Understanding Transport Phenomena in Soft Matter Systems using Multiscale Molecular Simulation Methods

Abstract

Transport properties in soft matter systems underpin a wide range of natural processes and technological applications, from flavour release in foods to drug delivery in nanomedicine. This thesis investigates some of the fundamental mechanisms of molecular transport in polymers, polymeric vesicles, and lipid bilayers using a multiscale
simulation framework, bridging atomistic to mesoscale resolution.

We performed atomistic simulations to investigate the binding of flavour molecules to a a chewing gum polymer matrix made of polyvinyl acetate (PVAc), a widely used
base polymer for chewing gum formulations. We initially compared two different Amber force fields (GAFF and Lipid17) with the combination of two charge models (RESP and AM1-BCC) based on their ability to reproduce experimental density and solubility parameters of PVAc. The GAFF force field and the AM1-BCC charge model reproduced a higher solubility parameter with a lower density, which arose from strong cohesive interactions between the polymer chains and increased cohesive energy density (CED). We performed a potential of mean force calculation to insert a single sorbitol molecule into the PVAc slab from the water region. This revealed a high free energy barrier to insertion into a dry PVAc due to the steric strain that developed when the sorbitol penetrated the dense polymer slab. By local heating of the sorbitol molecule to an elevated temperature, the steric strain was removed through relaxing the local polymer chains, and this significantly reduced the energy barrier for the sorbitol insertion. For a wet polymer slab, significantly improved flavour insertion was seen as the water molecules act as a molecular plasticiser for PVAc by softening the polymer structure and increasing the local free volume.

After understanding this behaviour at the atomic scale, we performed dissipative particle dynamics (DPD) simulations to evaluate the release dynamics of various hydrophobic flavour molecules, such as ethyl hexanoate, carvone, and linalyl acetate, from a model of a realistic chewing gum base composed of polyvinyl acetate and polyisobutylene. Each molecule shows different diffusion trends from phase separated polymer blends. Carvone showed the fastest release, ethyl hexanoate showed moderate release rates, while linalyl acetate showed a very slow release of molecules.

We also investigated the negative chemotaxis of a PEG-b-PLA polymersome in a lactic acid gradient. Atomistic simulations revealed that lactic acid molecules form hydrogen bonds with PLA segments, resulting in polymer expansion. At the mesoscale, our DPD simulations reproduced this result, and for a PEG-b-PLA bilayer in water, demonstrated bilayer expansion and partial exposure of PLA segments to water. This causes the interfacial tension between the membrane and water to increase, leading to the movement of polymersomes away from lactic acid gradients.

Building on these findings, we next investigated the interaction of a functionalised Fe3O4 nanoparticle containing an organic coating with a lipid bilayer to understand the translocation pathways across a cell membrane. Atomistic simulations revealed that cinnamaldehyde and folic acid molecules strongly adsorb onto the nanoparticle surface through hydrogen bonding, thereby stabilising an organic coating. Using the Martini 3 force field, we constructed a coarse-grained model to capture these interactions, incorporating distance restraints to maintain the stability of the coating. Simulations of the coarse-grained models revealed that the nanoparticles can adsorb at a membrane interface and become partially wrapped with a bilayer. The simulation highlights possible mechanisms for nanoparticle uptake and delivery.