Effects of injector geometry and sample matrix on injection and sample loading in integrated capillary electrophoresis devices.

Abstract

The double-T injector design employed in many microchip capillary electrophoresis devices allows for the formation of very small (50-500 pL) sample plugs for subsequent analysis on-chip. In this study, we show that sample plugs formed at the channel junction can be geometrically defined. The channel width and injector symmetry prove to be of great importance to good performance. A unique pushback of solvent into the side channels can be induced when the side channels have a very low resistance to flow, and this helps to better define the injected sample plug. Samples and running buffers of differing ionic strength (e.g., 10 mM KCl buffer and 20 mM KCl sample) can yield widely variable results in terms of plug shape and amount injected (variations of 1.5 to 10x). Applying bias voltages to all the intersecting channels aids in controlling the plug shape. However, when the ionic strengths of buffer and sample are not matched, the actual amount injected (up to 10x variations) can be inconsistent with the appearance of the plug formed in the injector (up to only 30 % variations). Operating at constant pH and ionic strength produced the most consistent results. This report examines the effects of altering the injector geometry and solution ionic strengths, and presents the results of using bias voltages to control plug formation. The observed results should provide a benchmark for modeling of the fluid dynamics in channel intersections.