Martin S. High

A group contribution equation of state for polymer solutions

Abstract A new group contribution lattice-fluid (GCLF) equation of state has been developed for polymers and polymer solutions. The lattice fluid theory of Panayiotou and Vera has been revised to include group contributions for solvent and polymer molecules. A lattice model containing holes is a convenient means of representing a polymer solution. The holes in the lattice serve as a means of varying the density of the fluid and are analogous to the free volume in generalized van der Waals models. An external degree of freedom parameter, the “C” parameter is not required. The GCLF equation of state reduces to forms similar to well established models for polymer solutions such as the Flory-Huggins equation. The group contributions for the model were calculated from the saturated liquid density and vapor pressures of small molecules. These group contributions are able to predict liquid densities, vapor pressures and VLE of small molecular systems with reasonable accuracy. The extension of the GCLF model to polymer solutions requires no pure component data and only one binary parameter.

Prediction of solvent activities in polymer solutions

Abstract The Holten-Andersen equation of state and the UNIFAC-FV (Free Volume) activity coefficient models are compared using a polymer solution data base. Both the Holten-Andersen equation of state and the UNIFAC-FV models have a foundation in the equation of state developed by Flory for polymer solutions. The Holten-Andersen equation of state is presented here with some corrections from the original publication. The UNIFAC-FV model is an extension of the UNFAC model for polymer solutions and includes the free volume expression from the Flory equation of state. The Holten-Andersen group contribution equation of state and the UNIFAC-FV activity coefficient model are the only predictive models available that were developed specifically for polymer solutions. Unlike the older correlative methods, these two models have not been previously tested and compared using an experimental data base.