Journal of Chemical Physics, Vol.114, No.13, 5937-5948, 2001
Monte Carlo simulation for the adsorption of diblock copolymers. I. In nonselective solvent
Extensive Monte Carlo simulations have been performed to study the adsorption of diblock copolymers from a nonselective solvent on an impenetrable surface. The efforts were concentrated on depicting the microstructure of adsorption layers. In simulations, diblock copolymer molecules are modeled as self-avoiding linear chains composed of r(A) segments of A and r(B) segments of B, where the former is attractive to the surface while the latter is nonattractive. The adsorption information including segment density profiles, adsorption amount and isotherms, adsorption layer thickness, bound fraction and surface coverage were obtained by detailed analyses on comprehensive simulation data under various conditions. The microstructure of adsorption layers, primarily the profiles of the adsorbed segments corresponding to tails, loops, and trains, and the size distributions of these adsorption configurations are presented. As a whole, the adsorption layer thickness is mainly determined by the length of the nonattractive block. The effect of the adsorption energy and the chain composition f, the latter is the proportion of attractive segments A in a diblock copolymer chain, on various adsorption properties has been inspected. Comparisons between results of this work and those of previous simulations as well as corresponding experiments were made and many useful conclusions have been drawn. It is shown that the adsorption amount increases monotonically with the increase of f when the adsorption energy is relatively small. However, if the adsorption energy has a larger value, the adsorption amount exhibits a maximum at certain value of f dependent on the length of the block A and the magnitude of the adsorption energy. This trend coincides well with the experimental results of Tiberg [Langmuir 10, 2294 (1994)] and Evers 's SCF calculations [J. Chem. Soc., Faraday Trans. 86, 1333 (1990)]. Why some previous work failed to simulate this phenomenon is also explained. (C) 2001 American Institute of Physics.