# Two-dimensional nanometric confinement of entangled polymer melts

Entry by Yuhang Jin, AP225 Fall 2011

## Contents

## Reference

Yves Termonia, *Polymer*, 2011, **52**, 5193.

## Keywords

entangled polymer melts, two-dimensional confinement, Monte Carlo simulation

## Introduction

Behavior of polymer chains under a two-dimensional confinement is of interest and importance to a variety of engineering applications, and has significant relevance to biological macromolecules as well. Both experimental studies and numerical approaches have been devoted to this topic, with a focus on thin polymer films whose thickness is smaller than the radius of gyration <math>R_g</math>. One important result of these efforts is that for film thickness <math>h<R_g</math> the polymer chains retain their Gaussian behavior.

In this paper, simulation at extreme confinements, <math>h<<R_g</math>, is performed for the first time. A Monte-Carlo approach based on previous work is employed.

## Model

In the simulation, the polymer film of thickness <math>h</math> is represented by a simple cubic lattice with periodic boundary condition along the lateral direction. A dense melt at high thickness <math>h>R_g</math> is first generated, with chains of length <math>N</math> and a small number of vacancies placed on the lattice. The chains are then equilibrated by elementary rotations and exchanges of segments between adjacent chains. It is guaranteed that the process obeys detailed balance, a necessary condition for thermodynamic equilibrium. Full equilibrium is considered to be achieved when the average radius of gyration for the chains and the eigenvalues of their gyration tensor approach steady state values. Next, the film thickness is slowly decreased by randomly picking a small number of chains and converting them into vacancies so that the density remains the same. A new equilibrium can be achieved. The whole process is repeated until a sufficiently small <math>h</math> is attained, as is illustrated in Fig. 1.

## Results

Polymer chains appear as flattened ellipsoids with their large semiaxis being <math>\sqrt{\lambda_1}</math> where <math>\lambda_1</math> is the largest eigenvalue of the gyration tensor **G**. Renormalized quantities are defined: <math>h^*=h/R_{g,bulk}</math> and <math>R_g^*=R_g/R_{g,bulk}</math>. The dependence of <math>\lambda_1/\lambda_3</math> and <math>\lambda_2/\lambda_3</math> on <math>h^*</math> is studied and depicted in Fig. 2. At large <math>h^*</math>, the curves converge towards their limiting values for Gaussian chains. When <math>h^*</math> is below 1, the two curves go up dramatically. However, <math>\lambda_1/\lambda_2</math> remains close to the value expected for Gaussian chains, as is seen in insert. The dependence of <math>R_g^*</math> on <math>h^*</math> is plotted in Fig. 3. A minimum of <math>R_g^*</math> is found around the critical value <math>h^*~1</math>.