Study On Discriminant Determination Methods For Edible Oil Oxidation

Edible oils are prone to oxidize and to decompose because of unsuitable storage temperature,exposure to light and oxygen during processing,transport and storage.The primary oxidation products that accumulate in oils are hydroperoxides,which may break down to form lower molecular weight compounds,for example alcohols,aldehydes and ketones.Long-term intake and consumption of these compounds have harmful impact on human health.Therefore,detecting technique of edible oil quality is significantly important.The oxidations of edible oils are measured according to the Chinese national standard.The edible oils oxidation status qualitative analysis were performed with electronic conductivity,electronic nose(E-nose),and dielectric spectroscopy technique by using cluster analysis(CA),principal component analysis(PCA)and linear discriminant analysis(LDA).The results are as follows:(1)The different edible oils(rapeseed oil,soybean oil,two types of peanut oil,flaxseed oil,sunflower oil,and corn oil)were chosen as materials.The electrolyte in the oil was extracted by water;electrical conductivity(EC)was determined in the aqueous phase;peroxide value and acid value were used as indexes to evaluate degree of oxidation.A relation model between the oxidation status of edible oil and EC was established.Results showed that the optimal detection conditions for conductivity were 80 s of oscillation time and 5:1 of water/oil ratio at room temperature.The established model classifies all oil samples based on different oxidation states.Oil samples with EC values less 6.9 ?s/cm were non-oxidized oils.Those in the range 6.9-8.9 ?s/cm were suspect oils that required further analysis.Samples with EC values exceeding 8.9 ?s/cm were oxidized oil.The total recognition rate can reach91.2%.This indicates that the conductivity method is feasible as a rapid detection of the oxidation degree of edible oils.(2)Seven edible oils were allocated randomly based on mass ratios of 1:1(two oils)and1:1:1(three oils)to produce 11 kinds of blend oils.Seventeen kinds of oil samples were prepared for rapid assessment of the oxidation degree in edible oils.The oxidation degree of the oil was evaluated according to the relevant national standards,and 20% of the anhydrous ethanol was used to elute the flavors in the oil.An electronic tongue(E-tongue)combined theCA,PCA and LDA was used to develop the calibration,and the calibration was validated.The recognition rates of PCA,CA and LDA were 93.4%,98.3% and 100%,respectively.The results showed that E-tongue combined with PCA,CA and LDA could distinguish different oxidation degree of edible oils.(3)The dielectric properties of rapeseed oil,soybean oil,peanut oil and corn oil were investigated in the frequency range of 1MHz-3GHz under normal temperature.The results showed that the relative dielectric constant has a good linear correlation with the peroxide value of edible oil in the range of 200 ~ 3500 MHz.PCA was used to classify edible oil according to its dielectric spectral characteristics.The recognition rate of calibration is up to94%.The dielectric characteristics of edible oil have a certain value of rapid detection and application.