Study On The Identification Of Trace Vegetable Oils Based On Gas Chromatography Combined With Chemometrics

The identification of trace oil evidence at the scene of arson,intentional injury,and other cases is of great significance for determining the nature of the case and restoring the original scene.The study of this issue is also a key and hot issue that urgently needs to be solved in the field of forensic science.In this paper,the identification of trace vegetable oils was studied,and eight identification models of trace vegetable oils were constructed using gas chromatography and chemometrics.Firstly,establish a method for the determination of fatty acids in eight types of vegetable oils.Secondly,analyze the changes in fatty acid content in eight trace vegetable oils under different environments.Finally,five main fatty acids(hexadecane acid,octadecanoic acid,oleic acid,linoleic acid,linolenic acid)were selected as identification indicators,and a variety of plant oil identification models were constructed by combining chemometrics methods.The pre-treatment and detection methods selected by the research institute are efficient and convenient,with accurate recognition models.They effectively solve the problem of difficult identification of plant oil types after changes in fatty acid content in plant oil,and provide technical support and scientific guidance for related cases involving trace plant oil evidence identification.Mainly including the following studies:(1)The analysis method of fatty acids was established by gas chromatography,and the three factors of alkali concentration,alkali volume and reaction time in the conditions of fatty acid methyl ester were optimized.The results showed that the optimal conditions for methylation were using an alkali concentration of 0.4 mol/L,an alkali volume of 10m L,and a reaction time of 30 minutes.Under this condition,the corresponding fatty acids measured have a good correlation coefficient within their linear range,with R~2>0.9931 and a recovery rate between 93.48%and 101.01%.The method was applied to the analysis of fatty acids in eight plant oils,and the results showed that the separation effect of fatty acids in the eight plant oils was good,which can be used for the detection of fatty acids in conventional plant oils.However,the analysis of trace amounts of fatty acids in plant oils under different environments still needs further research.(2)The optimal extraction time for eight types of vegetable oils on four carriers was determined based on the content of fatty acids,and the changes in fatty acid composition and content were specifically investigated when eight trace amounts of vegetable oils were placed in different environments for 1 day,3 d,7 d,14 d,30 d,45 d,60d,and 90 d,respectively.The results showed that the optimal extraction time for eight plant oils on four carriers using n-hexane was 15-20 minutes for iron and plastic sheets,and 25-30 minutes for cloth and paper sheets,respectively.The change rule of fatty acid content in eight kinds of micro vegetable oils under different environments is that the content of saturated fatty acid is on the rise,and the content of polyunsaturated fatty acid is on the decline.And the higher the temperature is,the greater the range of change is.At the same time,by comparing the content of oleic acid,linoleic acid and linolenic acid,it can be found that the oxidation decomposition rate of unsaturated fat acid is related to degree of unsaturation,and the higher the degree of unsaturation is,the faster the oxidation rate is.(3)Based on five kinds of fatty acids(hexadecane acid,octadecanoic acid,oleic acid,linoleic acid,linolenic acid)and hierarchical clustering analysis,the clustering effect of eight kinds of trace vegetable oils in different environments was investigated,so as to screen the sample range.The results showed that the accurate identification of eight trace vegetable oils(camellia oleifera seed oil,rapeseed oil,corn oil,peanut oil,sesame oil,soybean oil,sunflower seed oil)and flaxseed oil(flaxseed oil)within 90 d at low temperature(4℃),60 d at normal temperature(25℃),45 d at high temperature(38℃)and 30 d at high temperature could be achieved.The degree of difficulty and recognition accuracy of different recognition models are different.Based on comprehensive analysis,the random forest algorithm has greater advantages,with a recognition accuracy of98.23%.The model parameters are simple to set,which can not only achieve more accurate recognition of eight trace vegetable oils,but also assess the importance of each fatty acid to the classification results.The overall recognition accuracy of Fischer discriminative model,back-propagation neural network,short-term memory neural network and convolution neural network model is not as good as that of random forest,but the recognition accuracy of specific vegetable oil is close to or even higher than that of random forest model.Therefore,in police practice,the extracted detection data of unknown vegetable oil samples can be substituted into multiple models in the paper at the same time,Based on the recognition results of multiple models,comprehensively determine the category of vegetable oil.