Pros and cons of three approaches to the study of diffusion in zeolites: cellular automata, Networks and second-order Markov models

The problem of diffusion in zeolites is addressed form different theoretical perspectives, aiming at building coarse-grained models on the general background of Energy Landscapes theory.A Lattice Gas Cellular Automaton model is applied to the study of mixtures of gas adsorbed in zeolite ITQ-29. A more reductionistic approach, the Central Cell Model, is also presented, with a detailed study of the associated Displacement Autocorrelation Function.The PES of argon, xenon, and methane has been investigated, and represented by means of Disconnectivity Graphs, revealing low kinetic barriers for intra-cage motion at room temperature.The inner space of zeolite pores was coarse-grained, by clustering Molecular Dynamics trajectories, calculating the transition matrix for microstates, and applying Perron Cluster Cluster Analysis, in order to find metastable macrostates. This allowed the fundamental discrete events for diffusing molecules to be defined. On the basis of this events classification, a second-order Markov Model of dynamics was developed, and applied to various adsorbate species in zeolite ITQ-29.Finally, the weighted digraph corresponding to the event transition matrix was analyzed, applying concepts and techniques of Networks Theory.