In Reply to: Diffusion of H2 Gas through Epoxy Block posted by Kyle C. Smith on October 31, 2006 at 20:46:43:

Dear Kyle,

Further to your questions and background information, we would like to inform you on the following:

In case of a steel containment that holds the hydrogen, the hydrogen will diffuse into the welds of the containment. In there, the hydrogen will split in 2 x H+ and 2 electrons. Because the diffusion rate of H+ in this material is very high, the welds will suffer severely by hydrogen initiated cracks, increasing the corrosive degradation of the system.

Lining of such a system with a cast epoxy resin may be a good idea. Hydrogen will diffuse as one molecule into this liner with a diffusion rate that is lower than H+ in the case described above. This will generate an essential lag time for the start of the described corrosion rate (in presence of water and oxygen).

Furthermore, the chemical activity of the hydrogen on the epoxy – metal interface might be lower. However this depends on the rate of the reaction and tightness of the epoxy - metal interface. If the hydrogen decomposition rate is fast compared to diffusion rate, you may assume that the chemical activity / chemical potential becomes lower due to the lining.

The ultimate effect depends on the type of epoxy resin, the chosen thickness, and the interface between the epoxy resin and the metal.

Glass reinforcement does, from our point of view, not offer benefits from a diffusion and chemical resistance point of view in this specific case. However, in some cases these reinforcements may be required in this specific case, for mechanical and composite-steel interface reasons. Please contact us for more information, also with regard to further interface facilitation.

Hydrogen – epoxy diffusion measurements are definitely required as the type of diffusion behaviour, the solubility and swelling of epoxy under the load of hydrogen at 300 bar and 120 degrees Celsius has to be determined. We would advice you not to take shortcuts by extrapolation of permeation figures form handbooks or electronic sources, because it is expected that these extrapolated permeation figures are totally wrong in the conditions stated. Contact us for more experimental information on cast or composite solid materials.

Regards,

Composite Analytica