General Principles of Metallurgy

Metallurgy looks like industrial trivia until you notice that every choice in it is a thermodynamic decision. Why is aluminium extracted electrolytically and iron in a blast furnace? Why does coke reduce iron oxide only above about 1000 K? Why is magnesium used to displace titanium rather than the reverse? All four questions have the same answer — compare the two ΔG lines and pick the one that lies lower — and that is what the Ellingham diagram is for. Read this chapter as applied thermodynamics rather than as a list of processes, and the number of facts you have to remember drops by more than half.

Concentration of the Ore

Choosing a concentration method
MethodProperty exploitedUsed for
Hydraulic washingDifference in densityHeavy oxide ores such as haematite
Magnetic separationOne component is magneticMagnetite; removing wolframite from tin ore
Froth flotationThe ore is wetted by oil, the gangue by waterSulphide ores — galena, zinc blende, copper pyrites
LeachingThe ore dissolves selectively in a reagentBauxite in NaOH; gold and silver in cyanide
🎯 Why bauxite needs leaching rather than washing
Bauxite's impurities are iron oxide and silica, whose densities are too close to alumina's for gravity separation to work and which are not magnetic. But alumina is amphoteric and the impurities are not, so hot concentrated NaOH dissolves the alumina as sodium aluminate and leaves Fe₂O₃ behind as red mud. Passing CO₂ through the filtrate then reprecipitates pure hydrated alumina, which is calcined to Al₂O₃. This is the Bayer process, and it is a clean example of the general principle that a chemical difference beats a physical one whenever the physical properties are similar.

After concentration the ore is converted to oxide, because oxides are the easiest form to reduce. Calcination heats the ore in limited air and is used for carbonates and hydrated ores — it drives off CO₂ or water. Roasting heats it in excess air and is used for sulphides — it drives off SO₂ while oxidising the metal. So ZnCO₃ is calcined and ZnS is roasted, and confusing the two is a standard one-mark loss.

The Ellingham Diagram and Choice of Reductant

An Ellingham diagram plots ΔG° of oxide formation against temperature. Because oxidation consumes gaseous oxygen, ΔS is negative for most metals, so −TΔS is positive and the lines slope upward. The rule that makes the diagram useful is simple: a metal whose line lies lower will reduce the oxide of a metal whose line lies higher, because the coupled reaction then has a negative overall ΔG.

🎯 Why the carbon line slopes downward, and why that changes everything
For the reaction 2C + O₂ → 2CO, one mole of gas becomes two, so ΔS is positive and the line slopes downward — the only important line on the diagram that does. Consequently the C/CO line eventually crosses below every metal-oxide line, and above the crossing temperature coke will reduce that oxide. Iron oxide is reduced above roughly 1000 K, which is exactly why a blast furnace runs hot; zinc oxide needs about 1200 K. The same reasoning explains why coke is useless for aluminium: their lines would only cross above about 2500 K, far beyond what a furnace can sustain, so aluminium must be won electrolytically. One downward-sloping line therefore explains the entire division of metallurgy into carbon reduction and electrolysis.
Which extraction route, and why
MetalRouteReason
Fe, Zn, Pb, SnCarbon reduction in a furnaceThe C/CO line crosses below their oxides at attainable temperatures
Al, Na, Mg, CaElectrolysis of the molten salt or oxideTheir oxides are too stable for any chemical reductant
Cr, MnAluminothermite — reduction by aluminiumAl lies below them and the reaction is strongly exothermic
Ti, ZrDisplacement by Mg or Na (Kroll process)Carbon would form a carbide instead of the metal
Cu (from sulphide)Self-reduction — no external reductantCu₂S + 2Cu₂O → 6Cu + SO₂ once partly roasted
Ag, AuCyanide leaching then displacement by zincThey occur native and their oxides are barely stable
⚠️ The Ellingham diagram tells you nothing about rate
A negative ΔG means the reduction can happen, not that it will happen quickly — the diagram is pure thermodynamics, and kinetics is absent from it. This is why real furnaces run far hotter than the crossover temperature demands: the extra heat buys rate, not feasibility. It is also why the diagram cannot warn you that carbon reduces titanium oxide to titanium carbide rather than to the metal; that is a matter of which product forms, and it is the reason the Kroll process uses magnesium despite carbon looking adequate on the diagram. When a question asks why a thermodynamically permitted route is not used industrially, the answer is almost always rate or an unwanted product.

Refining the Crude Metal

Refining methods and what each depends on
MethodPrincipleExample
DistillationThe metal is far more volatile than the impurityZn, Hg
LiquationThe metal melts and flows away from the impuritySn
Electrolytic refiningImpure metal as anode, pure as cathodeCu, Ag, Al, Zn
Zone refiningImpurities stay in the melt rather than the solidSi, Ge for semiconductors
Vapour phase (Mond, van Arkel)A volatile compound forms and then decomposesNi via Ni(CO)₄; Ti and Zr via the iodide
ChromatographyDifferent adsorption on a stationary phaseLanthanoid separation
🎯 Why zone refining gives semiconductor-grade purity
Impurities are more soluble in a molten metal than in its solid, so when a narrow molten zone is moved slowly along a rod, the impurities travel with the melt and concentrate at one end, which is then cut off. Repeating the pass drives the purity to parts per billion — the level silicon needs before doping, because a semiconductor's conductivity is set deliberately by added dopants and any accidental impurity would swamp that control. The requirement is not fussiness: at one part per million of unintended impurity, a silicon chip simply does not work as designed.

Two industrial specifics are worth holding. The Hall–Héroult cell dissolves alumina in molten cryolite with fluorspar, which lowers the working temperature from over 2000 °C to about 950 °C and makes the process economic at all; the carbon anode is consumed as CO₂, which is why aluminium smelting is so carbon-intensive. And in a blast furnace the roles are layered by temperature: coke burns to CO at the base, CO reduces the oxide higher up, and limestone decomposes to CaO which removes silica as a calcium silicate slag — so limestone is a flux, not a reductant, a distinction questions test directly.

Key takeaways

  • Froth flotation is for sulphide ores; leaching is used when the ore differs chemically rather than physically.
  • Calcination is limited air for carbonates; roasting is excess air for sulphides.
  • On an Ellingham diagram, the metal whose line lies lower reduces the oxide of the metal above it.
  • The C/CO line slopes downward because 1 mole of gas becomes 2 — that is why coke works above a crossover temperature.
  • Al, Na, Mg and Ca must be extracted electrolytically, since no chemical reductant lies below their oxides.
  • The diagram is thermodynamics only — it cannot predict rate, nor that carbon gives titanium carbide.
  • Zone refining exploits impurities being more soluble in the melt, and gives semiconductor-grade silicon.
  • In a blast furnace limestone is a flux that slags off silica, not a reducing agent — CO does the reduction.

Practice questions (8)

Attempt each one before opening the answer. Every explanation names the tempting wrong option as well as the right one, because that is where marks are lost.

  1. Froth flotation is particularly suited to concentrating:

    1. Oxide ores
    2. Sulphide ores
    3. Carbonate ores
    4. Native metals
    Show answer

    Answer: B — Sulphide ores

    Sulphide ore particles are preferentially wetted by pine oil while the gangue is wetted by water, so the ore rises with the froth and is skimmed off. Oxide and carbonate ores are wetted by water like the gangue, so flotation cannot separate them and gravity or magnetic methods are used instead. Galena, zinc blende and copper pyrites are all concentrated this way.
  2. On the Ellingham diagram the line for 2C + O₂ → 2CO slopes downward because:

    1. The reaction is endothermic
    2. One mole of gas becomes two, so ΔS is positive
    3. Carbon is a strong reducing agent
    4. CO is unstable at high temperature
    Show answer

    Answer: B — One mole of gas becomes two, so ΔS is positive

    For most metals oxidation consumes gaseous oxygen and ΔS is negative, so −TΔS is positive and the line rises. Here the gas mole count increases from one to two, making ΔS positive and −TΔS increasingly negative as T rises — so the line falls. That downward slope is precisely why coke becomes a viable reductant above a crossover temperature, and why the entire distinction between furnace and electrolytic extraction exists.
  3. Aluminium cannot be extracted by carbon reduction because:

    1. Al₂O₃ is amphoteric
    2. The C/CO line crosses below the Al₂O₃ line only above about 2500 K
    3. Aluminium reacts with carbon monoxide
    4. Bauxite contains too much iron oxide
    Show answer

    Answer: B — The C/CO line crosses below the Al₂O₃ line only above about 2500 K

    Alumina is exceptionally stable, so its Ellingham line lies very low and the carbon line only drops beneath it at temperatures no industrial furnace can sustain. Electrolysis sidesteps the problem by supplying electrical rather than thermal driving force. Alumina being amphoteric matters at the leaching stage of the Bayer process, not at reduction, which is why that option is true but irrelevant here.
  4. The role of limestone in a blast furnace is to act as:

    1. A reducing agent
    2. A flux that removes silica as slag
    3. A catalyst
    4. A source of extra heat
    Show answer

    Answer: B — A flux that removes silica as slag

    Limestone decomposes to CaO, which is basic and combines with the acidic silica impurity to give a molten calcium silicate slag that floats on the iron and is drawn off. The reduction itself is done by carbon monoxide generated from coke. In fact limestone decomposition is endothermic, so it absorbs heat rather than supplying it — the opposite of the last option.
  5. Zone refining is used for silicon and germanium because:

    1. They are highly volatile
    2. Impurities are more soluble in the molten metal than in the solid
    3. They form volatile carbonyls
    4. They are magnetic
    Show answer

    Answer: B — Impurities are more soluble in the molten metal than in the solid

    Because impurities prefer the liquid phase, moving a narrow molten zone along a rod sweeps them to one end, which is discarded; repeated passes reach parts-per-billion purity, which is what semiconductor doping requires. Volatile carbonyl formation is the basis of the Mond process for nickel, and volatility itself is what distillation exploits for zinc and mercury — different methods for different properties.
  6. ZnS is converted to ZnO by:

    1. Calcination in limited air
    2. Roasting in excess air
    3. Leaching with NaOH
    4. Electrolysis
    Show answer

    Answer: B — Roasting in excess air

    Roasting heats a sulphide ore in excess air, oxidising the metal to its oxide and expelling the sulphur as SO₂. Calcination is the parallel process in limited air and is used for carbonates and hydrated ores, so ZnCO₃ would be calcined while ZnS is roasted — mixing the two up is one of the most frequent single-mark losses in this chapter.
  7. Titanium is extracted using magnesium rather than carbon because:

    1. Carbon cannot reduce TiO₂ at any temperature
    2. Carbon would form titanium carbide instead of the metal
    3. Magnesium is cheaper than coke
    4. TiO₂ is not reducible
    Show answer

    Answer: B — Carbon would form titanium carbide instead of the metal

    The Ellingham diagram suggests carbon could reduce titanium oxide at high temperature, but the actual product is a very stable carbide rather than the metal, which the diagram cannot warn you about because it says nothing about which product forms. The Kroll process therefore converts TiO₂ to TiCl₄ and reduces that with magnesium or sodium. Magnesium is in fact considerably more expensive than coke, so cost is not the reason.
  8. Cryolite is added to alumina in the Hall–Héroult process to:

    1. Act as the reducing agent
    2. Lower the melting point and make the melt conducting
    3. Remove iron oxide impurity
    4. Prevent the anode from being consumed
    Show answer

    Answer: B — Lower the melting point and make the melt conducting

    Pure alumina melts above 2000 °C, which is industrially impossible to sustain; dissolving it in molten cryolite with a little fluorspar brings the working temperature down to about 950 °C and gives a well-conducting electrolyte. Iron oxide is removed earlier, during Bayer leaching, and the carbon anode is consumed as CO₂ regardless — which is exactly why aluminium smelting carries such a large carbon footprint.