A predictive model of organic acids separation by chromatography with strong anionic resins in sulfate form.

ABSTRACT

Organic acids commonly have quite symmetrical chromatography profiles at low pH (< 1.5) with strong anionic resins, but a significant tailing can be observed with succinic and citric acids. Classical adsorption models, like the Langmuir model, fail to predict this behavior, which can have a major influence on mean retention times and profile shapes, therefore on chromatography performances. A new retention model was developed to better predict organic acid separation with strong anionic resin. This model combines a refined Langmuir adsorption model and an ion-exchange model. Organic acid adsorption is assumed to be due to hydrogen bonding with sulfate and hydrogen sulfate counter-anions on the resin. The adsorption capacity depends mostly on molecular size up to sixteen formic acid molecules could be adsorbed per counter-anions, meanwhile only two succinic acid or one citric acid molecules could be adsorbed. This adsorption model was then embedded in a generic and accurate modeling approach (continuous column with mass balance equations solved by the conservation element/solution element (CE/SE) method). All parameters of this column model were identified by fitting the simulation to experimental results (equilibrium curves and pulse tests). Then, the column model was validated with original experimental results from a binary mixture pulse test (formic and succinic acids). Results show that simulations are much more predictive for multi-component pulse tests, both in terms of profile shape and retention time, which cannot be captured without considering ion-exchange.