Abstract: Response surface methodology (RSM) was applied to optimize the extraction of loquat seed oil with supercritical carbon dioxide. A quadratic polynomial mathematical model was set up and validated describing the effects of pressure, temperature and time on oil yield. The fatty acid profile of the oil was analyzed by GC-MS. The optimal extraction conditions were 44.16 MPa, 62.02 ℃ and 21 min. Under these conditions, the predicted and actual oil yields showed good consistency, suggesting that the model can fit the experimental data well. A total of 11 fatty acids were identified in the oil, with a relative content of 94.37% in total fatty acids. Linoleic acid (C18:2, 45.91%) was the most predominant fatty acid in the oil, followed by palmitic (C16:0, 26.01%), oleic (C18:1, 6.19%), linolenic (C18:3, 5.01%), stearic acid (C18:0, 3.64%) and other 6 fatty acids (total about 7.61%).

Key words: loquat seed, supercritical carbon dioxide, response surface methodology, optimization, fatty acid