Acid Hydrolysis of Damaged Wheat Grains: Modeling the Formation of Reducing Sugars by a Neural Network Approach

Abstract

In this work, hydrochloric acid, phosphoric acid, nitric acid and sulphuric acid were screened for their relative potential for hydrolysis of damaged wheat grains. Inhibitor concentration and reducing sugar were taken as performance parameters. Concentration of four inhibitors namely, furfural, 5-hydroxymethyl furfural, acetic acid and formic acid were measured by high pressure liquid chromatography. Initial screening demonstrated that HCl was the most potent acid for hydrolysis. Subsequent experiments with different substrate (10 %, 15 %, 20 % w/w) and acid concentrations (1%, 3%, 5% w/v) were carried out to identify suitable hydrolysis condition for maximum conversion of substrate to reducing sugars (RS). Results showed that 3% HCl with 15 % substrate concentration produced highest RS (116.29 mg/mL) after 45 min of hydrolysis. Early formation of inhibitors was observed at 5% HCl which diminished the RS formation. Although hydrolysis with 1% HCl yielded RS comparable to that of 3% HCl concentration, the time of hydrolysis was higher. Artificial neural network (ANN) and second-order models were applied to the experimental data to map the variation in RS with hydrolysis. ANN performed well in predicting RS after hydrolysis with good accuracy (R2 = 0.939). The obtained model can be used to predict the variation in RS over a wide range of process variables thereby making the selection of hydrolysis process parameters easier.