All organic materials become very unstable at high temperatures. They will crack, coke, polymerize, especially on hot solid surfaces and on some more so than on others. They can also react with the surfaces. Some of these pyrolysis or reaction products can be good solid lubricants. They don't last long, but then again under friction and wear new active surfaces and more lubricants can be formed. This is a concept of solid lubricant regeneration.
Our work has proved experimentally that this concept has merit, perhaps as a result of partial graphitization, under selected conditions. In particular, in an environmental chamber, on heated nickel or nickel alloy and palladium surfaces in inert atmospheres, friction and wear coefficients were found to drop by an order of magnitude or more when as little as 1% of ethylene gas was introduced. Diffusion of elemental carbon through the metal lattice appears to be the rate-controlling step in the process.