“…In short, the number of constraints is smaller than number of unknowns as already noted in ref. [19] for 0 → p a step change, while this is not the case for symmetric membrane where the unknown parameters can be determined as demonstrated by Grant et al [1,2,7]. To avoid this obstacle in the case of asymmetric membrane, the value of solubility is most often taken from the literature despite the considerable scatter and unreliability of K S values at low hydrogen pressures that can introduce a significant error in the subsequent determination of K L,1 and γ.…”