New strategies to improve the sensitivity of capillary electrophoresis for carbohydrate analysis

This study describes various strategies to improve the sensitivity of carbohydrate analysis in capillary electrophoresis (CE). The use of pH stacking is investigated in conjunction with 5-aminofluorescein as a derivatisation agent for the sensitive analysis of simple sugars such as glucose, lactose and maltotriose by CE. The derivatisation agent was selected on the basis of its extremely high molar absorptivity, its compatibility with a 488 nm light-emitting diode (LED) and the fact that it has two ionisable groups making it compatible with on-line stacking using a dynamic pH junction. The influence of both acetic and formic acids were investigated with regard to both derivatisation efficiency and the ability to stack using a dynamic pH junction. Superior sensitivity and resolution was obtained in formic acid over acetic acid. Simulation studies combined with experimental data showed the separation with the best resolution and greatest sensitivity when the carbohydrates were derivatised with the 95 mM formic acid. Using this method efficiencies of 150,000 plates and detection limits at 8.5 x 10‚ÄövÖ¬™‚ÄövÖ‚àè M for mono-, di- and tri-saccharides were achieved. The current system demonstrates a 515 times improvement factoring sensitivity when compared to using a normal deuterium (D‚Äövává) lamp, and 16 times improvement over other systems using lightemitting diodes (LEDs). A novel fluorescent tag, 0-2-[aminoethyl]fluorescein, was developed for the separation of sugars by CE with laser-induced fluorescence (LIF) detection using an argon ion laser. The tag was synthesised using three consecutive steps namely: esterification, alkylation and hydrolysis, specifically designed to offer a flexible way in which to make an assortment of fluorescent tags from cheap and readily available starting reagents (typically less than $1 per g of fluorescent tag). 0-2-[Aminoethyl]fluorescein was equipped with a spacer group to lower steric effects between the fluorescein backbone and the reducing end of the carbohydrate which were anticipated to improve the reactivity of the tag. Fluorescence studies of the novel tag revealed a quantum yield (QY) of 0.24, when using fluorescein as a standard. Kinetic studies were also conducted to compare and assess the performance of aromatic and aliphatic amines using the novel tag and two commercial fluorescent fluorescein motifs where the aromatic amine derivative demonstrated better labelling performance. The separation performance of all the tags was also benchmarked using a range of corn syrup oligosaccharides. The application of the novel tag to a set of oligosaccharides produced a baseline separation of seven different sugar units, with 1 nM detection limit for maltoheptaose. A CE method was designed with on-line concentration which can be translated directly to a microchip format allowing preconcentration via dynamic pH junction. Optimisation of the separation selectivity yielded best separations using a 170 mM ammonium borate buffer at pH 8.60 in an acrylamide coated capillary. When using the current system via LIP, limits of detection (LODs) as low as 0.13 nM for maltose were obtained, which were 10 times lower than could be achieved without on-line concentration. In order to implement this system in a glass/polydimethylsiloxane (PDMS) microchip, the low pH sample was introduced into the microchannels via a cathodic pH independent electro-osmotic flow (EOF) generated using a polyelectrolyte multilayer coating. Microchip separations of maltose, glucose, galactose and allose with dynamic pH junction, were achieved within 120 s, with the limit of detection (LOD) of maltose using a light-emitting diode induced fluorescence (LEDIF) detection system being 790 nM. This is the first implementation of on-line concentration via a dynamic pH junction in a microchip, and significantly, the improvement in sensitivity achieved when translated to the microchip was equivalent to that achieved in capillaries.