Optimization Of The Culture Medium For Lipid Accumulation In Chlorella Vulgaris

A major bottleneck in microalgal biodiesel production is the serious contradiction between lipid content and biomass. Although some algal species have been known to accumulate up to 75% of their dry cell weight (dcw) in lipids, the low lipid yield, which depends on both lipid content and biomass and is the determinant of lipid productivity, lead to a limitation on the industrialization of microalgae-produced biodiesel.To solve this problem, several optimization methods including One-Factor-at-A-Time Design (OFAT), Box-Behnken Design (BBD), Response Surface Methodology (RMS) and Artificial Neural Network-Genetic Algorithm (ANN-GA) were employed in this research.First, determination using Spectrofluorometer was combined with gravimetric analysis for a rapid quantitative measurement of lipid yield of Chlorella vulgaris. Then the culture medium was optimized by OFAT experiment as the first step of the optimization for a high lipid yield. From OFAT experiment, an optimized TAP medium was obtained contain a nitrogen (N), phosphorous (P) and sulfur (S) concentration of 6.13,5.23 and 4.04mg/l respectively. Second, BBD experiment was made based on the result of OFAT. As a basic data, The result of BBD was then used in the construction of RMS model and ANN-GA model. Further, predictions were made by these two models. RMS model reported a best lipid yield of 112.9mg/l would be obtained in an optimum culture medium contains a N, P and S concentration of 6.33 mg/1,4.19mg/l and 4.66mg/l respectively. While, the report from ANN-GA model is that culture medium contains a N, P and S concentration of 6.45mg/l, 4.23mg/l and 5.12mg/l respectively would lead to a highest lipid yield of 124.8mg/l. Finally, the predictions from these two models were investigated by experiment, and a lipid yield of 111.60 mg/1 and 104.10 mg/1 were obtained in media optimized by RMS and ANN-GA respectively, while only 12.88mg/l was obtained in a non-optimized TAP used as control. This result suggests that both RMS and ANN-GA have leaded a lipid yield boom and can be employed in the optimization of culture medium for lipid production.