Although Charles Hall conducted his first successful run of the Hall-Héroult aluminium smelting process in the woodshed behind his house, it has ever since remained mostly out of reach of home chemists. It does involve electrolysis at temperatures above 1000 ℃, and can involve some frighteningly toxic chemicals, but as [Maurycy Z] demonstrates, an amateur can now perform it a bit more conveniently than Hall could.
[Maurycy] started by finding a natural source of aluminium, in this case aluminosilicate clay. He washed the clay and soaked it in warm hydrochloric acid for two days to extract the aluminium as a chloride. This also extracted quite a bit of iron, so [Maurycy] added sodium hydroxide to the solution until both aluminium and iron precipitated as hydroxides, added more sodium hydroxide until the aluminium hydroxide redissolved, filtered the solution to remove iron hydroxide, and finally added hydrochloric acid to the solution to precipitate aluminium hydroxide. He heated the aluminium hydroxide to about 800 ℃ to decompose it into the alumina, the starting material for electrolysis.
To turn this into aluminium metal, [Maurycy] used molten salt electrolysis. Alumina melts at a much higher temperature than [Maurycy]’s furnace could reach, so he used cryolite as a flux. He mixed this with his alumina and used an electric furnace to melt it in a graphite crucible. He used the crucible itself as the cathode, and a graphite rod as an anode. He does warn that this process can produce small amounts of hydrogen fluoride and fluorocarbons, so that “doing the electrolysis without ventilation is a great way to poison yourself in new and exciting ways.” The first run didn’t produce anything, but on a second attempt with a larger anode, 20 minutes of electrolysis produced 0.29 grams of aluminium metal.
[Maurycy]’s process follows the industrial Hall-Héroult process quite closely, though he does use a different procedure to purify his raw materials. If you aren’t interested in smelting aluminium, you can still cast it with a microwave oven.
Sweet! Always good to see industrial processes being made more accessible. Still waiting for someone to try the FFC Cambridge process to extract titanium or silicon.
I enjoy seeing these kinds of experiments a lot. I think it would be good to mention the work by Cody (from the YouTube channel Cody’s lab) who has a video with nearly the exact same procedure from 8 years ago.
It’s a real shame that it turned out the internet was not to be for more things like this. Cody himself has said that Youtube these days just isn’t very welcoming towards sharing anything that could be perceived as unsafe. Youtube would be much happier if people only made heaps of low effort content, which has tainted the reputation of all Youtubers as people who only make heaps of low effort content.
Al2O3 + 3Cl2 + 3CO —> 2AlCl3 + 3CO2
Reduction of gaseous AlCl3 to metal requires less energy than reduction of Al2O3. From patent.
Neat! Now try it with aluminum oxide dissolved in cryolite.
If I understand the description correctly, I think that is exactly what he did — the text described it as “using cryolite as a flux”. Since this was done specifically to overcome the high melting point of aluminum oxide, that implied to me that the oxide was being dissolved in molten cryolite. I could be wrong though.
I’ve played with little electric melters like the one pictured — you certainly aren’t going to be melting straight aluminum oxide in that. They max out around 1000-1200 C. Alumina is melting another 1000 degrees higher, give or take. I don’t think we have any resistive heating elements that haven’t melted themselves by that point, and a lot of the materials used for furnace walls are having a tough time too. People have gotten there in their backyard, but it’s challenging and usually involves a flame and compressed air.
As a side note, the term “flux” in the context of metallurgy seems to be frustratingly ambiguous. As best I can gather, it seems to just mean “magic powder we dump in the melt to achieve some desired effect”. That might mean selectively dissolving impurities, acting as a sacrificial acceptor of oxygen, lowering a melting point, etc. Perhaps it reflects that some of these processes predate our understanding of the actual chemistry? I am sure that modern metallurgists understand very specifically what they are trying to achieve in each case, but I feel like it leads to a lot of confusion in hobby efforts.
I work in the industrial production of aluminium, the cryolite flux in this case is Na3AlF6. It is generally referred to as bath.
Not quite a flux in a typical metallurgical sense. But you are right in being a eutectic, solubility of alumina, as well as being slightly lower density when molten compared to aluminium – allowing the metal to sink to the bottom of the cell.
Hall and Heroult were truly geniuses of their time.
Yes. The term is exactly that, an additive to do “something”, whether it be caustic, etc. see: bauxite.