Assessment Of Animal Oils Quality By LF-NMR

This work was supported by the National Natural Science Foundation of China(NSFC31201365), Innovation Program of Shanghai Municipal Education Commission(11YZ109) and the key scientific and technological projects of Science and Technology Commission of Shanghai Municipality(11142200403). In this paper, firstly, the measurement parameters of low field NMR(LF-NMR) was established to assess the quality of solid oil. And, the correlation model between LF-NMR results and the physicochemical indexes was established during the frying process of different frying load. Then, with adulteration ratio of fried lard increased, the correlation model between LF-NMR results and the fatty acid composition was studied. At last, a method of discrimination of oil samples of different kinds and quality was developed according to LF-NMR relaxation characteristic using principal component analysis. The major results were as follows:1. The effects of the LF-NMR instrument parameters, detection temperature and volume on the stability and reproducibility of oils’ relaxometry characteristics results have been investigated with taken soybean oil、palm oil and lard as the samples. The results indicated that good stability and reproducibility can be guaranteed if choosing the following parameters, such as TR=2000ms, τ=250μs, EchoCount=6000, and before the measurement, a sample of 2.5mL should be stabilized at 50℃. Then the method was employed to discriminate 8 solid oil samples, 10 liquid hogwash oil samples. The result indicated that the proposed methodology is a useful tool to authenticate the hogwash oils, and the correct recognition rate was 75% and 100%, respectively, and the false positive rate hase been effectively avoided. Meanwhile, the difference of S21 peak area could reflect the adulterated ratio, which means that, the developed method has a good possibility to be used to assess the oil quality in the future.2. The correlation model between several physicochemical indexes(TPC concent, conjugated diene, conjugated trienes, peroxide value) and the LF-NMR T2 relaxation characteristics(the relaxation time T21, T22, T23, the corresponding peak area S21, S22, S23 and the single component relaxation time T2W) of lard was established and experimentally verified during the frying process of different frying load. The results indicated that as the frying process prolonged, the TPC content increased exponentially, conjugated diene and conjugated trienes increased linearly, peroxide value increased firstly and decreased later. And a good binomial relationship was obtained between S21 or T2 W and frying time. No obvious relationship was found between frying time and S22, S23, T21, T22 or T23. With the increase of the frying load, the TPC content and K270 increased signally, T2 W decreased markedly.Regression analysis showed that TPC content, conjugated diene, conjugated trienes, peroxide value had good relations with S21 and T2 W. The models could be used to predict the changes of physicochemical indexes by the LF-NMR characteristics of frying oil.3. With adulteration ratio of fried lard increased, the correlation model between LF-NMR results and the fatty acid composition of adulterated lard was established and experimentally verified. The results indicated that as the adulteration ratio of fried lard increased, the relaxation time T21 decreased signally, S21 and S22/S23 increased linearly, S23, T2 W, the content of C18:1+ C18:2 and relative unsaturation U/S decreased linearly. And a good binomial relationship was obtained between S22 and the adulteration ratio of fried lard. No obvious relationship was found between the adulteration ratio of fried lard and T22 or T23. Regression analysis showed that C16:1, C16:0, C18:2, C18:0, the content of C18:1+ C18:2 and relative unsaturation U/S had good relations with S21, T21, S22, T23 and T2 W.4. Principal component analysis(PCA) was used to analyze the LF-NMR(T2) relaxation characteristics(T21、T22、T23、S21、S22、S23) of 4 kinds of normal edible oil samples, fried oil samples, frying process and adulteration oil samples. The results indicated that application of PCA to the LF-NMR data allowed a good differentiation between edible soybean oil and cottonseed oil; while palm oil and lard could not be obviously distinguished. Different kinds of frying oil samples could be poperly differentiated, regardless of the frying degree. As the frying process prolonged, the distribution of the fried oil samples moved upper right based on the principal component scores. And the fried oil samples could be obviously devided into two groups: palm oil and lard, soybean oil and cottonseed oil. In addition, it was easy to distinguish adulterated lards from edible lard according to their corresponding principal component scores. With adulteration ratio of fried lard increased, higher distinguish ability was found and experimentally verified correctly.