Lattice Monte Carlo simulation for the partitioning of a bimodal polymer mixture into a slit

Partitioning of a bimodal polymer mixture of different chain lengths into a narrow slit has been examined by using lattice Monte Carlo simulations in two approaches: the regular canonical ensemble simulation with a twin box and the chain insertion method. The regular canonical ensemble simulation determines the partition coefficients of the two components directly from the simulation. The results reveal that the short chains have an enhanced partitioning and the long chains have a reduced partitioning into the pore in the bimodal mixture compared with the monodisperse systems at the same concentration. In addition, the partition coefficients of the two components in the bimodal mixture maximize their difference in the semidilute concentration, substantiating the separation mechanism of the high osmotic pressure chromatography. The chain insertion method determines the chemical potentials of the chains in the bimodal mixture in the bulk and in the slit. It was found that the chemical potential depends strongly on the total volume fraction of the polymers but only weakly on the composition of the mixture in both the confined and the bulk solutions. The partition coefficients in the bimodal mixture were estimated from the chemical potentials by neglecting the dependence of the chemical potentials on the composition. The latter produced a less pronounced enhancement and reduction in the partition coefficients of the two components compared with those obtained from the simulations using the twin box.