Calculation of ternary polymer solution phase diagram via compressible lattice fluid model extended to specific interactions

Abstract

Ternary polymer solutions are currently employed for the production of porous structures (e.g. membranes and scaffolds for tissue engineering) via liquid-liquid phase separation techniques. The relation between processing conditions and thermodynamic features of the considered polymeric system determines the morphology of the as-prepared membrane. Therefore, in order to achieve a better control on a porous structure, a deep characterization of the phase behavior of a polymer solution is needed. The experimental derivation of phase diagrams is highly cost and time consuming, thus usually the cloud point curves of the system are directly measured. In this work, the compressible lattice fluid extended to specific interactions was applied on a ternary polymer solution for the derivation of a complete phase diagram, where the cloud point data were employed for fitting the model parameters. The investigation was focused on the poly-L-lactide (PLLA)-dioxane-water system, as it has been employed for the preparation of both microfiltration membranes and scaffolds for tissue engineering purposes. Although the wide and growing scientific interest upon this system (and, in general, on polymer-solvent-nonsolvent systems), a complete phase diagram of the system is not available in literature. The main feature of the considered model is the possibility to take into account for specific interactions between the species, thus improving the system description. The original formulation was derived and checked for a binary system: in this case, the applicability is tested on a ternary system. Hansen solubility parameters were scrutinized to select the appropriate interaction parameters to be taken into account: the total number of fitting parameters used was reduced to three, upon six available, by considering specific interactions between the PLLA-dioxane pair and nonspecific interactions for the PLLA-water pair. The adjustable parameters were fitted on the basis of experimental cloud point data. The binodal curves were calculated by equating the chemical potentials of species in the separated phases, and by solving the system with the equation of state for each phase. The cloud points were determined by the intersection of the binodal curve with a straight line corresponding to a fixed solvent/nonsolvent proportion. The model is able to reproduce qualitatively and quantitatively the cloud point curves of the system. The as-obtained interaction parameters were used to calculate a complete phase diagram of the system. Thus, from simple experimental data, as cloud point curves are, a more complex information was derived. In principle, the proposed method can be employed for the derivation of the phase diagrams of similar systems, thus improving the information supporting the microstructure control of polymeric foams.