Diffusion-Based Separation Method Using Bidirectional Electroosmotic Flow

We present a novel method and microfluidic device that leverages non-uniform electroosmotic flow patterning to achieve highly efficient diffusion-based separation of molecules. In contrast to existing separation methods that rely on microfabricated elements, which may suffer clogging and adsorption of sample to the wall, our method relies solely on modulation of flow through the interaction of an electric field with non-uniform surface charge in an unobstructed microfluidic chamber.

We believe this method could serve as an in-situ sample preparation step for lab-on-chip applications.

Figure 1. Working principle of the diffusion-based separation. The separation is performed in a microfluidic channel where bidirectional electroosmotic flow is established. At t = 0, we place low-diffusivity (red) and high-diffusivity particles (blue) in the left reservoir. The blue particles, due to their high diffusivity, rapidly transverse across stripes and therefore experience a net-zero velocity. Contrary, the red particles diffuse slower and maintain their advection trajectory (t = t1) towards the right reservoir where they can be collected, and thus extracted from the mixture (t = t2).

Figure 2. Experimental results demonstrating separation of low-diffusivity 1-μm beads (D = 4.1·10-13 m2s-1) from high diffusivity rhodamine B (D = 4.2·10-10 m2s-1). (a) Schematic of the device composed of a microfluidic chamber, glass substrate patterned with stripes of different zeta potential, and electrodes immersed in the reservoirs to create an electric field. (b) Image of the glass substrate (dark stripes) after deposition of FITC-labeled PAH stripes (bright stripes). (c) Upon application of an electric field, a bidirectional EOF is established (white arrows). Rhodamine B rapidly diffuses between streamlines resulting in a dispersed yet stationary front, while beads remain on their streamlines and are advected towards the opposite reservoir. The experiments were carried out at a field strength of ~16 V/cm using a buffer composition of 5 mM bistris and 2.5 mM HCl (pH 6.5).

Figure 3. Monte Carlo simulation results showing the efficiency and the purity of separation of 15-μm cells (D=3.3·10-14 m2s-1) from proteins (D=7.8·10-11 m2s-1) for three different channels length. Whereas the channel length does not affect the efficiency (i.e. the ratio of extracted cells compared to their initial concertation), longer channels leads to higher purity of cells (i.e. the ratio of extracted cells compared to the extracted proteins). For a 20 mm long channel, the separation mechanism would be capable of nearly 97% extraction efficiency and 96% purity, in less than 2 minutes.