Applications of response surface methodology and artificial neural network for decolorization of distillery spent...

Ethanol production from sugarcane molasses yields large volume of highly colored spent wash as effluent. This color is imparted by the recalcitrant melanoidin pigment produced due to the Maillard reaction. In the present work, decolourization of melanoidin was carried out using activated carbon prepared from pepper stem (Piper nigrum). The interaction effect between parameters were studied by response surface methodology using central composite design and maximum decolourization of 75 % was obtained at pH 7.5, Melanoidin concentration of 32.5 mg l-1 with 1.63 g 100ml-1 of adsorbent for 2hr 75min. Artificial neural networks was also used to optimize the process parameters, giving 74 % decolourization for the same parameters. The Langmuir and Freundich isotherms were applied for describing the biosorption equilibrium. The process was represented by the Langmuir isotherm with a correlation coefficient of 0.94. The first-order, second-order models were implemented for demonstrating the biosorption mechanism and, as a result, Pseudo second order model kinetics fitted best to the experimental data. The estimated enthalpy change (DH) and entropy change (DS) of adsorption were 32.195 kJ mol-1 and 115.44 J mol-1 K which indicates that the adsorption of melanoidin was an endothermic process. Continuous adsorption studies were conducted under optimized condition. The breakthrough curve analysis was determined using the experimental data obtained from continuous adsorption. Continuous column studies gave a breakthrough at 182 mins and 176 ml. It was concluded that column packed with Piper nigrum based activated carbon can be used to remove color from distillery spent wash.