Metals want to be ores again
Here is the uncomfortable truth behind every shiny object: most metals are not stable in air. Iron, aluminium, zinc, copper — left to themselves with oxygen and water around, they slide back toward the oxides and sulfides we dug them out of. A lump of native gold can survive in a riverbed for a million years precisely because gold is the rare exception, sitting at the unreactive bottom of the [[electrochemical-series|electrochemical series]]. Everything more reactive than gold is, in the long run, fighting a losing battle against its own chemistry. Extraction and corrosion are really the same story told in opposite directions: extraction is us forcing electrons onto a metal ion to make the metal, corrosion is nature pulling those electrons back off.
You already have the tools to see this clearly. From the earlier guides in this rung, you know that [[metal-extraction|reducing a metal ion]] means handing it electrons — Fe2O3 has iron in the +3 state, and to get iron metal we must drop that oxidation state from +3 to 0. The whole question of how to extract a metal collapses into one practical worry: who is willing to be the [[oxidizing-and-reducing-agents|reducing agent]] — the thing that hands over those electrons — and how much will it cost in heat, energy, or money? The Ellingham diagram is the map that answers exactly that.
The Ellingham diagram: a race of oxides
The [[ellingham-diagram|Ellingham diagram]] plots, for the formation of each metal oxide, the free energy change delta-G against temperature. Crucially, every line is drawn per mole of O2 consumed, so the diagram puts every metal on the same footing — they are all competing for the same oxygen. The lower a metal's line sits, the more negative its delta-G, the more stubbornly that metal clings to oxygen. And here is the engine of the whole diagram: if you want metal A to reduce the oxide of metal B (stealing B's oxygen), A's line must lie below B's line at the working temperature. The element with the more negative delta-G wins the oxygen.
Why do the lines tilt? Almost every metal-oxidation line slopes gently upward — its delta-G grows less negative as temperature rises — because a gas (O2) is being consumed, so the reaction loses entropy, and that minus-T-delta-S penalty bites harder when T is large. Now look at the one line that does the opposite: the reaction 2C + O2 -> 2CO. Here you turn one mole of O2 gas into two moles of CO gas, so entropy increases, and the carbon line slopes downward. That single downward line is why carbon is the great industrial reducer: as you heat the furnace, carbon's line eventually dives beneath the lines of iron, zinc, and many others, and from that crossover temperature onward carbon will rip the oxygen away.
delta-G / per mole O2 (more negative = lower = clings to O harder)
0 +------------------------------------------------
| Ag2O (high up: easy to reduce, even mild heat)
| Cu2O
-200| 2C + O2 -> 2CO (slopes DOWN \)
| 2Fe... -> Fe oxide \
| ZnO \ carbon line dives
-500| ... \ below Fe, Zn at high T
| Al2O3 (low down: very hard to reduce)
| MgO
-1100| CaO (lowest: C cannot reach -> use electrolysis)
+------------------------------------------------
low T --------- temperature ---------> high TReading the diagram: carbon, hydrogen, or electrolysis?
The practical payoff is that one glance at the diagram tells you which reducing agent to use. For iron, carbon's line crosses below iron's around 700-1000 degrees C — exactly the blast-furnace regime, where coke and the CO it makes pull oxygen out of Fe2O3. For the very stubborn oxides at the bottom — Al2O3, MgO, CaO — carbon's line never gets low enough at any practical temperature, or only at temperatures so extreme they would form unwanted carbides. Those metals refuse to be reduced by heat-and-carbon, so we hand them electrons directly with electricity: aluminium is won by electrolysis of molten Al2O3 dissolved in cryolite, an enormously energy-hungry process. That single fact — that aluminium sits below carbon's reach — is why recycling a drink can saves about 95 percent of the energy of making one from ore.
Where does hydrogen fit? Hydrogen's oxidation line (2H2 + O2 -> 2H2O) sits at an intermediate level, so hydrogen can cleanly reduce moderately easy oxides — tungsten and molybdenum, where you need a pure metal free of carbon contamination, are won this way. It is the reducing agent you choose when purity matters more than cost. So the diagram quietly sorts the metals into three bands: high oxides reduced by gentle heat or even carbon at low temperature, middle oxides by carbon (cheap) or hydrogen (clean), and the deep oxides reachable only by electrolysis.
Rust: a tiny battery on every iron surface
Now flip the story. [[corrosion-and-passivation|Corrosion]] is extraction running backward and for free — the environment acts as the oxidizing agent that nature could not afford to lose. The beautiful, slightly surprising fact about the rusting of iron is that it is not a simple smear of oxygen onto metal. It is electrochemistry: a damp piece of iron behaves like a swarm of microscopic galvanic cells, with anode regions and cathode regions wired together through the metal itself and through a film of water acting as the electrolyte. Where iron is strained, dirty, or oxygen-poor, it becomes the anode and dissolves; where oxygen is plentiful, that spot becomes the cathode. This is why rust so often blooms in a crevice or under a paint chip rather than out in the open air.
- At an anodic patch, iron gives up electrons and dissolves: Fe -> Fe2+ + 2 electrons. This is the oxidation half, and it is why pitting starts here.
- The electrons travel through the metal to an oxygen-rich cathodic patch, where oxygen and water are reduced: O2 + 2 H2O + 4 electrons -> 4 OH-. Note the cathode reaction consumes the oxygen, which is why oxygen access controls the rate.
- Fe2+ ions drift out into the water film and meet the OH- and more oxygen, oxidizing further to Fe3+ and precipitating as hydrated iron(III) oxide — the flaky, reddish-brown rust, roughly Fe2O3 times n H2O.
The reason rust is so destructive — far worse than the tarnish on aluminium — is that this hydrated oxide is flaky and porous. It does not stick; it spalls off and exposes fresh metal, so the corrosion keeps eating inward. You can even read the whole process off the iron part of a Pourbaix diagram from the earlier guide: it maps, against pH and potential, where iron stays as the bare metal (immune), where it dissolves as Fe2+ (active corrosion), and where it sits under a protective oxide (passive). Rust prevention is, in essence, the art of keeping iron out of that middle active region.
Three ways to win the redox fight
Once you see corrosion as a galvanic cell, the cures organize themselves. The first and most obvious is the barrier coating: paint, grease, plastic, or electroplated tin keeps water and oxygen off the metal so no cell can form. The catch is brutal — a barrier only works while it is whole. Scratch a tin-plated steel can and the tin actually makes things worse, because tin is less reactive than iron, so the iron becomes the anode and corrodes faster at the gap. The protection is only as good as its weakest scratch.
The second cure turns that same logic into a weapon: the sacrificial anode. Bolt a block of a more reactive metal — zinc or magnesium — onto the iron, and you deliberately make the iron the cathode. Now the electrochemical series works for you: zinc has a more negative standard reduction potential, so zinc gives up its electrons first and corrodes in the iron's place. This is exactly why galvanized (zinc-coated) steel survives a scratch — even with bare iron exposed, the surrounding zinc keeps feeding it protective electrons. Ship hulls, pipelines, and water heaters all carry blocks of zinc or magnesium that are designed to be eaten and periodically replaced.
The third cure is the most elegant, and it is why aluminium does not crumble in air the way the Ellingham diagram might frighten you into expecting. Passivation is corrosion that protects: aluminium and chromium are so eager to oxidize that they instantly grow a thin, dense, tightly-bonded oxide skin (Al2O3, or Cr2O3 on stainless steel) that seals the surface and stops oxygen from reaching the metal beneath. The difference from rust is entirely about the film's character — adherent and pore-free versus flaky and porous. This is also the deep reason stainless steel is stainless: enough chromium in the alloy to grow a self-healing chromium-oxide passive layer. Anodizing aluminium is just this same trick, done deliberately and thickened by electrolysis.
One arrow, two industries
Step back and the unity is striking. The drive to lower energy that the Ellingham diagram charts as delta-G against temperature is the very same drive that the electrochemical series charts as reduction potentials in solution — thermodynamics speaking with two accents. Extraction is humanity spending enormous energy (a third of it, for aluminium, just to run the electrolysis) to push metals uphill, against their nature, into the useful elemental form. Corrosion is the same metals quietly rolling back down. Every anti-corrosion strategy is a clever way to slow that descent, and the bill for losing — replacing rusted bridges, pipes, and ships — runs into the trillions worldwide each year. Redox bookkeeping, the humble business of counting electrons, turns out to underwrite a staggering slice of the physical economy.
This guide closes the Redox rung. You began with oxidation states as a bookkeeping device, learned to balance half-reactions and read potentials, and saw descriptive redox chemistry compressed into the elegant pictures of Latimer, Frost, and Pourbaix diagrams. Now you have watched those same ideas walk out of the textbook and into a blast furnace and onto a ship's hull. Electron transfer is not an abstraction — it is how we mine the modern world and how the world slowly takes it back.