초록
Waste Sorghum Leaves (WSL) are considered as agricultural waste residue and a potential feedstock for biofuel production. This study investigates the optimum recovery of xylose and glucose from WSL using six Response Surface-based models. Furthermore, the optimum physico-chemical set-points for hydrogen production from these fermentable sugars are determined and a preliminary scale up is assessed. Models for saccharification were based on HCl (HCl-model), H<SUB>2</SUB>SO<SUB>4</SUB> (H<SUB>2</SUB>SO<SUB>4</SUB>-model) and HNO<SUB>3</SUB> (HNO<SUB>3</SUB>-model) pre-treatments subjected to input variables of acid concentration, heating time, solid to liquid ratio and acid exposure lag time in the ranges of 1-6%, 70-240 min, 30-50% and 0-24 h respectively. The models gave high coefficients of determination of up to 0.93. The HCl-model showed the highest recovery of xylose and glucose, with yields of 54.05 g/L and 15.98 g/L respectively, corresponding to 77% hemicellulose solubilisation and a shorter pre-treatment time in comparison to the other two acids. Optimization of physico-chemical variables for biohydrogen production gave set-points of 50% inoculum concentration, process time of 83 h, 11 min with an initial pH of 7. Process scale up in a 13 L bioreactor resulted in a peak hydrogen fraction of 43.75% and a volumetric hydrogen yield of 213.14 ml/g of fermentable sugar (FS), which is comprised of xylose and glucose. These findings illustrate the potential of sorghum leaves which are generally considered as agricultural waste for large scale production of fermentative hydrogen.