APPLICATION OF ARTIFICIAL NEURAL NETWORK FOR THE MAXIMIZATION OF ISOAMYL ISO VALERATE ESTER YIELDS IN N-HEXANEG.V. CHOWDARY1, M.N. RAMESH' AND S.G. PRAPULLA
Synthesis of isoamyl isovalerate, an apple flavour ester, in n-hexane using immobilized lipase from Rhizomucor miehei has been carried out. Artificial Neural Network based model has been used to arrive at the ideal combination of parameters for maximization of yields. Considering the least possible percentage error of the trained model, the effect of the parameters has been studied for the maximization of the process. The statistical significance of the model confirmed the goodness of the fit. The maximum esterificat ion obtained through the network trained model was 87% at alcohol concentration of 0.45 M, isovaleric acid of 0.35 M, temperature of 40°C for a reaction time of 120 h at an E/S ratio of 45 g/mol. The experimental value for the same conditions was 90%.
Enter your contact information below to receive full paper.