Patterning Electro-osmotic Flow with Patterned Surface Charge

Abstract

This Letter reports the measurement of electro-osmotic flows (EOF ) in microchannels with surface charge patterned on the 200 mm scale. We have investigated two classes of patterns: (1) Those in which the surface charge varies along a direction perpendicular to the electric field used to drive the EOF; this type of pattern generates multidirectional flow along the direction of the field. (2) Those in which the surface charge pattern varies parallel to the field; this pattern generates recirculating cellular flow, and thus causes motion both parallel and perpendicular to the external field. Measurements of both of these flows agree well with theory in the limit of thin double layers and low surface potential. PACS numbers: 47.54. + r, 47.65. + a, 61.20.Lc, 82.45. + z Microfluidic devices are important in a range of applications: microanalytical systems for genomics and proteomics, research tools for cell biology, and low-inventory chemical synthesis. Developments in microfluidics will require sophisticated methods for handling fluids in the low Reynolds number (Re) regime that is imposed by the dimensions of the microchannels in these devices. Electrically driven flow [electro-osmotic flow (EOF)] is an alternative to pressure driven flow in microchannels, but has been used only for the simplest case, that is, for flow driven by uniformly charged channel walls. In this Letter, we describe EOF driven by surface charge patterned on the channel walls [1]. The geometries that we explore demonstrate two new types of fluid behavior at low Reynolds number: multidirectional flow along the axis of the channel and the applied field, and controlled cellular flow with bulk motion both perpendicular and parallel to the axis of the channel and the applied field. We believe that these new types of flows offer opportunities for fine control of fluid motion and chemical transport in microfluidic devices, and also suggest the use of fluids to drive micromechanical elements via shear-generated torques.The experimental challenge of patterning surfaces inside microstructures has hindered the investigation of EOF in the presence of patterned surface charge. For this study, we used two soft lithographic techniques [2] to solve this problem in patterning. We used laminar flow patterning [3] to generate longitudinal patterns, and a modified version of micromolding in capillaries to generate transverse patterns [4]. In both cases, the electrostatic adsorption of organic polymers having charged side groups [poly(ions)][5] onto the inside walls of the channels generated the required regions of positive and negative surface charge, and ensured that these regions had charge densities of similar magnitude [6].Electro-osmotic flow [7] uses an externally applied electric field parallel to a surface with net surface charge density to drive bulk motion in a fluid. The field interacts with the net charge carried by the mobile screening layer (the Debye layer) in the fluid above the surface. A channel with walls having fixed charges thus generates...

Induced-charge electro-osmosis

“…Anderson & Idol (1985) explored electro-osmotic flow in inhomogeneously-charged pores, and found eddies and recirculation regions. Ajdari (1995;2001) and Stroock et al (2000) showed that a net electro-osmotic flow could be driven either parallel or perpendicular to an applied field by modulating the surface and charge density of a microchannel, and Gitlin et al (2003) have implemented these ideas to make a 'transverse electrokinetic pump'. Such transverse pumps have the advantage that a strong field can be established with a low voltage applied across a narrow channel.…”

Electrokinetic generation of microvortex patterns in a microchannel liquid flow

“…A variety of techniques have been proposed to locally generate various charge patterns. Laminar flow patterning was used to generate longitudinal patterns, while a modified version of micromolding in capillaries was used to create transverse patterns [13]. In another approach, polymer coating by an electrostatic-self-assembly (ESA) combined with standard photolithography was used to generate several different surface charge patterns [14].…”

Three-Dimensional Structure of Electroosmotic Flow over Heterogeneous Surfaces

“…Results from the three-dimensional model were verified against the well-known classical results for homogeneous surfaces 34,35 as well as the one-dimensional experimental results presented by Herr et al 27 In addition, the BLOCS code used in this study has been checked against several classical flow test problems, such as flow over a backward step and flow in a square cavity, and experimentally verified in previous works. 36 Of immediate interest here would be a qualitative comparison with the two-dimensional results of Ajdari 20 and Stroock et al 25 using the lengthwise and crosswise strip patterns as shown in Figure 1a,b (note that Ajdari actually considered a sinusoidally varying surface charge density as opposed to the step changes examined here; however, as will be shown, a similar flow structure exists). To do this, we consider a computational domain with dimensions l x ) 50 µm, l y ) l z ) 25 µm (the x-y-z coordinate system is as shown in Figure 2) containing a 1 × 10 -5 M KCl solution (thus a double-layer thickness, 1/κ, of 0.1 µm) and an applied driving voltage of 500 V/cm.…”

“…The results for the lengthwise strips are shown in Figure 3 and Figure 4 for the crosswise strips. 25 the flow is directed along the negative x-axis over the heterogeneous strips and the positive x-axis over the homogeneous patch, in both cases becoming weaker and approaching zero near the transition point. Also of interest is the change in the magnitude of the induced flow velocity with increasing distance from the surface.…”

“…The first two of these patterns, Figure 1, parts a and b, are in essence one-dimensional surface patterns and are only considered here in order to provide a validation of the results via a qualitative comparison with the aforementioned studies of Ajdari 20 and Stroock et al 25 Since the pattern is repeating, the computational domain is reduced to that over a single periodic cell, demonstrated by the dashed lines in Figure 1. Note that this periodic boundary condition ignores any entrance/exit effects and thus is not valid for the first and last periodic units on the surface.…”