(224h) Computational Studies of Local Friction Coefficients in Confining Geometries

Sturtevant, B. D. - Presenter, Purdue University
Corti, D. S. - Presenter, Purdue University

As devices become smaller and we ever-increasingly rely upon molecular or colloidal particles to self-assemble, an important problem that remains largely un-explored yet vastly important is determining the relative effect of bounding surfaces on the dynamics of suspended entities. There have been studies, both experimental as well as computational, of the influence of planar surfaces on the dynamics of individual particles. More complicated situations, however, have not been considered in great detail, such as a particle diffusing in a corner, or two or more particles, both of which are interacting hydrodynamically with a planar wall, diffusing relative to each other. The purpose of this work, then, is to investigate the dynamics of model colloidal particles moving in corners as well as groups of colloidal particles interacting with planar bounding surfaces. More specifically, we calculate local friction coefficients of colloidal particles in these geometries as these quantities can be directly related to local diffusion coefficients. To probe the long-time scales that are necessary to study colloidal motion, we rely on coarse-grained meso-scopic simulation techniques such as Stochastic Rotation Dynamics (SRD)1. Transport coefficients, such as the local friction coefficient, are extracted from simulation data by relying upon a Generalized Langevin Equation framework to capture the essence of the coarse-graining procedure and focus on the dynamics of the colloidal particle themselves2.

While use of meso-scopic techniques is essential to the computational exploration of colloidal behavior, the effect of coarse-graining, where there is some loss of detail regarding the surrounding fluid, raises some issues that have not yet been satisfactorily answered. Although techniques like SRD have been shown to correctly replicate hydrodynamic effects on long-length scales, they are not capable of replicating the features of molecular structure that occur on the much smaller, atomic length scales. Furthermore, they fail to resolve hydrodynamic fields at small length scales. The investigation of local fluid structural effects as well as small length-scale hydrodynamics is achieved by performing full-scale molecular dynamics simulations of large particles suspended in smaller fluid molecules. We simulate the large particles, which are more aptly classified as nano-colloids, in the same physical scenarios as we performed meso-scopic simulations: simulations of a colloidal particle in a corner and simulations of pairs of particles interacting with a planar surface. We also rely on a Generalized Langevin Equation framework to isolate the dynamics of the colloidal particles from the dynamics of the fluid. In both meso-scopic simulations as well as molecular dynamics simulations we compare our results to experimental data or theoretical results for colloidal particles diffusing near a surface.

1: J.T. Padding and A.A. Louis. Phys. Rev. E., 74:031402 (2006). 2: J.T. Padding and W.J. Briels. J. Chem. Phys., 132:054511 (2010).