Abstract:
The nutritional and medical use of palm kernel oil is in high demand and has ignited the search for extraction of the oil at the optimum. The study investigated the use of ethanol as solvent, constituting one of the independent factors used in the research, which are; Sample Weight, Solvent Volume, Extraction Time to extract oil. An experimental design was employed to optimize the oil extraction in the Box-Behnken design, and the physiochemical characteristics of the extracted oil were obtained. 17 experimental runs were generated through Response Surface Methodology (RSM) and Artificial Neural Networks (ANN) methods. The results reflected that low solvent volume and high extraction time are required for an increase in the oil yield percentage for extraction of 40g weighted sample, 175 ml solvent volume, and 50 min extraction time. RSM predicted that the optimal yield of palm kernel seed to be 36.67 % while ANN predicted a yield of 37.693 % at the optimum sample weight of 40 g, the volume of 150 ml, and extraction time of 60 min. The results demonstrated that RSM and ANN software is effective for optimization oil extracted by ethanol, with a higher percentage of yields from ANN. The properties of the oil revealed yellowish-brown oil at room temperature, a fat content of 42%, and an Iodine value of 87.85(I2g/100g oil), the Saponification value of 140.125 (mg KOH/g Oil), with low acid and high FFA. Hence the extracted oil can be used for medical and cosmetic purposes.
|