Simulation of Rodlike Particles in Field-Flow Fractionation (FFF)

The manufacturing of single wall carbon nanotubes (SWNTs) results in mixtures that are polydisperse with respect to both size and chirality, the latter of which strongly affects the tube electronic properties. For nanotubes to achieve their full potential in applications, it is desirable to be able to separate them according to both of these physical characteristics. Field-flow fractionation (FFF) (1-3) is an liquid based technique used to separate various macromolecular, colloidal, and particulate materials ranging from 10 to 10 μm in size and has recently been applied in a number of studies (49) to the separation of nanotubes. In FFF, a mixture to be separated is driven through a channel (the “flow”, in FFF) while a field is applied in a direction perpendicular to the streamwise direction. The perpendicular field may be either another flow field, an electric field, or a temperature gradient, amongst a number of possibilities. The perpendicular field is chosen so that the interaction between the field and the streamwise parabolic profile, promotes a separation of components based on their mobility in the field. In this work, we develop a Brownian dynamics simulation of prolate ellipsoidal particles to investigate the separation of rodlike particles in FFF. The particle motions are governed by stochastic forms of a linear momentum balance with orientation dependent drag and diffusion coefficients, and the Jeffery equation with rotational diffusion (10; 11). The simulation shows that nanotube scale particles would be expected to elute by a normal mode mechanism, based on a particle diameter of 1 nm. Elution profiles and average velocity through the device as a function of particle size, and throughput and cross flow flowrates are examined.