# How fast a surface oxide layer is formed on bulk chromium?

If a chromium piece is cut in half, how long would it take for an oxide layer to form on the new surface (room temperature)? i.e. is it a matter of seconds/minutes/hours/days/weeks?

The closest study I found is by G. Thurner and P.H. Holloway (1992), but I can't really tell from their results what would happen at ambient conditions:

(I found several publications about high temperature oxidation, but none for surface oxidation at ambient conditions)

First, a definition of the Langmuir unit, used in the graph in the question, which is not often run across except by vacuum (particularly UHV) folks. One Langmuir is the exposure of a surface to a gas pressure of $10^{-6}$ Torr for one second. It roughly corresponds to one monolayer equivalent of gas hitting the surface. At normal atmospheric pressure, with 20% oxygen in air, 1 Langmuir of oxygen hits the surface in about 6.6 nanoseconds (yes, nanoseconds).

Another paper that could be of interest is C. Palacio et al., "AES, XPS and EELS study of the initial oxidation of polycrystalline chromium", Surface Science 182(1-2) 41-55 (1987). They see chemisorption of oxygen in the first 1-2 L, nucleation and lateral growth of oxide islands up to 10 L, and continued growth of an oxide film above 10 L.

Overall, there is not a huge difference between the two papers. By ~10 L of exposure to oxygen there is a continuous oxide on the surface at 300K. Following from the numbers in the first paragraph, that occurs in less than 100 nanoseconds. Basically, it oxidizes as it is cut.

• Great expansion! How did you get 6.6 nanoseconds? I'm getting 0.16 nanoseconds – Sparkler Mar 29 '17 at 23:24
• I went with 760 Torr times $10^6$ times 20%, getting 152,000,000 L in one second (for oxygen). Then, one L happens in (1/152000000) of a second. – Jon Custer Mar 29 '17 at 23:28
• ohhh it's pressure multiplied by time. makes sense now. – Sparkler Mar 29 '17 at 23:38
• Right. So 1 Langmuir could be a microTorr for 1 second or a nanoTorr for 1000 seconds. – Jon Custer Mar 29 '17 at 23:58