Sample plug induced peak splitting in capillary electrophoresis studied using dual backscattered interferometry and fluorescence detection

The appearance of unexpected peaks in capillary electrophoresis (CE) is common and can lengthen the time of method development as assay conditions and experimental parameters are varied to understand and mitigate the effects of the additional peaks. Additional peaks can arise when a single-analyte zone is split into multiple zones. Understanding the underlying mechanism of these phenomena, recognizing conditions that favor its presence, and knowing how to confirm and eliminate the effect are important for efficient method optimization. In this study, we examine how the overlap of analyte zones with the sample plug can lead to peak splitting. This is explored experimentally using dual detection CE, which enables both the sample plug and analyte zones to be independently and simultaneously measured from the same detection volume. Simulations performed via COMSOL Multiphysics confirm the origin of the splitting and help guide experiments to reduce and eliminate the effect. Our findings show that this peak splitting mechanism can arise in separations of both small and large molecules but is, especially, prevalent in separations of slowly migrating macromolecules. This effect is also more prevalent when using a short length-to-detector, as is commonly found in microfluidic applications. A simple diffusion-less model is introduced to develop strategies for reducing peak splitting that avoids modifying the apparatus, such as by lengthening the separation length, which can be difficult. Decreasing the sample plug length and slowing the electroosmotic flow can both reduce this effect, which is confirmed experimentally.