Thanks to its favorable reduction scale law, and its easy integration, electrokinetics has emerged over the last fifteen years as one of the major solution to drive flows in fully integrated lab-on-chip. At microscale, an efficient mixing is a keystep which can dramatically accelerate bio-reactions. For thirty years, Dynamical System theory has predicted that chaotic mixing must involve at least 3 dimensions (either time dependent 2D flows or 3D flows). However, in microfluidics, few works have yet presented efficient embedded micromixers. This paper presents experimental and theoretical study of 2D time dependent chaotic mixing using AC electrothermal fluid flows. Experiments and numerical simulations are performed on a top view device and a sideview device. In both devices, a sinusoidal electric signal is applied between 3 interdigitated gold electrodes. A phase signal Vpp = 11V and a ground are switched between the two side electrodes using a step function, whereas the opposite phase signal –Vpp is steadily applied to the center electrode (Figure 1). Flow velocity is measured by micro particle image velocimetry μ PIV. The velocity profile shows a dramatic asymmetry between the two vortices. Therefore, during the switch, vortices overlap, leading to stretching and folding flows required to obtain chaotic mixing (Figure 3 and 4). The experimental measurements validate our electrothermal models based on our previous work [1]. The mixing efficiency of low diffusive particles is studied at multiscale using the mix-variance coefficient (MVC) [2] to evaluate mixing at different scales (Figure 4). To do so, the domain is successively divided in boxes along the x and y direction up to nx and ny boxes, respectively. For each box configuration, average bead concentration is computed. The variance of these concentrations is then evaluated: MVCs=1nxny∑i=1ny∑j=1nxρij-0.52. The result of numerically evaluated MVC in Figure 2 show a dramatic increase of mixing efficiency with blinking vortices compared to steady flow.
Theoretical, experimental and simulation results of the mixing process will be presented.