Method For The Quality Analysis Of Flaxseed Oil Based On Fiber Raman Spectroscopy

Flaxseed oil is a natural vegetable oil with health benefits.The unique quality characteristics and high nutritional value render flaxseed oil more expensive than other common vegetable oils in the market.Hence,the adulteration of flaxseed oil and repeated use of fried flaxseed oil are unavoidable.It is to be noted that the lower quality of flaxseed oil inevitably affects the healthy development of the flaxseed oil industry and the physical and mental health of consumers.The quality of flaxseed oil has an important impact on human health.However,traditional methods have the disadvantages of complicated operational steps and costly and time-consuming processing,which fail to meet the market demand for testing.Therefore,the quality control of flaxseed oil requires a more rapid,simple,and accurate identification and analysis method.The current study applied Raman spectroscopy combined with“oil microscopy method”and machine learning to the quality determination of flaxseed oil.The contents and results of this study are primarily composed of the following two parts.1.Rapid identification of flaxseed oil using fiber-optic Raman spectroscopy combined with oil microscopy.The Raman spectral data of 125 samples containing flaxseed,canola,cottonseed,and adulterated oils were obtained using a portable fiber-optic Raman spectrometer,and then their spectra were analyzed.Finally,the following 11 characteristic bands were selected based on differential spectrum,principal component analysis（PCA）loading,and derivative analyses for the“oil microscopy method.”The Raman spectra of flaxseed and canola oils showed noticeable peak shifts,whereas the flaxseed and cottonseed oils exhibited relatively similar peak positions.Clear peak intensity differences were observed in the flaxseed,cottonseed,and canola oils mainly at 868,1022,1265,and 1655 cm^-1,with Raman shift intensities in the following order:I_{flaxseed oil}>I_{cottonseed oil}>I_{canola oil}.Similarly,the peak intensities of the flaxseed and adulterated oils also exhibited certain differences(at 868,1022,1265,and 1655 cm^-1),and the Raman shift intensity tended to decrease gradually with increasing canola and cottonseed oil content in flaxseed oil.Additionally,the results of Raman spectroscopy combined with the“oil microscopy method”exhibited large variations in the radar patterns of the flaxseed,canola,and cottonseed oils,whereas the radar patterns of the flaxseed and adulterated oils are highly similar.The Raman shift intensities increased with increasing canola and cottonseed oil concentrations,showing positive correlations at 868 and 1265 cm^-1,whereas a negative correlation was observed at 845 cm^-1.The results indicate that Raman spectroscopy in combination with oil microscopy more effectively reveals subtle differences in Raman shift intensity,serving as a more visual and comprehensive approach for differentiating quality variations between pure flaxseed oil,other oil species,and the adulterated oil.2.Quality analysis of fried flaxseed oil using fiber-optic Raman spectroscopy.Fiber-optic Raman spectroscopy was used to obtain spectral data for 80 flaxseed oil samples and fried flaxseed oil.Moreover,the fatty acid composition,peroxide value,and acid value of flaxseed oil and fried flaxseed oil samples were determined.The measurement results were then compared with the Raman intensity ratio,characteristic peak area,and other parameters.Finally,the Raman spectra were combined with PCA,ratio method,and partial least squares discriminant analysis（PLS-DA）to discriminate between flaxseed oil and fried flaxseed oil.The spectral analysis results showed that the Raman ratio intensities(I_{（1655/1438）},I_{（1265/1438）},and I_{（970/1438）})at the characteristic peaks related to fatty acids at 970,1265,and 1655 cm^-1did not vary significantly with temperature,number of frying cycles,or frying duration.The characteristic peak areas related to peroxide and acid values showed significant differences with temperature,number of frying cycles,and frying duration,which mostly appeared at 1022,1078,1265,1298,1438,and 1655 cm^-1.The model analysis results show that the ratio method can more effectively distinguish between flaxseed oil and fried flaxseed oil than PCA.The identification effect of PLS-DA model is the best,achieving 100%identification accuracy between flaxseed oil and fried flaxseed oil.