We studied the fractionation of poly(ethylene oxide) (PEO) in isobutyric acid and water at different temperatures below the critical point.  To do so, different molecular weight PEO with different end groups have been used.  All the solutions were prepared at the critical composition of the binary liquids.  During the fractionation, the polymer distributes itself in the lower and upper phases as a function of molecular weight.  The results of the higher molecular weight polymer confirmed previous studies done by Shresth.  The current studies show that the molecular weight and the end group of the polymer seem to have no effect on the fractionation.  The fractionation of the lower molecular weight PEO with OCH3 termination is most powerful when compare to the higher molecular weight with OH end group.

       Two non-linear equations were then used to fit the number fractions molecular weight distributions in order to determine the type of distribution.  The relative mass of polymer in each phase is then calculated.  Most of the polymer migrates to the upper phase leaving almost little in the lower phase.  The ratio of the mass fraction in the upper phase to that in the lower phase.

Poly(ethylene oxide)

       Since Albertsson confirmed their use in bioseparation in the late 50s,  poly(ethylene oxide) (PEO) and poly(ethylene glycol) (PEG) have attracted the wide attention of the scientific world.  Showing interesting and peculiar characteristics in organic and aqueous solutions, PEO has been extensively studied.  Poly(ethylene oxide) with smaller molecular weight (200-20000 g/mol) is often referred as PEG.  In addition, any PEO that has hydroxyl groups at each extremity of the molecule is also called PEG.  The appellation of PEO is generally used for higher molecular weights or molecules with methyl oxide groups at each end.

Theoretical fractionation

       When a polydisperse polymer is dissolved in a binary solution of two liquid phases, we expect the polymer to distribute itself as a function of molecular weight between the upper phase and the lower phase at equilibrium.  The Flory-Huggins (FH) theory, which pioneered the understanding of this effect, cannot quantitatively describe such polymer fractionation because, like Van der Waals, FH ignores all non-mean field effects. Several refined theories have helped deal with the thermodynamics of the polymer/solvent systems.  Stockmayer and co-workers have proposed a “bridging” expression that considers all the interactions. That expression resulted in the addition of nonlinear terms in the Flory equation.  Evans and coworkers presented a theory of a polydisperse system based on the perturbation of a monodisperse system.   This theory predicted the ratio of the moments of the parent and daughter in homogeneous solutions or coexisting phases.  In fact, they showed that the difference in number average molecular weight, Mn of the daughter phases should be proportional to the skewness of the parent distribution. This theory was later contradicted by Xu and Baus, “but

supported by measurements on colloidal particles.

       In 1995, Ten Brinke and Szleifer constructed a new theory that takes into account the non-mean field intramolecular and mean-field intermolecular interactions.  Using Monte Carlo calculations, this theory predicts the full molecular weight distribution and the distribution coefficient of the polymer.