Separation and Determination of Carbohydrates in Drinks by Ion Chromatography with a Self-Regenerating Suppressor and an Evaporative Light-Scattering Detector.

Jing Li,Meilan Chen,Yan Zhu
DOI: https://doi.org/10.1016/j.chroma.2007.02.080
IF: 4.601
2007-01-01
Journal of Chromatography A
Abstract:Analysis of glucose and other carbohydrates are often performed by use of normal phase HPLC methods with acetonitrile as major eluent coupled with evaporative light-scattering detector (ELSD) or by use of anion-exchange ion chromatography (IC) methods with NaOH as eluent coupled with pulsed amperimetric electrochemical detector. In this work, a novel method for the determination of carbohydrates by IC in conjunction with a self-regenerating suppressor and an ELSD detector was investigated. Three carbohydrates (glucose, fructose, and sucrose) were separated using a KOH eluent generator to avoid the effect of carbon dioxide absorption in the alkaline eluent. Due to the use of the suppressor, non-volatile components were removed and a low salt background (K+ approximately 0.070 microg/mL) can be obtained so the suppressed eluent could directly go into an ELSD detector without obvious interference of inorganic salts. After examining the changes in retention and resolution, an optimized method was established (for IC: using 32 mM KOH as the eluent at a flow rate of 1 mL/min; for ELSD: operated at 95 degrees C, 4.0 bar nitrogen with a gas flow rate of 2.0 L/min) and the linearity, reproducibility, and the limit of detection (LOD) for the three carbohydrates were further evaluated. Regression equations revealed acceptable linearity (correlation coefficients=0.994-0.998) across the working-standard range (100-1000 microg/mL for glucose and sucrose, 150-1000 microg/mL for fructose) and LODs of glucose, fructose, and sucrose were 93, 126, and 90 microg/mL, respectively. This method has successfully been applied to the determination of the three carbohydrates in carbonated cola drinks and fruit juices. The recoveries were between 95 and 113% (n=3) for different carbohydrates.
What problem does this paper attempt to address?