Solvent Effect on the Diffusion of Unentangled Linear Polymer Melts

Abstract

We conducted molecular dynamics (MD) simulations to study how solvent chains affect the diffusion of linear polymers in the unentangled regime. For monodisperse solvent chains, the self-diffusivity of a tagged chain scales with its chain length. The solvent chain length affects both the prefactor and the exponent, the latter of which ranges from -0.79 to -0.85. The scaling exponent here deviates from -1 as predicted by the Rouse model, which may suggest that the friction coefficient increases with the solvent chain length. In addition, we carried out diffusion simulations on two polydisperse melts, one with the Flory-Schulz distribution and the other with the Gaussian distribution. The measured diffusivity as a function of the tagged chain length agrees with a simple proposed model accounting for the heterogeneous medium.

Fluid-solid phase transition of n-alkane mixtures: Coarse-grained molecular dynamics simulations and diffusion-ordered spectroscopy nuclear magnetic resonance

“…Molecular simulation studies can only shed light on the incipient stages of the formation of the crystalline phase and allow us to study only the first stages of the progression towards solidification facilitated by the decrease in temperature. We compare our results of cooling from a liquid state with simulation models for n-alkanes reported in the literature corresponding to higher-fidelity united-atom (UA) models, where a unique force center is employed for each carbon center and its associated hydrogen atoms 38,41,53,54 . At this level of description, diffusion coefficients have been favorably compared to NMR results for smaller alkanes 55 .…”

“…Studies are challenged by the fact that while crystallization is a process that spans several length and time scales, classical simulation techniques have a spatial and temporal resolution of only a few nanometers and nanoseconds. Although recent advances in molecular modelling, mainly employing coarse-graining strategies, have reduced the limitations associated with atomistic models 37,38 , they, nevertheless, remain insufficient to study crystallization fully in complex multicomponent mixtures.…”

A graph theory assisted Monte Carlo algorithm for simulating the topology structure of LDPE and connection with molecular dynamics simulation

“…35 In literature, researchers usually predefine the structure of target macromolecules or assume some particular distribution of properties. 36 In other words, these investigations lack the reactor configuration part to some degree, which density property can be predicted from our modeling method. In addition, more properties supposed to be calculated similarly in theory, thus they are is also put in the MD part.…”