Now it becomes a clearer situation, at least for me.
[•] Coming back to the subject of nitrogen-inflated tires. It is not a diffusion process as claimed by the nitrogen supporters, it is permeation, as highlighted by kenvlach, a totally different concept. Permeation refers almost exclusively to polymers and is caused by differences in chemical potential. Example: loss of carbon dioxide from plastic bottles for carbonated soft drinks.
[•] Permeation is characterized by a coefficient P, which is the product of two seemingly counteracting transport effects: the diffusion coefficient D through the pores (favouring nitrogen), and the solubility (molecular transport) coefficient S of the permeant gas in the matrix (adding high mobility to oxygen).
[•] The process of permeation through a polymeric barrier is said to consist of four steps: absorption of the gas into the polymer wall; solubility in the matrix; movement by diffusion along a concentration gradient; and desorption from the outer wall.
[•] It appears that the polarity of the polymer functional groups, chain symmetry, attraction between polymer chains, glass transition temperatures, and crystallinity all affect the permeation rates. Example: oxygen permeates 10 million times faster through poly(dimethylsiloxane) than through poly(acrylonitrile).
[•] Permeability is also influenced by temperature, differential pressures, humidity, fillers, additives, thickness, etc. Some of them, by increasing the pathway and its tortuosity, decrease the permeation of a gas molecule.
[•] Gases, in general, have lower permeation coefficients in thermoplastics than in elastomers because segmental movements of the former are frozen-in below the transition temperatures.