Binary gas adsorption in molecular sieves
This thesis is concerned with the development of sorption-effect chromatography as a rapid method for the determination of binary gas-mixture adsorption isotherms.There are many alternative non-chromatographic methods but these have inherent disadvantages the direct experimental methods require excessive equilibration times and the predictive methods require the respective pure-component isotherms and an ideal adsorbed phase. A computer simulation has shown that for an alternative chromatographic method, good results will only be obtained if both binary isotherms are close to linear. Sorption-effect chromatography is characterised by the flowrate retention time (TN) which measures the change in column inventory when a perturbation is made to the system. Along with the standard composition retention time (Tx), this extra measurement allows the gradient of each binary isotherm to be evaluated. Subsequent integration will give the respective mixture isotherm. Three gas systems (nitrogen-argon, nitrogen-helium and argon-helium) have been investigated over zeolite 5A at different temperatures. The results confirm that the adsorbed phase amounts decrease, with increasing temperature and that there are degrees of component interaction. Experimentally, thermal fluctuations in the oven will cause noise on the flowrate record making TN determination difficult. Isolation of the column from direct air flow was seen to reduce the noise level. Also, using a computer simulation model, the heat of adsorption for the above zeolite 5A systems will be easily dissipated preventing any unwanted gas temperature rises; the comparatively small column diameter was found to be a significant factor. The employment of delay lines (empty tubes) in various locations has been investigated. To directly determine TN it is necessary to used delay lines downstream of the column. Also, the chromatographic method has been extended to determine mixture isotherms by considering the change in average column pressure rather than the motion of a composition front through the column. Delay lines situated upstream of the column are able to separate these two effects, and preliminary results are satisfactory. However, the use of delay lines anywhere in the system changes the measured retention times and the theory has to be adjusted to account for this.