Assessing the authenticity about the botanical and geographical origins of food is an important content for food safety research .Amino acids are the most important nutrients of honey .The types and contents of amino acids are different in various honey samples .Thus they can be used as one of the important parameters to discriminate the honey variety and quality .In this article ,amino acids in honey were first derived with formaldehyde and acetyl acetone solution .In the following step ,three‐di‐mensional fluorescence spectrums combing with multidimensional pattern recognition methods were used to distinguish the kinds of honey .Five kinds of honey (total 150 honey samples) from different botanicals were studied in this research .Before fluores‐cence detection ,the effect of the amount of derivation reagent ,the time of reaction ,temperature and pH to the derivation pro‐gress of honey samples were first studied .Research showed that the fluorescence intensity of derivatives of honey was the stron‐gest when the amount of derivation reagent was 4.0 mL ,the time of reaction being 2 h ,pH being 4 and the temperature being 100 ℃ .The derivatives of honey were then scanned with three‐dimensional fluorescence spectrometry .The collection of fluores‐cence intensity values occurred within excitation‐emission ranges of 300~500 and 380~580 nm .A 150 × 41 × 101 cube matrix data sets can be acquired .The three‐dimensional fluorescence data sets were decomposed with multilinear pattern recognition methods , such as multilinear principal components analysis (M‐PCA ) , self‐weight alternative trilinear decomposition (SWATLD) and multilinear partial least squares discriminant analysis (N‐PLS‐DA) methods .All of these multilinear pattern recognition methods showed the clustering tendency for five different kinds of honey .Compared with the other two methods , N‐PLS‐DA got more accurate and reliable classification results because it made full use of all the fluorescence information of the derivative honey samples .Its total recognition rate reached 88% .The result is acceptable for the complexity of the honey sam‐ples .It showed this method could be applied to identify the varieties of honey .Compared with the chromatographic analysis method ,this method is relatively simpler and more sensitivity .It avoided the chromatographic separation and reduced the con‐sumption of organic solvent .Thus it can be regarded as a kind of relatively green honey classification method .This research will provide a new idea to directly fluorescence analyze for no or weak fluorescence natural substances .%确认食品的种类和产地等信息是食品质量与安全研究的重要内容。基于蜂蜜中的氨基酸种类和含量的不同，应用甲醛和乙酰丙酮与蜂蜜中氨基酸发生荧光衍生反应对源自不同花源的蜂蜜进行种类辨别研究。实验首先研究了衍生剂用量、时间、温度及pH值对蜂蜜中氨基酸荧光衍生反应的影响。然后在激发光波长范围为300～500 nm和发射光波长范围380～580 nm下对衍生后的蜂蜜进行三维荧光检测，每个蜂蜜样品经检测后得到一个三维荧光光谱矩阵数据，五种蜂蜜共150个样品最后得到一个三维立方数据。将检测获得的三维荧光数据结合多维主成分分析（M‐PCA）、自加权交替三线性分解法（SWATLD）及多维偏最小二乘辨别分析（N‐PLS‐DA ）等多维模式识别方法，进行数据处理并获取五种蜂蜜的识别信息。三种多维模式识别方法分别都展示出不同种类蜂蜜的聚类趋势，相比较而言，由于N‐PLS‐DA充分利用了衍生蜂蜜荧光的所有信息，所以取得了更加准确和可靠的分类结果，对五种蜂蜜的总识别率为88％。研究结果显示基于氨基酸荧光衍生的多维模式识别方法可以用于蜂蜜种类的识别研究。该研究为无荧光或弱荧光自然物质的直接荧光辨别分析提供了一种新思路。
展开▼
