My inorganic chemistry is at a basic level, so my question is, if I heat alumina ($\ce{Al2O3}$), to its melting point around 2000 C, will I wind up with pure Al? I think the oxygen will release from the $\ce{Al2O3}$ when the temperature approaches 2000 C.

I bought 5 kilograms from a friend who works at a refinery near my town (he works in logistics and does not have a chemical background) with 1.75$ of almost pure alumina ~99.95%, and I want to know if I melt that white powder will I obtain 2.5kg of 99.95% aluminium?


2 Answers 2


You've figured out one half of the Hall-Heroult process. Just melting alumina on its own won't liberate the oxygen. The reason is difficult to explain without getting into chemical thermodynamics - one way to think about it is that the lowest energy state is one where oxygen is bound to alumina - where the electrons from alumina have been "taken" by the oxygen. In diatomic oxygen, the electrons are "shared" (the bonds are covalent) and so the total energy is higher. This is a very brief and not-very-detailed explanation, but the end result is that oxygen and aluminum, if given the opportunity, would rather form a bond than not.

When you add thermal energy, unless it changes that balance (it doesn't, at least not until you get way past the melting point), all you are doing is making it more possible for the reaction to happen. If you kept going, you could eventually add enough thermal energy to completely dissociate the alumina, at which point (depending on how you did it) you might be able to remove the oxygen and recover the aluminum.

However - the amount of energy it would take to do that is extremely large. Instead, it turns out to be much easier to simply force oxygen to give its electrons back to aluminum using an electric potential. When you do that, you electrolyze the compound and create diatomic oxygen at the anode and aluminum at the cathode.

The Hall-Heroult process takes this another step to make this work at an even lower temperature (1000 C) by dissolving the alumina in molten cryolite. As a result, it is one of the most (maybe the most) efficient industrial process for producing aluminum, and is responsible for the widespread availability and low cost of aluminum in modern times (and also for the existence of Alcoa).

Your idea about using solar energy isn't bad - although you might be surprised at the amount of energy you need to produce aluminum - both to melt it, and to electrolyze it. One thing you might look into is using solar thermal heating (with a collector, for example) to melt the alumina/cryolite mixture, and then use photovoltaics just for the electrolysis step. That would be quite a bit cheaper than using photovoltaic electricity alone to run an electric heater.

Industrially, the cryolite mixture is kept molten by passing a large electric current through it. The electricity is usually produced by power plants, since the amount required is so large (15 kWh/kg according to Alcoa, and that's just for the electrolysis step!). Although coal and gas power isn't renewable, the advantages are a relatively low cost per unit of power (which might not stay true in the next few decades) and on-demand power generation (likely to always be a big advantage over solar).

There is also nuclear, hydroelectric, and geothermal power - all of which provide large amounts of on-demand power very inexpensively, once the plants are built.

  • $\begingroup$ Aluminum is the third most abundant element, however it was difficult to extract and purify prior to the Hall–Héroult process which came about in the 1880's. Despite its natural abundance, aluminum was considered a precious metal before the 1880's. Napoleon's most important guests were given aluminum cutlery while more ordinary guests dined with silver. $\endgroup$
    – ron
    Commented Jul 12, 2014 at 0:41
  • $\begingroup$ @ron - another interesting fact: the Washington monument was originally capped with aluminum, because it was so valuable. There is also a legend that the roman emperor Tiberius had a man executed who had claimed to have discovered a way of refining aluminum (or what is assumed to be aluminum), because he was afraid it would make his stockpile less valuable. $\endgroup$
    – thomij
    Commented Jul 12, 2014 at 0:48
  • $\begingroup$ Now I understand what I had missed in my mind, so the process is very expensive but just in my opinion is much more expensive using chemical agents like cryolite (AlF3,3NaF), cryolite goes at about 600-700 usd a tonne FOB. about 700 two tonnes of alumina ~= 1400 usd without electrolisys which is intensive power drain (consumable carbon anodes) water, what about Environmental Taxation and so on... At LME aluminium spot price is 1936 usd a tonne, and now I come with a second question, what really is the profit there? I mean a refinery will work with profits like 100 usd / tonne? $\endgroup$
    – LXSoft
    Commented Jul 12, 2014 at 1:26
  • $\begingroup$ I don't know what the profit margins are, but I would imagine they are thin since it is a very energy intensive process and since the core technologies are out of patent protection. It will also depend strongly on commodity prices and worldwide demand - but in general I would guess that you need a large, efficient operation with lots of capability to absorb price fluctuations - which is probably why Alcoa has survived by buying up competitors. About the cryolite though - remember most of it can be recycled. $\endgroup$
    – thomij
    Commented Jul 12, 2014 at 1:31
  • $\begingroup$ Because its position in the reactivity series of metals, aluminium cannot be extracted using carbon because it is above carbon in the reactivity series (more reactive than carbon in the series) carbon is not reactive enough to displace aluminium from its compounds such as aluminium oxide. But what about using calcium metal? It is more reactive than aluminium. [EDIT: Lowest melting point than aluminium, but what about dissolute aluminium oxide in some acids and then reactive series?] $\endgroup$
    – LXSoft
    Commented Jul 12, 2014 at 2:44

Passing aluminium oxide through a hot enough flame will result in the formation of sapphires, ruby and corundum (amongst others) type gem stones also known as the 'Verneuil process'.
A lot of man made stones are made in similar way. Here is a video to explain this.

  • $\begingroup$ Welcome to chemistry.se! If you have questions about how to beautify your posts, have a look at the help center. Do you want to know more about this site, please take the tour. $\endgroup$ Commented Jan 9, 2015 at 7:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.