Valorization of corncobs into cephalosporin C through predictive modeling of the dilute acid hydrolysis and detoxification process

Corncobs (CC) are an abundant lignocellulosic residue with considerable potential as a low-cost source of xylose for microbial production of high-value biochemicals. This study aimed to develop an integrated biorefinery process that produces xylose from CC and applies it as a fermentation substrate for the production of cephalosporin C. The dilute acid hydrolysis of CC for xylose recovery was modelled and optimized by using response surface methodology (RSM) and artificial neural network (ANN), with reaction time, H 2 SO 4 concentration, and particle size as the major process variables. The dataset of 20 experimental runs generated through the RSM design was used for RSM and ANN modeling. The xylose concentrations of corncobs hydrolysate (CCH) produced under the optimal conditions derived from RSM (38.7 min, 0.69% H 2 SO 4 , and a CC particle size of 290 μm) and ANN (50 min, 0.75% H 2 SO 4 , and a CC particle size of 234 μm) were found to be 26.88 g/L and 29.22 g/L, respectively. Within the investigated design space, the ANN model showed lower prediction errors than the quadratic RSM model and was therefore selected for hydrolysate production. Fermentation inhibitors in the resulting CC hydrolysate, including acetic acid, formic acid, 5-hydroxymethylfurfural, and furfural, were subsequently removed by liquid–liquid extraction using n-butyl acetate, with removal efficiencies of 74.92%, 100%, 100%, and 100%, respectively. Finally, 1.14 g/L of cephalosporin C was produced in the detoxified CCH-based medium (about 1.2-fold higher than control) using Acremonium chrysogenum . This study demonstrated that CC can be utilized as an alternative fermentation substrate for the production of cephalosporin C by integrating hydrolysis optimization using an ANN, hydrolysate detoxification, and microbial fermentation.