Co-nonsolvency Transition in Polymer Solutions: A Simulation Study

Abstract

We study the phase segregation in polymer solutions in the presence of a co-nonsolvent (CNS) using molecular dynamics simulations, where CNS particles are a preferential solvent for the polymers. To investigate the condensation transition, we use a movable wall that exerts an osmotic pressure on the polymers but is permeable with respect to the CNS. We focus on the semidilute state of the polymer solution in the absence of CNS. Increasing the amount of CNS results in a condensation of the polymer−CNS phase, which is related to a sharp drop of the polymer volume, followed by a re-entry behavior at higher concentrations of CNS. Simulation results for different interaction energies between monomers and CNS particles are compared with the prediction of the adsorption−attraction mean-field theory under constant osmotic pressure. By calculating the radius of gyration and the structure factor of the individual chains, we found that the conformation statistics of the chains is consistent with the behavior in semidilute solution at the given concentration, resulting from the co-nonsolvency response. Concerning the diffusion dynamics of the polymer chains, we found a monotonous slowing down with increasing CNS concentration and no indication for network-like dynamics or exceptional slow dynamics at the condensation point. Our results are consistent with the collective nature of the co-nonsolvency transition due to weak and transient monomer−CNS−monomer bridging which can be rationalized using mean-field-like concepts.

Phase Separation of Homopolymers in Binary Mixed Solvents─The Co-Nonsolvency Effect

“…Further, using eq , we obtain ϕ A,c u ≈ 0.432, ϕ B,c u ≈ 0.568 and ϕ A,c u = 0.333, ϕ B,c u ≈ 0.667 at the upper critical point in the limit of N → ∞ for the two parameter sets in Figure a,b, respectively. Finally, we note that the critical behaviors obtained here are rather different from those predicted by the adsorption–attraction model. ,, Specifically, the adsorption–attraction model predicts a type-II phase transition that arises from the negative third virial coefficient but with the second virial coefficient being positive.…”

“…Finally, we note that the critical behaviors obtained here are rather different from those predicted by the adsorption−attraction model. 36,38,41 Specifically, the adsorption−attraction model predicts a type-II phase transition 68 that arises from the negative third virial coefficient but with the second virial coefficient being positive.…”

“…Significant efforts have been devoted to explaining these phenomena, and currently, there are two main categories of opinions on the molecular driving forces. The first category regards the collapse of a single homopolymer chain or phase separation of homopolymer solutions to the preferential adsorption of one type of solvent over the other to the polymer backbone. ,− Specifically, Tanaka and co-workers attributed the co-nonsolvency effect of a PNIPAM chain in the mixed solvent of water and methanol to the competitive hydrogen bonding by water and methanol molecules onto the polymer chain. − This explanation was supported by the subsequent experiment by Wang et al The coarse-grained molecular dynamics study by Mukherji et al suggests that the co-nonsolvency effect is a generic phenomenon due to the preferential attraction of the better solvent to monomers. ,, By combining the preference attraction concept with the Alexander–de Gennes brush theory, , Sommer developed an adsorption–attraction mean-field model for polymer brushes and found that this preferential attraction can also lead to a swell–collapse–reswell transition in homopolymer brushes in mixtures of two good solvents . He further generalized this theory to homopolymer solutions and found that the preference attraction also results in a re-entrance phase separation in mixtures of binary good solvents .…”