Use of Experimental Design and Artificial Neural Network in Optimization of Capillary Electrophoresis for the Determination of Nicotinic Acid and Nicotinamide in Food Compared with High-Performance Liquid Chromatography

Guangfen Mu,Feng Luan,Huitao Liu,Yuan Gao
DOI: https://doi.org/10.1007/s12161-012-9429-z
IF: 2.9
2012-01-01
Food Analytical Methods
Abstract:A simple capillary electrophoresis (CE) method for the analysis of nicotinic acid (NA) and nicotinamide (NAM) was developed. Important experimental factors were optimized with response surface approach. Statistical models were constructed by Box–Behnken design as well as artificial neural network using the three selected variables (buffer concentration, sodium dodecyl sulfate (SDS) concentration, and voltage). Optimum separation conditions obtained were 10 mM sodium borate/10 mM SDS buffer, pH 9.0, 25 kV applied voltage, and UV detection at 210 nm. Linear relationships between peak-area and the content of the analytes were obtained in the concentration range (1.0–100 μg mL−1), and the detection limit was less than 0.15 μg mL−1. The method was applied to the determination of NA and NAM in different kinds of samples with satisfactory separation in less than 5 min. In addition, a comparison with high-performance liquid chromatography demonstrated that the developed CE method was comparable with regard to linearity, sensitivity, precision, and accuracy, but the analysis time and reagent consumption for CE were decreased significantly.
What problem does this paper attempt to address?