chemically bonded to oxygen. In order to turn cassiterite into tin, it has to be reduced. Reduction is the opposite of burning, were the metal is releasing it's oxygen. Reduction is a chemical reaction. To reduce cassiterite into tin, two things are needed: heat and carbonmonoxide (CO). This gives the following chemical reaction:

The fuel used is charcoal, which provides both the heat and the carbonmonoxide.

Concentrating the tin ore

photo by Hans Splinter

In order to concentrate the tin ore, it first has to be crushed to powder. For this I used an old broken saddlestone, originally used for grinding corn. The ore is stamped and ground to a fine powder. This took a couple of hours work.

photo by Hans Splinter

The photo above shows the crushed ore being concentrated by panning. This works the same as panning gold: as the ore is much heavier then the rock in which it's embedded (matrix, consisting of f.e. quartz), by shaking the ore in a bowl filled with water, the ore will sink to the bottom faster. The matrix will sit on top, and by washing water over it, it will wash the matrix off the ore.

photo by Hans Splinter

Here you can see the ore in the wooden bowl used to pan the ore. The light blueish grey and white material is the matrix, the dark grey material below is the cassiterite.

Above you can see the concentrated tin ore. There is still a lot of matrix included, but clearly the tin ore contents has increased due to the color being much darker then before the panning. Of about 1.5kg of the ore which I concentrated, I have 700grams of material left which will be smelted. I didn't know how far I'd have to go, as I don't know the original tin concentration. Before I started the panning, I wasn't really sure either what was the tin ore, so I was learning along te way, and hoping that I'd have it right.

Here you see me operating the smelting furnace. The furnace is similar to the copper smelting furnace I used (see here), but a smaller size. The smaller size is both due to having less ore to process, and because the ore dust heats up quickly, so it would react faster before reaching the bottom. As I wanted to prevent any reduced tin from oxidizing again, I made a hole at the front of the furnace, so the tin could run out directly. This did not happen though. The reason was that I got a lot of slag from the remaining matrix mixed with the ore, which blocked the tin from running out. I only got a few droplets running out of the front. The furnace was preheated with charcoal, and then layers of charcoal and hands of tin ore powder were added alternatingly. The charcoal I used was finer, to prevent the fine powder from falling through. This is also a reason why I kept the size of the furnace small, as with the small charcoal, the heat zone is much smaller.

After the smelt, I flipped the furnace over. Sorting out the contents, I found a lot of slag with tin embedded. I crushed up all the slag, and got all of the tin out. This resulted in a total of 60 grams of tin.

Here is most of the tin. I can definately see one advantage of increasing the tin concentration, as it will result in the tin joining together in larger amounts, making it easier to recover. If the tin contents would have been as high as in the original tinstones, it would have run out of the furnace freely, with much less slag forming.

After this I did two more experiments. The first was smelting the left over material from the concentrating process. After I emptied the furnace, I found only tiny tin particles, too tiny to recover. Crushing up the slag revealed nothing further. So it seems the concentrating process was succesful.

In the final experiment, I used the last remaining tin ore, which I only crushed to about 5mm pieces average and smelted that without concentrating. This gave only about 5 grams of tin max, from roughly 500 grams of ore (1% tin). The concentrated ore had a much higher yield, with 60 grams from roughly 1500 grams of ore (4% tin). Now it has to be considered that the last smelt had a lot of the ore fragments that I noticed had less ore in them, and I kinda rushed the last smelt by bellowing faster (hotter temperature, more oxygen increasing the chance of tin burning again). So these are probably factors that decreased the yield as well.

As I'm aiming at making at least 1kg of bronze, I'm going to do some more tin smelts in the future (next season). I may do some more experimenting, to see what the effects are of other methods to increase the tin yield, such as adding lime to replace the tin in the slag.