Indirect photomeric detection of anions in capillary electrophoresis usingdyes as probes and electrolytes buffered with an isoelectric ampholyte

Citation
C. Johns et al., Indirect photomeric detection of anions in capillary electrophoresis usingdyes as probes and electrolytes buffered with an isoelectric ampholyte, ELECTROPHOR, 21(7), 2000, pp. 1312-1319
Citations number
15
Categorie Soggetti
Chemistry & Analysis
Journal title
ELECTROPHORESIS
ISSN journal
01730835 → ACNP
Volume
21
Issue
7
Year of publication
2000
Pages
1312 - 1319
Database
ISI
SICI code
0173-0835(200004)21:7<1312:IPDOAI>2.0.ZU;2-3
Abstract
The use of highly absorbing anionic dyes as probes and isoelectric ampholyt es as buffers in background electrolytes (BGEs) combined with the use of a light emitting diode (LED) as a light source has been studied for ultrasens itive indirect photometric detection in capillary electrophoresis (CE). Pot ential dyes and buffers were evaluated based on characteristics relevant to indirect photometric detection principles, such as the electrophoretic mob ility of the probe dye, its solubility and adsorption behaviour, and the is oelectric point and buffering capacity of the ampholytic buffer. Two dyes, tartrazine and naphthol yellow S, and histidine as the ampholytic buffer, w ere selected for detailed investigation. Purification of the probes was vit al to avoid anionic impurities interfering with the detection. For the elec trolytes containing a purified probe (0.5 mM) and histidine as the isoelect ric buffer (p/7.7), hydroxypropylmethyl cellulose (similar to 0.05%) was ef fective in suppression of the electroosmotic flow (EOF). Analytical method performance characteristics were determined. For both probes, experimentall y determined mobilities were generally close to literature values, excellen t peak shapes and separation efficiencies of up to 298 000 theoretical plat es were obtained, and detection limits were generally at the sub-mu M level . For the naphthol yellow S-histidine BGE, linearity and reproducibility we re also evaluated, with excellent linearity being observed over a range of 5-500 mu M, and reproducibility (relative standard deviation, RSD) less tha n 1% for migration times and 2-8% for normalised peak areas. The approach d eveloped was applied successfully to several real samples including tap wat er, mineral waters, and beer.