Modeling of zeolite membranes from atomistic principles

Membranes made of zeolites offer many potential advantages for the separation of gas and liquid mixtures. Hundreds of different zeolite structures are known but despite progress in synthesizing zeolite membranes and studying their properties, reproducible membranes can be synthesized and used only for a handful of different zeolite structures. In this thesis we present a general methodology for predicting the macroscopic flux of multicomponent mixtures through zeolite membranes with information obtained from detailed atomistic simulations. The feasibility of such atomistic simulations is based on the known detailed atomic level structure of zeolites. We first apply this methodology to predict the permeances of methane and perfluoromethane as pure gases through silicalite membranes and compare quantitative with experimental measurements. Having established a quantitative agreement between the experiments and the calculations we proceed by applying our methodology to predict the transport properties of a methane-perfluoromethane mixture through a silicalite membrane. A comparison between the calculated and the experimentally measured selectivity of the silicalite membrane for the separation of this mixture is presented and discussed. The transport properties of various light gas species through silicalite and other silica zeolites are calculated from atomistic simulations. It is shown that the transport properties of light gas species vary dramatically from system to system. A brief discussion for the differences observed between the gas species for transport in silicalite is given. We then extend our atomistic simulations to study the transport properties of methane and hydrogen in carbon nanotubes. The diffusion rates observed in nanotubes are found to be be 3–5 orders of magnitude higher than zeolites. Finally a comparison between the fluxes of methane observed in a model nanotube membrane and silica zeolites is presented.