Abstract: One-factor-at-a-time method and response surface methodology (RSM) were used in combination to optimize the extraction of arabinoxylans from hull-less barley bran with aqueous sodium hydroxide solution. The independent variables investigated were solid to solvent ratio, temperature, extraction time and sodium hydroxide concentration, while the response variable was the yield of arabinoxylans. Each independent variable was tested at three levels. The optimum conditions obtained were as follows: solid to solvent ratio, 1:25 (g/mL); extraction temperature, 55 ℃; extraction time, 3 h; and sodium hydroxide concentration, 15 g/L. Under the optimized conditions, the yield of arabinoxylans was 14.31%, which agreed well with the predicted value of 14.27%. Thus, the optimized process was stable and feasible. Analysis of purity and structural properties by high-performance gel permeation chromatography (HPGPC), UV absorption spectroscopy, IR spectroscopy, high-performance anion-exchange chromatography (HPAEC) showed that the arabinoxylans were homogeneous and contained a small amount of protein but no nucleic acid and they were mainly composed of four monosaccharide units: arabinose, galactose, glucose and xylose with a molar ratio of 11.61:1:3.62:18 ( molar ratio of arabinose to xylose = 0.645), which were linked via α and β-glycosidic bonds.

Key words: response surface methodology, hull-less barley bran, arabinoxylan, structure