Gas permeation through carbon membranes: Model development and experimental validation

Abstract

With a growing interest in carbon membranes for gas separation, understanding their performance and behaviour is essential for proper design of the membrane separation. Currently not many models exist that correctly describe transport through carbon membranes due to its complex nature. This work attempts to implement a general modelling approach which describes several key transport phenomena inside carbon membranes. The approach assumes a membrane wall to be a bundle of pores with parallel transport mechanisms using the pore size distribution as a weight factor to sum the different transport phenomena. This work adapts this approach specifically for carbon membranes, additionally accounting for molecular sieving and pore blocking effects. Imposing realistic boundary conditions, the model is solved using global optimization algorithms. For testing, four different CMSMs have been produced with hydroquinone and novolac precursors. Pure- and mixed gas permeation tests are done for these CMSMs with H2, N2, and CO2 and the model is fit to this permeation data. Fitting results with pure gas measurements show the model is able to predict the contributions of different mass transport mechanism for the different membranes. This is validated by comparing these results to gas-pair permselectivity data. The model is furthermore fit to mixed gas data. Existence of multi-component effects shows that the model could be further improved. Overall, the model presented in this work is shown to be able to describe complex mass transport behaviour for various different carbon membranes.