A LATTICE MONTE-CARLO STUDY OF LONG-CHAIN CONFORMATIONS AT SOLID POLYMER MELT INTERFACES

In this paper we present a comprehensive lattice Monte Carlo study of long chain conformations at solid-polymer melt interfaces. Segmental scale interfacial features, like the bond orientational distribution were found to be independent of surface-segment energetics, and statistically identical with Helfand's predictions for the full-occupancy, infinite chain length limit. Conformational statistics of chains longer than 5-6 statistical segments followed the predictions of the Scheutjens-Fleer theory and the same power laws as a single ideal chain at the critical value of the surface-segment adsorption free-energy. Our simulations tested the predictions of random walk next to a ''reflective'' surface statistics for the spatial variations of chain dimensions and chain center of mass density. It was found that these statistics furnish the correct long chain limit, independently of surface-segment energetics. The random walk next to a ''reflective'' boundary predictions for the ''adsorbed'' amount and the distributions of tail, loop, and train number, as well as tail and loop size were in quantitative agreement with the simulation data. The correspondence between random walks and real (or simulated) chains required the knowledge of a single, microscopic parameter, the number of chemical segments per statistical segment, a(infinity). This quantity was very close to the average length of adsorbed sequences (trains), in the long chain limit. Our simulations tested thoroughly and established firmly the validity of ''reflective'' boundary statistics in the melt. The inevitability of these statistics has broad implications on ''desorption'' kinetics, chain mobility, and chain relaxation, which are currently under study.