Molecular Dynamics Assessment of Small Molecule Diffusion in Medical Plastics
The ability to accurately predict solute diffusivity in polymeric matrices has far-reaching implications, particularly in the capacity of predicting the behavior of polymer matrix composites used in medical devices. We have recently demonstrated that it is possible to use all-atom molecular dynamics simulations to acquire these values. However, as the size of the solute increases or polymer matrix becomes more glassy, the simulation time required becomes prohibitively long, and the observed dynamics are restricted to the sub-diffusive regime. Thus, we seek to explore potential relationships between the sub-diffusive and macroscale (Brownian) behavior. Here, we have characterized the dynamics of a model system within the context of generalized Langevin dynamics. Specifically, we compute the Laplace frequency dependence of the memory kernel Γ (s) that quantifies non-trivial drag and random forces in the observed dynamics. Our observations span the inertial, sub-diffusive, and Brownian regimes overa wide range of temperatures. We find in all cases, that Γ (s) is constant as s → 0 or ∞, suggesting ideal dynamics, i.e “instantaneous” drag force and white noise, in these regimes. For intermediate values of s, Γ (s) exhibit power-law relationships with exponents that increase monotonically with system temperature. These observations may serve as a basis to link the sub-diffusive and macroscale dynamics in these systems.
Keywordsdiffusion polymers molecular dynamics memory kernel
Unable to display preview. Download preview PDF.
- Christopher Forrey, David M Saylor, Joshua S Silverstein, Jack F Douglas, Eric M Davis, and Yossef A Elabd. Prediction and validation of diff coefficients in a model drug delivery system using microsecond atomistic molecular dynamics simulation and vapour sorption analysis. Soft Matter, 2014, in press.Google Scholar