The smelter: which reducing agent works?
Pair ores with coke, carbon monoxide, hydrogen, aluminium powder and magnesium in a virtual smelter, set the furnace temperature, and let Ellingham thermodynamics decide whether metal comes out.
Aim Find which reducing agents reduce each oxide, and above which temperature; explain why carbon can smelt iron and zinc but not aluminium.
Procedure
- Start with FeO and coke at 800 K. Note the sign of ΔG°, then raise the temperature 100 K at a time until ΔG° changes sign.
- Repeat for ZnO: above 1180 K the zinc comes off as vapour. Switch “Include metal melting/boiling” on and off and compare the crossover temperatures.
- Switch to Fe₂O₃ with carbon monoxide (the blast furnace’s real reducing agent) and check ΔG° between 600 and 1800 K.
- Try every reducing agent on Cr₂O₃, TiO₂ and Al₂O₃, and log the results.
- Compare with the Ellingham diagram and explain your log by which line lies below which.
What to notice
- Carbon (making CO) reduces FeO above about 1020 K and ZnO above about 1220–1260 K; for Al₂O₃ and MgO, ΔG° stays far above zero at ordinary furnace temperatures.
- CO reduces Fe₂O₃ at every temperature, but for FeO + CO the ΔG° is slightly positive at high temperature — it needs a gas that is mostly CO.
- Aluminium reduces Cr₂O₃ (thermite), and magnesium reduces TiO₂: their lines lie below those oxides.
- The furnace glow runs from dull red through orange-yellow to near white — black-body colour from Planck’s law, which is how smiths judged temperature by eye.
In a real lab
- Real smelting is also about rate: reactions between solids are extremely slow below about 700 K even when ΔG° < 0.
- Carbon forms carbides with many metals (TiC, Al₄C₃), so the product is not the pure metal; titanium is made industrially by chlorination and magnesium reduction (the Kroll process).
- Thermite exceeds 2000 °C and throws out molten metal, so it is demonstrated only under professional conditions; carbon monoxide is highly toxic.
- The straight Ellingham lines ignore heat capacities and oxide melting, so crossover temperatures can be off by tens of kelvin.