Optimization of Xylose Production by Alkaline Hydrolysis of Water Hyacinth Biomass Using Response Surface Methodology and Artificial Neural Network.

ABSTRACT

This paper entails a comprehensive study on production of xylose by alkali pretreatment of water hyacinth biomass. Artificial Neural Network (ANN) and Response Surface Methodology (RSM) were used for optimization in order to enhance xylose yield by alkali pretreatment using NaOH. In the present study, water hyacinth (Eichhornia crassipes), a fast growing, aquatic lignocellulosic weed was exploited for the production of xylose .The application of RSM in the study for xylose yield by alkali hydrolysis process utilizing optimization and modeling facilitated the analysis of interaction effects of the process variables on the response. Artificial Neural Network modeling was also studied to validate and compare the results obtained from RSM. The optimized experimental result of the influencing parameters are NaOH concentration of 3%, agitation speed of 130 rpm, treatment time of 11 min, treatment temperature of 60 oC and soaking time of 3h which produced a xylose yield of 84.56 mg/g of dried biomass. The yield dropped drastically on increasing the temperature above a certain point which may be due to the degradation of xylose. The analysis was done by Design Expert 9.0.3 software and Matlab, R2009a (The Math Works, Inc., MA, USA) for optimization by Response Surface Methodology and Artificial Neural Networking.