Linking pore diffusivity with macropore structure of zeolite adsorbents. Part II: simulation of pore diffusion and mercury intrusion in stochastically reconstructed zeolite adsorbents

Abstract

In the present study we complete the evaluation of three dimensional digitized reconstructions of a binderless zeolite adsorbent with improved mass transfer rates, by performing simulations of pore diffusion and Hg-intrusion porosimetry in these structures. It is seen that an excellent agreement with the experimental diffusivity is achieved (relative error of 1.2 %) for a pore structure that matches, besides low order correlations, chord length distribution functions that account for higher order correlations. Furthermore, simulations on a variety of reconstructed samples indicate that matching chord length distribution functions is a necessary (though probably not sufficient) condition for accurate structural representation. The average tortuosity factor is 2.68 and is nearly constant over a broad spectrum of pressures, when properly normalized. Hg-intrusion porosimetry simulations, performed with a pure morphology method, show a good agreement with the experimental curve for normalized cumulative intrusion volumes in the range of 50–88 %, but cannot make a distinction between structures with differences in higher order correlations. It is believed that SEM micrographs, properly obtained to represent realistic 2D sections of the material, contain sufficient structural information that can distinguish among pore structures with different mass transfer rates, when combined with stochastic reconstruction methods. Evidently, the direct link between these structural parameters and pore diffusivity will provide the necessary route to improve the mass transfer rate of porous adsorbents.