Research On Method For Detecting Phospholipid Content In Soybean Oil Based On Multiple Modified Enzyme Electrodes

Degumming is an important step in the refining process of soybean oil,in order to take off the phospholipids and other colloids in the soybean crude oil,which lay the foundation for subsequent decolorization and deodorization.Therefore,the phospholipid content in soybean oil needs to be continuously detected during the process.The traditional methods of detecting phospholipid are all based on chemical analysis in laboratory,slow detection speed,high detecting costs,much more affected by operators and experimental environment,pollute environment,especially not suitable to on-line detection and control.Whereas electrochemical analysis technology possessed of many advantages,such as fast detecting speed,higher measuring accuracy,not destructing samples,simple preprocessor and easily implementing on-line detection,and so on.Thus,it is applied widely in many fields for detection and analysis in recent years.The paper presents a new method for detecting phospholipid content in soybean oil based on electrochemical analysis.First,40 mixed soybean oil samples with different content of phospholipids are prepared,and the standard value of phospholipid contents were precisely detected by molybdenum blue colorimetry.There is an issue that it’s no electron transfer in the process of phospholipase hydrolysis.The paper develops the electrochemical sensor modified by multienzyme in order to obtain electrochemical signals,which combined with electrochemical workstations,the soybean oil samples were scanned by cyclic voltammetry,and the current-potential voltammetry curves are obtained.And the paper makes a thorough study of electrochemical data pretreatment,the establishment of correction model and regression prediction model,and so on.Firstly,the partial least squares correction model is established based on the original electrochemical data,and abnormal samples are removed according to the deviation of the predicted values and the actual value.Then,in order to improve the prediction accuracy of model,the method of denoising the electrochemical data for the phospholipid content in soybean oil based on the Savitzky-Golay smoothing filter and dbN series wavelet is researched.Through the comparison and analysis,it is found that the denoise effect based on db6 wavelet and three-layer decomposition is best.Finally,four methods are used respectively to set up regression model between the electrochemical data and phospholipid concentration,namely that the linear fitting based on reductive peak ciirrent and phospholipid concentration,principal component regression model and partial least squares regression model and support vector machine regression model.Modeling effect for the 4 methods are analyzed and compared,which show that the prediction accuracy of the model with SVR based on radial basis kernel function is the best.The Coefficient of Determination(R~2)is 0.9987,Root-Mean-Square Error of Prediction(RMSEP)is 0.2889 and Relative Standard Deviation(RSD)is 2.55%,respectively,all of the indices achieved the actual testing requirements.The study of this paper demonstrated that the electrochemical sensor modified multiple enzyme for the detection of phospholipid in soybean oil is feasible,which established foundation for developing special instruments of electrochemical analysis and further realizing real time on-line detection and control for phospholipids in soybean oil.