Permeation with a double mode of sorption.
It may be the case of solvent molecules at low vapor pressures. At the beginning of the kinetics, most the molecules cross the membrane without being trapped and the strong interaction sites become to be more and more occupied very slowly. The permeation kinetics curve likes roughly a fickian one if time is not too long. The time-lag seams to be shorter than in the previous case but its measurement is difficult and one should wait for a long time before the stationary regime is completly reached to fit the asymptote. Anyway, the estimation of the real diffusion coefficient from the time-lag by Fick law will be wrong and one should obtain only a pseudo diffusion coefficient using the pseudo stationary part of the kinetics. The concentration profile rates slowly to final stationary one after a long time and then should be finaly the same as in example 3.
Example 4, case of strong interaction sites and a low kinetics of trapping :