Correlation and prediction of liquid-phase adsorption on zeolites using group contributions based on adsorbate-solid solution theory

Citation
C. Berti et al., Correlation and prediction of liquid-phase adsorption on zeolites using group contributions based on adsorbate-solid solution theory, LANGMUIR, 15(18), 1999, pp. 6035-6042
Citations number
25
Categorie Soggetti
Physical Chemistry/Chemical Physics
Journal title
LANGMUIR
ISSN journal
07437463 → ACNP
Volume
15
Issue
18
Year of publication
1999
Pages
6035 - 6042
Database
ISI
SICI code
0743-7463(19990831)15:18<6035:CAPOLA>2.0.ZU;2-4
Abstract
Both correlation as well as prediction of experimental data for the adsorpt ion of various binary liquid mixtures of alkanes and alkenes on NaX at diff erent temperatures are presented. The theoretical background is based on th e adsorbate-solid solution theory which conceives the adsorbed phase to be a mixture of the adsorbed species (adsorbate) and the adsorbent as an addit ional component. With the introduction of the Gibbs excess energy G(E)* for this hypothetical mixture, activity coefficients and composition of the ad sorbed phase may be calculated. The Gibbs excess energy and thus the activi ty coefficients of the adsorbed species depend strongly on the energetic he terogeneity of the solid surface which may be described by use of so-called group contribution models. These approaches, until now widely applied to p redict fluid-phase equilibrium, are derived from statistical thermodynamics and take into account the energetic interactions between the respective co mponents. For the application of this approach on thermodynamics of adsorpt ion zeolites have to be divided into different functional groups such as Si O2, AlO2-, and the respective cations. The interaction energies between the se active sites and the functional groups of the adsorbed liquid molecules represent fundamental parameters of activity coefficient models based on gr oup contributions such as UNIFAC. These parameters were determined by fitti ng four different adsorption systems. With the fitted values, six other sys tems were predicted. Both correlation and prediction include adsorption dat a at different temperatures. All calculations show excellent results with a mean relative deviation of 4.2% for the correlation and a mean deviation i n the range of 8-17% for the predictions.