The oxidation of carbon-carbon composite is characterized by several intermediate reactions. Some oxygen atoms are not immediately oxidized and remain bonded to the carbon furface to anneal the defects caused by the removal of carbon atoms by the oxygen. Carbon monoxide and dioxide are then formed via a sequence of reactions involving mobile surface oxygen complexes at active sites. According to Bacos, the oxidation of carbon unfolds according to the following steps:
- Oxygen diffusion through the boundary layer.
- Oxygen diffusion through cracks and pores of the carbonaceous surface.
- Chemical reaction between carbon and oxygen and formation of oxidation products.
- Gaseous product diffusion through the cracks and pores of the carbonaceous surface.
- Gaseous diffusion of the products through the boundary layers.
Since it involves gaseous products and reactants, the oxidation reaction rate is expected to be influenced by both pressure and temperature conditions. Different levels of temperature and pressure can determine which of the five steps presented above becomes rate-limiting. Using a series of micrographs, Bacos has shown that at low temperatures, the slowest and thus rate-determining step is the chemical reaction between carbon and oxygen, while the fastest step is the oxygen and product diffusion through the boundary layer. However, the temperature where the change in the mechanism took place is not given in that investigation. In fact, no kinetic data were presented either and these deductions on the rate-controlling mechanisms were based entirely on observations using sectioned samples. At low temperatures; (1) transverse cracks and fiber-matrix de-bonding both of which contributed mainly to mass transport and (2) pores within fiber bundles affecting mainly the chemical reaction. The oxygen molecules are able to penetrate into the pores of the material and they facilitate the formation of crevices within the bulk. These crevices then become longitudinal channels reducing the bulk to a highly porous skeleton where effectively only the fiber reinforcement remains. This observation implies that at low oxidation temperatures there is little or no shrinkage of the composite until the latter parts of oxidation. At high temperatures, the slowest step has been reported to be the oxygen diffusion through the boundary layer and the fastest is the chemical reaction. Additional support for this observation has been provided by Yasuda et al. who reported that the rate-controlling step changes from chemical reaction-control to diffusion-control of the gaseous species through the boundary layer at higher temperatures. Kinetic data presented by these authors showed that as the temperature of oxidation increased, the relationship between weight loss with time changed from linear to parabolic at temperature between 662C and 770C. This indicated that the rate-limiting step changed from chemical reaction control to diffusion-control of oxygen through the boundary layer at the surface of the composites at these temperatures.
Micrographs presented by Bacos show that the oxidation process at higher temperatures affected only the surface of the material. This implies that the oxidation process at high temperatures results in shrinkage of the composite due to carbon loss from the surface. In this case the oxidation triggers the formation of cracks in the matrix as well as bundles/matrix and bundle/bundle de-bonding followed by cracks within orthogonal bundles. The precise means by which the C-C composite degradation unfolds at low or high temperatures depends also on the reactivity of the matrix and the fibers. However, this can vary considerable in accordance to the manufacturing process employed to produce the C-C composite. In their investigation, Yasuda et al. reported that the matrix oxidation rate occurred at a higher rate than the fibers.