Polymer Science Faculty Research

Title

Monte Carlo simulation of solvent effects on the threading of poly (ethylene oxide)

Document Type

Article

Publication Date

4-19-2004

Abstract

Recently, we described a coarse-grained model of poly(ethylene oxide) (PEO) and then employed that model to study the amount of spontaneous threading of cyclic molecules by linear chains in the melt (Helfer et al Macromolecules at press). We now use that coarse-grained model to determine how the amount of threading is affected by dilution of the PEO melt with good, Θ, or poor solvents. We employ two distinctly different methods for incorporating the solvent. One method employs explicit coarse-grained solvent molecules, which have the same size as a coarse-grained bead of the polymer. Solvents of different quality are produced by changes in the values of the interaction energies between solvent and polymer beads. The other method does not employ explicit solvent particles but simply modifies the attractive part of the discretized Lennard-Jones potential that describes the pair-wise long-range interaction of coarse-grained PEO while leaving unchanged the rotational isomeric state model that describes the short-range intramolecular interactions. The thermodynamic quality of this implicit solvent is inferred from its effect on the mean square radius of gyration of PEO. The intermolecular pair correlation functions show that cyclic and linear molecules aggregate together in a poor solvent, which contributes to the increase in the amount of threading, relative to the situation found at the same concentration in a good solvent. However, the magnitude of this effect of the poor solvent is insufficient to overcome the decrease in threading upon increasing dilution expected from simple application of the law of mass action. Although the qualitative effects are similar for the two different methods of treating the solvent, quantitative comparison suggests that the use of explicit solvent particles gives a stronger connection with the anticipated behaviour of real systems. The present results for equilibrium structures in a Monte Carlo simulation are not as sensitive to the details of the treatment of the solvent as were earlier results (Chang and Yethiraj 2001 J. Chem. Phys. 114 7688) for the dynamics of the collapse of a single chain in a poor solvent.

Publication Title

Modelling Simulation Material Science and Engineering

Volume

12

First Page

S59

Last Page

S71