extraction of metals from ores

Almost every metal we use sat for ages in the ground not as shiny metal but as a compound — an oxide, sulfide, or carbonate — because over geological time most metals have reacted with oxygen, sulfur, or water. Extracting a metal means winning back the electrons it lost: chemically reducing the ore's metal ions, M-n+, all the way down to the free metal, M. The whole story is one giant reduction, and the choice of method follows from how stubbornly each metal clings to its lost electrons.

There is a rough ladder of methods set by reactivity. Very unreactive metals (gold, sometimes copper) occur as the metal itself and just need physical separation. Moderately reactive metals are often smelted: carbon (as coke) reduces their oxides in a furnace — iron from Fe2O3 in a blast furnace is the textbook case — because hot carbon is a cheap, powerful reducing agent, as an Ellingham diagram makes precise. The most reactive metals (sodium, magnesium, aluminium) hold their electrons too tightly for carbon to pry loose economically, so they are extracted by electrolysis of a molten salt or oxide, where electricity supplies the reducing power directly (aluminium from molten alumina is the giant industrial example). Sulfide ores are usually roasted to the oxide first, and some metals are won 'wet' by leaching and then displacement or electrolysis.

Extractive metallurgy is one of the oldest and most consequential applications of redox chemistry — the Bronze and Iron Ages are literally named for it, and aluminium stayed a precious curiosity until cheap electrolysis arrived. The honest framing is that the method is dictated by thermodynamics (how negative the oxide's free energy of formation is) tempered by economics and kinetics: the 'best' reductant on paper may be too slow, too dirty, or too costly, which is why real plants are compromises, and why recycling — which skips the reduction entirely — is so much cheaper than primary extraction.