Predicting the structural parameters of integrally skinned porous membranes

Abstract

The objective of this study is to propose a novel approach for predicting the structural parameters of integrally skinned porous membranes using gas permeation data. It is intended to overcome the limitation of the conventional gas permeation testing (GPT) method, since the latter method seems to suffer from several conceptual drawbacks. In particular, a comparison is made between the theoretical calculation and the experimental data to show the superiority of the newly proposed model. The new model is a modification of Wakao et al.'s model, in which, unlike the conventional GPT method, the contribution of the slip flow is considered. Although Wakao et al.'s model was found superior to the conventional GPT method, the model fitting to the experimental data was not completely satisfactory. It was likely that the slip flow, although its effect cannot be neglected, is not fully developed. Therefore, a factor ψ is introduced to show the extent of the contribution of the slip flow mechanism to the total gas permeation rate. As a result, the new method can overcome the shortcomings of the conventional GPT method by manifesting the following advantages: (i) it can cover the entire range of J versus P diagram, (ii) it can specify the contribution of the individual mechanisms involved in the total gas permeation and (iii) unlike the conventional GPT method, it is not limited by any constraints or conditions of data acquisition. In summary, the model can predict pore size and effective porosity, and also simulate the experimental J versus P trends with sufficient accuracy (within ~2% over the pressure range studied) for all types of membranes, i.e. NF, UF, MF, MD and membrane contactors. In view of this fact, the proposed model is simpler to apply than Rangarajan et al.'s model and more accurate than the conventional GPT method.