Improving repeatability of capillary electrophoresis-a critical comparison of ten different capillary inner surfaces and three criteria of peak identification

Abstract

A poor repeatability of migration times caused by the fluctuations of electroosmotic flow (EOF) is an inherent weakness of capillary electrophoresis. Most researchers endeavor to prevent this problem using relative migration times or various capillary coatings which are expensive and not easy in comparison. Herein, we present an original approach to this problem: we apply a model sample designed to induce significant EOF instability, in order to critically compare ten capillary types with different physicochemical characteristics. Moreover, we accompany capillary modification with the evaluation of various criteria of peak identification: migration time, migration times ratio, and electrophoretic mobility. Our results show a great effectiveness of a dynamic coating in the stabilization of migration times, with the average RSD(%) value reduced from 3.5% (bare silica capillary) down to 0.5%. The good outcomes were also obtained for the surfactant-modified silica and amine capillaries. For the capillaries exhibiting significant instability of EOF, electrophoretic mobility turned out to be a more universal and reliable criterion of peak identification than relative migration time. It can be explained by an intrinsic dependency of migration times ratio on EOF change, which should always be considered during the selection of an internal standard.

Keywords: Capillary coating; Capillary electrophoresis; Electrophoretic mobility; Micellar electrokinetic chromatography; Relative migration times; Repeatability.

Fig. 1

Schematic illustration of ten physicochemically different capillaries including their inner surface and the direction of EOF. Symbols: - negative charge; + positive charge; ( ) dynamic polyionic layer (renewable); N₁ polyacrylamide (PAA) layer; N₂ polyvinyl alcohol (PVA) layer; II hydrophobic tail-tail interactions. MEKC-SDS method is presented as a distinct capillary modification (see the text)

Fig. 2

Theoretical simulation of the potential shifts of parameters in the qualitative analysis caused by the EOF change, (A) Using migration times obtained for four different compounds (1–4) exhibiting different migration velocity. (B) Using migration time and relative migration times obtained for one compound (4) and considering three different internal standards (IS, 1–3). The shifts were calculated as a relative change of the parameter upon the given EOF alteration. The inset graphics present schematic electropherogram and adding of vectors for the particular analytes (the initial electroosmotic mobility equals 100 and it varies +/− 10; electrophoretic mobilities are constant and they equal +20, 0, −20, −40 for the given analytes; the values were chosen arbitrarily to visualize the discussed phenomenon)

Fig. 3

Representative electropherograms obtained for all ten capillaries on the highest concentration level (500 μg × mL⁻¹). The negative peaks observed for the DC-silica, DC-amine and SDS-amine capillaries may stem from the lack of kit components and SDS molecules in the sample solution. The EOF strength was measured using always the positive DMSO peak

Fig. 4

The average values of electroosmotic mobility obtained for various capillaries, using the sample containing all analytes on the highest analyte concentration level (500 μg × mL⁻¹)