A Fast Algorithm to Predict Cell Trajectories in Microdevices Using Dielectrophoresis

Abstract

Prediction of accurate trajectories of biological particles is necessary for efficient design of microdevices for dielectrophoretic manipulation. Due to simplicity, a point-based method is generally used to simulate particle trajectory, but a point-based method provides significant distortion from the actual path when particle size is comparable to the device characteristic dimension. This article reports an efficient numerical model which can overcome these drawbacks of a point-based method and can be used for accurate predictions of particle trajectory. This model is formulated on a distributed Lagrange multiplier based fictitious domain (FD) approach for flow field and motion of particles, and the multi-domain method for electric potential. In this study, the dielectric forces are calculated from the Maxwell stress tensor. The accuracy of the proposed methods are validated separately with two test problems: terminal velocity of a particle in a stationary fluid while it is pulled with a constant force, and dielectrophoretic force acting on a particle when it is placed in between two planar electrodes. The capability of the proposed model is demonstrated by simulating trajectories of two biological particles (cells) of the same geometry and size but different dielectric properties in a microdevice. The effects of frequency, particle-particle initial separation distance, and particles’ relative position are investigated. Numerical results indicate that this model can capture the physics of particle manipulation better than the conventional point-based method. Moreover, this algorithm reduces the computational time significantly, which is a major bottleneck in 3-D simulation.

Acknowledgments

The research was supported by Washington State University.