All metals undergo oxidation and corrosion, as they always want to return to their initial, low-energy state. Generally, this is the oxide form.

    \[Metal + Oxygen + Energy \rightarrow Metal Oxide\]

If the applied energy is positive, the metal is stable. If it is negative, the metal will oxidise.

Aluminium has a large negative oxidation energy, whereas iron has a smaller negative energy – so why does iron oxidise more than aluminium?

Oxidation rate is the measurement of change of mass (as oxygen is added) per unit area over time, and is dictated by a fine balance of oxidation energy and form.

There are two types: parabolic and linear.

Parabolic Oxidation Rates (Dry)

    \[(\Delta m^)^2=k_P t\]

    \[k_P = A_P e^\frac{-Q_P}{\bar R T}\]

The oxide layer is non-porous, meaning that as they layer builds up, it prevents further oxidation. This is the most common form of oxidation.

Parabolic Oxidation Micro-Mechanisms

When the oxide layer sticks to the metal surface, further oxidation depends on ionic diffusion:

The metal releases electrons, forming metal ions.

    \[Metal \rightarrow Metal^{n+} + ne^-\]

These diffuse through the oxide layer, leaving vacancies in the metal and oxidising at the surface.

The released electrons are absorbed by oxygen, forming oxygen ions.

    \[Oxygen + 2e^- \rightarrow Oxygen^{2-}\]

These diffuse through the oxide layer too, oxidising at the metal-oxide boundary.

The parabolic oxidation rate therefore depends on both diffusivity and resistivity of the oxide layer.

The lower the diffusion coefficient and the higher the resistivity, the better the protective layer.

Linear Oxidation Rates (Dry)

    \[\Delta m = k_L t\]

    \[k_P = A_L e^\frac{-Q_L}{\bar R T}\]

The oxide layer is porous, so oxidation can continue through it.

Normally, k_L is positive, but in some metals (tungsten, vanadium, molybdenum etc.) the oxide layer is extremely volatile and evaporates or dissolves immediately. In

this case, k_L <0, and the metals have limited use in high-temperature environments.

Linear Oxidation Micro-Mechanisms

In some metals (like magnesium), the oxide layer occupies a smaller area than the metal. In this instance, cracks form to mitigate the additional stresses.

In other metals, the oxide layer occupies a larger area than the metal. In this instance, the layer peels away from the metal

In both instances, the oxide layer does not provide protection, and so oxidation continues at a more-or-less steady rate – the layer is said to be porous.

The Piling-Bedworth (P.B.) Ratio expresses the volumes of oxide and metal consumption:

    \[P.B. Ratio = \frac{V_{oxide, produced}}{V_{metal, consumed}}=\frac{M_{oxide} \times \rho_{metal}}{n(M_{metal} \times \rho_{oxide})}\]

  • M is the atomic or molecular mass (g/mol)
  • \rho is the density (kg/m³)
  • n is the number of metal atoms in the oxide
P.B. RatioOxidation RateOxide Layer
P.B Ratio ≈ 1ParabolicAdhesive, non-porous
P.B. Ratio << 1LinearMicrocracks
P.B. Ratio >> 1LinearDelaminates

Wet Corrosion of Metals

The above mechanisms only occur in dry, high-temperature environments. When there is water involved, everything changed: at room temperature, steel rusts a million times faster when exposed to moisture than when dry.

This is because there is (more or less) no oxide layer, so the electrons lost in the formation of metal ions react with the water to form hydroxide ions.

    \[O_2 + 2H_2 O +4e^- \rightarrow 4OH^-\]

These react with the metal to form metal hydroxide precipitates (reduction).

    \[Metal^{n+}+nOH^- \rightarrow nMetalOH\]

The corrosion is so much faster because:

  • The metal hydroxide precipitates do not form a protective layer, but flow away.
  • The metal and hydroxide ions diffuse faster in water than in the oxide layer.
  • Conduction of electrons is better in liquids than in oxide layers.

Kinetics of Corrosion

The weight of metal that has uniformly corroded in an aqueous solution, , is given by Faraday’s general equation of chemistry:


  • w is the weight of corroded metal (g)
  • I is the current flow (A)
  • i is the current density (A/m²)
  • A is the area (m²)
  • t is the time (s)
  • M is the atomic mass of the metal (g/mol)
  • n is the number of electrons produced or consumed per metal atom
  • F is Faraday’s constant, 96 500 C mol^{-1}

Types of Wet Corrosion

Crevice Corrosion

When the metal surface is in contact with another object at a point (like dirt, sand, or a rivet), there is a closed environment at this contact area.

Initially, the corrosion is uniform everywhere along the metal surface, but eventually the oxygen in the closed environment is used up and the reduction of oxygen to hydroxide ions cannot continue.

The metal continues to form cations, however: this excess positive charge is balanced out by the reaction with chlorine ions nearby, forming hydrochloric acid.

    \[MCl+H_2 O \rightarrow MOH+HCl\]

The acid acts as an auto-catalyst, making subsequent dissolution of metal faster. Therefore, a crevice forms at the contact area.

This can be avoided by avoiding rivets (using welds instead), cleaning the surface regularly, and using non-absorbing gaskets (using a rubber band as a gasket will lead to crevice corrosion, and is actually used as a method of cutting steel).

Pitting Corrosion

Pitting corrosion initiates at a scratch or surface defect rather than a contact point with another object (crevice corrosion). The mechanism is similar to crevice corrosion, in that there is an oxygen-depleted zone in which hydrochloric acid forms.

It is extremely localised and is one of the most destructive forms of corrosion because it is hard to identify (pits are often very small or covered in corrosion materials) and predict (failure tends to occur suddenly).

To avoid it, surfaces can be polished, or small amounts of other metals can be added to homogenise the composition distribution (e.g. 2% molybdenum in stainless steel).

Galvanic Corrosion

When two different metals are joined, galvanic corrosion will occur on the more reactive metal (the anode).

This can be avoided by using similarly reactive metals, avoiding large contact areas, electrically insulating the metals, or connecting a third, anodic (even more reactive) metal.

Order of reactivity for common metals:

Increasing Reactivity →

Galvanic corrosion can be advantageous: coating steel with zinc will cause the zinc to corrode in place of the steel, but before this, zinc will form an oxide layer (double protection).

Similarly, an anodic metal can be connected to the cathode to corrode instead of the cathode. This is known as sacrificial protection.

Intergranular Corrosion

Intergranular corrosion can cause failure along the grain boundaries of some metal alloys in specific environmental conditions.

For example, in 18-8 stainless steel between around 500 and 800°C, excess carbon becomes insoluble, and reacts with chromium in the solid solution to form chromium carbide near the grain boundaries. This creates chromium depleted zones, which are anodic to their surroundings, so the grain boundaries corrode.

The grains themselves do not corrode.

Intergranular Corrosion can be avoided by:

  • Heating steels to fully dissolve any carbide particles
  • Lowering the carbon content to avoid an excess carbon
  • Alloying the steel so that the excess carbon reacts with something other than chromium.
  • Dry metals will oxidise when there is net negative energy.
  • Parabolic oxidation rate decrease as the oxide layer thickens: (\Delta m)^2=k_P t where k_P=A_P e^\frac{-Q_P}{\bar R T}
  • Linear oxidation is less common, and occurs because the oxide layer is permeable: \Delta m=K_L t where k_L = A_L e^\frac{-Q_L}{\bar R T}
    • P.B.Ratio \ll 1: microcracking
    • P.B.Ratio \gg 1: delamination/spalling
  • Wet corrosion happens a million times faster than dry oxidation in steels: w=\frac{ItM}{nF}
  • Crevice corrosion occurs at rivets, dirt, gaskets etc where hydrochloric acid forms as an auto-catalyser in an oxygen depleted zone
  • Pit corrosion is similar but occurs at a surface defect
  • Galvanic corrosion occurs at the interface of two different metals – the more reactive (anodic) metal will corrode.
  • Intergranular corrosion occurs at the grain boundaries in some alloys