Metallothermic reduction
Taking a metal out of its compound by offering a hungrier metal in exchange — the route to the elements carbon and electricity cannot reach.
Some oxides are too stable for carbon to take the oxygen away, and some metals form carbides instead of reducing when you try. For those, the answer is to offer a metal that wants the oxygen more: magnesium, aluminium, calcium or sodium, depending on what is being made and what can be afforded.
The reaction is a straightforward exchange and it is usually violently exothermic. The best-known instance is thermite — aluminium powder and iron oxide, which reaches around 2,500 °C and produces molten iron, and which is used to weld railway track in place because it needs no external power at all.
Processing
The choice of reducing metal is an economic argument dressed as a chemical one. Aluminium is cheapest and reduces most oxides; magnesium is used where aluminium would contaminate the product, as in the Kroll process for titanium and zirconium; calcium and sodium are reserved for halides and for the most stubborn oxides, and are handled with corresponding care.
The heat is both the mechanism and the problem. Enough of it makes the reaction self-sustaining once started, which is convenient; too much and the product melts into the crucible or the charge sprays out of it. Industrial practice is largely about moderating a reaction that would otherwise run away.
Separating the product from the slag is the other half of the job. The reduced metal and the reducing metal's oxide have to part cleanly, which they do when the temperature is high enough for both to be liquid and their densities differ enough for one to settle.
Economic significance
This is the expensive way to make a metal, and the elements made this way are produced in tonnes or hundreds of tonnes a year rather than in millions. That is not a failure of the method — it reflects what these metals are used for, which is small quantities in specific places.
Its largest use by tonnage is not making a metal at all but making a ferroalloy: ferrovanadium, ferroniobium and ferrotitanium are aluminothermic products, and they go straight into steel without the element ever being isolated.
How we know: checked recently · only one source, so there is nothing to cross-check it against · stated directly by the source.
How this connects
Where a connection has been confirmed by an outside reference, that reference is named beside it.
produces
- Beryllium — element · magnesium reducing beryllium fluoride, and the reason beryllium is expensive before anyone accounts for how carefully it has to be handled
- Boron — element · magnesium reducing boron oxide, which gives an impure amorphous boron — the crystalline element needs a different route entirely
- Barium — element · aluminium reducing barium oxide under vacuum, the barium distilling off as vapour
- Strontium — element · the same aluminothermic route as barium, and in similarly small quantity
- Caesium — element · calcium or barium reducing caesium chloride under vacuum — caesium is too reactive to survive most alternatives
- Rubidium — element · reduced from its chloride the same way, and usually as a by-product of the caesium it accompanies in pollucite
- Titanium — element · the Kroll process is this reaction — magnesium reducing titanium tetrachloride — and this is the general case it belongs to
takes as input
- Magnesium — element · the reducing metal where aluminium would contaminate the product, which is why titanium and zirconium go the magnesium route
- Aluminium — element · the cheapest reducing metal and the commonest, and the aluminium in thermite
- Calcium — element · reserved for halides and the most stubborn oxides, where the cheaper metals will not do the job
was succeeded by
- Kroll process — process · the Kroll process is the industrialised special case, for titanium and zirconium specifically
Sources
- Material WorldOur own writing