(621c) Multiscale Framework for Predicting Cyclodextrin Assembly on End-Functionalized Polyethylene Glycol.

The non-covalent self-assembly of a ring molecule on a polymer axel is the first step in the formation of polypseudorotaxanes (PSRs) which have found applications in stimuli-responsive crosslinked networks. Functionalization of chain ends permits tunable linear density and distribution of cyclic molecules on polymer backbones, which impact the final assembly and properties of PSRs. Forays into the PSR design space require successive syntheses and characterizations for each possible end group moiety. The development of a computational model that will predict the linear density based on the functionalized axel would expedite the design and synthesis of high-performance PSRs. We employ a multi-scale approach to predict the assembly of alpha-cyclodextrin (aCD) on functionalized polyethylene glycol (PEG). Combining all-atom molecular dynamics simulations and two-dimensional (2D) umbrella sampling, we compute the free energy landscapes for CD threading onto the functionalized PEG. We show that our atomistic simulations and sampling method accurately capture the thermodynamics of the complexation of aCD with various moieties by comparing our simulation results with experimentally determined binding constants. With the predicted free energy landscapes, we estimate the rate constants for aCD threading, dethreading, and hopping along the PEG backbones. A kinetic Monte Carlo (kMC) model is then applied to predict the linear density and intra-chain distribution of aCD along the polymer axel. We expect our multiscale framework to help efficiently navigate the chemical design space for novel PSR materials.