It is my understanding that iron oxidation in stainless steel is prevented by adding chromium, which creates chromium oxides at the surface that shield the bulk of the material from further oxidation.

I have some stainless steel samples that I heated up to 400°C, and they came out of the oven with signs of oxidation. How can I determine at which temperate will iron start oxidation despite this protective chromium oxide layer? Note that I do not have enough samples to conduct an experiment to find the temperature, I'm looking for a theoretical approach.

  • $\begingroup$ Is that microwave oven $\endgroup$
    – reddot
    Dec 3, 2018 at 15:21
  • $\begingroup$ No, it is an electric resistance furnace. $\endgroup$ Dec 3, 2018 at 15:38
  • $\begingroup$ did the oxidation appear as a smooth transparent rainbow or dark blue or orange film, or was it rough and brown? $\endgroup$ Dec 3, 2018 at 16:17
  • $\begingroup$ @nielsnielsen It has a dark brownish color. $\endgroup$ Dec 3, 2018 at 16:20
  • $\begingroup$ was it completely smooth and tightly adherent or was it rough and flaky? $\endgroup$ Dec 3, 2018 at 16:22

2 Answers 2


There is no real, discrete temperature at which an oxidation reaction (or any other chemical reaction for that matter) starts.

Instead reaction rates are dependent on absolute temperature and follow the Arrhenius equation.

In short this means that at low temperatures (e.g. close to room temperature) reaction rates are usually (but not always) almost infinitesimally small and then increase as temperature is increased. There is however no discrete temperature at which the reaction can be said to 'start'.

To get a better idea of this rate/temperature dependence, chemists will usually perform an Arrhenius plot.

  • $\begingroup$ Thanks for the comment, do you happen to know any literature about Arrhenius plot for stainless steels? One of my colleges suggested me to look at Ellingham diagrams, does this make sense to you? $\endgroup$ Dec 3, 2018 at 16:28
  • $\begingroup$ Ellingham diagrams serve a different (but related) purpose. And Arrhenius plots assume the reaction proceeds throughout the reagents, not just on the surface. I would look for handbooks on corrosion of building materials such as SSs. These will provide practical guidelines for the use of SS at high Ts, w/o going into theory, which is difficult to apply here. $\endgroup$
    – Gert
    Dec 3, 2018 at 17:33
  • 1
    $\begingroup$ @lavirtuosacorcholatavoladora Since the oxidation of Cr always lies below the oxidation of Fe (up to the melting point) in the Ellingham diagram, which is essentially a plot of negative spontaneity vs. temperature, we know that elemental Cr in the material will always tend to oxidize (and also reduce oxidized Fe) before the elemental Fe oxidizes. $\endgroup$ Dec 3, 2018 at 21:12

Stainless steels are indeed protected by an extremely tightly-adherent, conformal coating of chromium oxide which forms the instant that the part is first exposed to air. This coating is optically transparent because of its thinness- only a few tens of atoms thick- but it protects the material underneath it because the rate at which oxygen can diffuse through it is extremely slow.

When it is heated as in your furnace, the diffusion rate is increased and so the creation of chrome oxide is "turned on" again until such time as the thickened oxide begins to choke off the diffusion of oxygen again, and the thickness of the oxide stabilizes.

This temperature-dependent, diffusion-limited and self-limiting pattern of behavior is called self-passivation and greatly complicates the dynamics of high-temperature corrosion.

As the oxide layer thickens, it begins to show optical interference effects between the front and back surfaces of the layer- and instead of appearing perfectly colorless, the film gets colored- first a light yellow which darkens into orange and then a brilliant blue as the film thickens. This color pattern then fades out and repeats; by counting these "fringes" you can accurately estimate the total thickness of the oxide.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.