(407d) Generalized Perturbation Method to Determine Multicomponent Isotherm for Saturated Systems through a Modeling-Design Concurrent Approach

Determining adsorption isotherms for multi-component systems requires substantial experimental effort. Single component adsorption as well as binary/ternary interactions need to be studied at many different compositions. An example is xylenes isomers on molecular sieve. Industrial xylene adsorption process interests in the competitive adsorption of five components (p-xylene, m-xylene, o-xylene, ethyl benzene and para-diethyl benzene) at saturation. The adsorption properties are usually determined by studying single component adsorption experiments and fitting the data to competitive adsorption models. These adsorption can be characterized by empirical models [1] or by Ideal Adsorbed Solution Theory (IAST) [2]. When the number of components is large, a number of experiments must be performed at various compositions, which requires substantial effort.

The perturbation method is an experimental technique that an equilibrium of an adsorbent with the liquid phase is perturbed by a small injection of a sample with different concentration [3]. This method, where only the retention time of the sample is measured, can reduce analysis effort for multi-component systems. While this method has been reported for a dilute system which consists of two solutes, applications for a larger number of components, or for saturated systems where interactions with the mobile phase cannot be ignored, have not been reported.

We extended the perturbation method to study concentrated (non-dilute) system to model the adsorption isotherm near or at the saturation. Furthermore, we also adopted a concurrent approach, where isotherm modeling and process design are carried out iteratively. In this proposed approach, a process model with preliminary isotherm parameters is used to design an adsorption process, and this model also predicts some representative compositions in the process. At these predicted compositions, isotherm measurement tests are carried out. The isotherm determined from the experimental data refines the process model, which leads to better prediction of representative compositions. These two steps, isotherm determination and process design, are carried out concurrently and iteratively.

We present a simulated moving bed (SMB) process for xylene isomer separation as an example. A rigorous dynamic model for the SMB process is implemented and optimized by a deterministic method. The internal concentration profiles are used to determine the compositions where experiments for isotherm determination is carried out. By this iterative approach, experimental effort is reduced to minimal, and model reliability can be increased.

1. Tournier, H., et al., Adsorption equilibrium of xylene isomers and p-diethylbenzene on a prehydrated BaX zeolite. Industrial & Engineering Chemistry Research, 2001. 40(25): p. 5983-5990.

2. Rota, R., et al., Generalized Statistical-Model for Multicomponent Adsorption Equilibria on Zeolites. Industrial & Engineering Chemistry Research, 1988. 27(5): p. 848-851.

3. Heuer, C., et al., Design of the simulated moving bed process based on adsorption isotherm measurements using a perturbation method. Journal of chromatography A, 1998. 827(2): p. 175-191.