Thermodynamic modeling of PEG 6000 – Dextran T70 Aqueous Two-Phase System Using NRTL Activity Coefficient Model and Particle Swarm Optimization Algorithm

As an extraction technique for the purification of biomaterials, aqueous two-phase systems (ATPSs) have been the focus of increasing attention in recent years. Their special application in overcoming low product yield in conventional fermentation processes is of note. Essentially, an aqueous two-phase system is formed because of the mutual incompatibility of two hydrophilic polymers or a polymer and a salt at specific range of compositions above the critical composition of the ATPS system of interest. ATPSs are especially unique in the fact that the low surface tension between the two phases results in an extraction process where the expensive protein and biomolecule products are subject to a relatively less harsh and damaging transfer between the phases. In this study the NRTL (nonrandom two-liquid) activity model is used, in a modified version in terms of mass instead of mol fractions, suitable for systems comprised of components with relatively high molecular weights. For the ternary system of interest, PEG 6000 – Dextran T70 – Water, the binary interaction as well as the non-randomness parameters of the model are estimated by fitting them to the experimental data by employing the global, evolutionary based Particle Swarm Optimization algorithm with acceptable results.