五常稻花香米的^1H-NMR波谱分析及掺假鉴别

^1H- NMR Spectroscopy and authenticity identification for Wuchang Daohuaxiang rice

结合核磁共振氢谱（^1H NMR）技术和多变量统计分析方法,研究五常稻花香米的掺假鉴别方法。采集稻花香米与掺假样品的^1H NMR谱,分析它们的化合物组成模式的差异,分别建立基于1H NMR谱和基于20种主要成分半定量数据的掺假鉴别模型。结果表明：五常稻花香米与掺假大米在α-葡萄糖、淀粉、蔗糖、甜菜碱、β-葡萄糖、胆碱、天冬酰胺、天冬氨酸、琥珀酸、谷氨酸、乙酸、γ-氨基丁酸、丙氨酸、苏氨酸、异亮氨酸和缬氨酸等16种主要成分上存在显著差异（p〈0.05）;基于-1H NMR谱的模型预测准确率为83.5%,而基于20种半定量化合物的模型预测准确率为70.1%。

Authenticity identification of Daohuaxiang rice was studied by combing of proton nuclear magnetic resonance（-1H NMR） and multivariate analysis in this paper.-1H NMR spectra were acquired for Daohuaxiang and adulteration samples. The compositional patterns of Daohuaxiang and adulteration samples were analyzed using their -1H NMR spectra.Then two Daohuaxiang identification models were built based on -1H NMR spectroscopy and semi- quantitative information of 20 compounds of the rice samples respectively. It turned out that there were 16 significantly different（ p〈0.05） compounds between Daohuaxiang and adulteration rice including alpha- glucose,starch,sucrose,betaine,beta- glucose,choline,asparagine,aspartate,succinate,glutamate,acetate,gamma aminobutyric acid（ GABA）,alanine,threonine,isoleucine,and valine. The predictive accuracies were 83.5% and70.1% for -1H NMR spectroscopy- based identification model and semi- quantitative information- based identification model respectively.