(717e) Low-Dimensional Dynamic Models of Colloidal Cluster Assembly Using Tunable Electric Fields

Electric fields can be used to facilitate the assembly of small numbers (10-1000) of colloidal particles into ordered structures that may be employed as photonic crystals and other meta-materials. Control of such assembly processes in real time would be enhanced by the availability of low-dimensional dynamic models that accurately capture the particle-level physics. In this paper we describe the building of such models for a specific system: 210 silica particles of micron size in aqueous solution under the influence of a quadrupole field with adjustable voltage. We use a combination of Monte Carlo umbrella sampling and diffusion maps (DMaps) to identify the slow, low-dimensional manifolds in this system. The DMap coordinates are correlated against set of candidate order parameters (OPs) to identify a suitable choice of observables. We build Fokker-Planck (Smoluchowski) models in the chosen OPs and generate free energy and diffusivity landscapes at various voltage levels. Strategies for incorporating this information into real-time process control algorithms are also discussed.