The subject of this dissertation is the adsorption of methane and ethane on a 5A molecular sieve. An experimental method, based on the fixed-bed flow-cell, was developed covering a wide range of adsorbate partial pressures in a continuous sequence. A Laporte type 4A molecular sieve failed to adsorb methane or ethane. Single component isotherms were established for a 5A molecular sieve and an activated carbon at 20°C. Binary isotherms were obtained for the 5A molecular sieve, at high fractional saturation, at 20°C. The results obtained for single components agreed well with an established mathematical model. Single component isotherms on the activated carbon were of the same order as have been found by other workers. All the single component isotherms could be expressed in terms of an empirical Langmuir model. None of the models used to describe the binary adsorption data on the 5A molecular sieve, from the single component model coefficients, gave satisfactory results. An empirical method was employed to model the binary adsorption data using single component empirical Langmuir models to express the isotherm of each component in each mixture. The Langmuir model coefficients were then expressed as a function of the gas phase mole fraction of the interfering species. Adsorbent particle macro-pore spectra and volume were determined using mercury porosimetry. Isothermal fixed bed experiments were carried out, and data from the isotherm experiments were used, to obtain breakthrough data for single and two component mixtures over a range of component relative mole fractions, total adsorbate mole fraction of the total flow, adsorbate concentration and flow rate. Mathematical models, using finite difference techniques, were used to model the fixed bed breakthrough data. In the instances where the total adsorbate concentration was a significant proportion of the total gas phase concentration a constant flow rate through the bed could not be assumed. A mathematical method was developed whereby a bed axial flow profile was calculated and used in the solution of the fixed bed flow equations with excellent results. Stability criteria were derived for equilibrium and linear lumped parameter kinetic models which gave more accurate predictions of the finite difference step lengths and also explained instability in the equilibrium control model at what were previously thought to be suitable step lengths. An equilibrium control model was found to give good results over the whole range of experimental conditions. Isothermal fixed bed experiments were carried out to determine the desorption kinetics of methane-ethane mixtures. A linear lumped parameter model was used to describe the desorption of both components. A series of computer simulations were run, using the results from these experiments to examine the feasibility of using the desorption step to effect a further separation, especially when the initial gas feed contained a small fraction of ethane.
|Date of Award||1978|