Open AccessWhole cell‐catalyzed transesterification of waste vegetable oil
Author(s) -
Xiao Man,
Obbard Jeffrey Philip
Publication year - 2010
Publication title -
gcb bioenergy
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.378
H-Index - 63
eISSN - 1757-1707
pISSN - 1757-1693
DOI - 10.1111/j.1757-1707.2010.01060.x
Subject(s) - response surface methodology , transesterification , biodiesel , central composite design , yield (engineering) , raw material , grease , coefficient of determination , fatty acid methyl ester , acid value , factorial experiment , chemistry , pulp and paper industry , mathematics , catalysis , chromatography , organic chemistry , materials science , biochemistry , statistics , engineering , metallurgy
Abstract Enzymatic transesterification of waste cooking oil, comprising fats, oil and grease (FOG), to produce fatty acid methyl esters (FAME) i.e. biodiesel, was investigated using a novel strain of the fungus Aspergillus niger , immobilized as a whole‐cell biocatalyst. Response surface methodology (RSM), with a five‐level‐three‐factor central composite rotatable design, was used to optimize the reaction and analyze the relationship of reaction variables and their coinfluent on the response i.e. FAME yield. Independent variables that affect the transesterification reaction include temperature, feedstock water content and enzyme amount. Using RSM, a second‐order polynomial equation was derived for FAME yield using multiple regression analysis. The second‐order polynomial regression model was highly significant ( P <0.001) in predicting the actual relationship between the response and the variables, where a linear relationship was apparent between observed and predicted values ( R 2 =0.9651). In addition, the predicted determination coefficient q 2 i.e. 0.7723, also proved that the model has a high predictive ability. The validation experiments, under optimized conditions, showed that the predicted value of maximum FAME yield (i.e. 91.3%) was in close agreement with the experimental value (i.e. 91.8%).