基于多维拉曼光谱的乳粉表征及判别分析

来源:用户上传      作者:顾颖 赵姝彤 熊蓝萍 季娜 洪浩 张正勇

　　摘要：针对乳粉质量管控需要，发展具有潜在鉴别应用价值的多维拉曼光谱表征与量化判别分析方法。以牛、羊乳粉为对象，获取样品的传统拉曼光谱信息，并运用二维相关光谱法构建三维拉曼光谱，从多维拉曼光谱刻画了研究对象的特征谱图信息。运用相似度分析算法，定量化分析了牛、羊乳粉样品间差异情况，提出了研究对象特征谱图分析思路。研究结果显示，在谱图分析层面，牛、羊乳粉的传统拉曼光谱、二维相关拉曼光谱含有丰富的样品化学特征信息及直观视觉特点；在相似度量化分析层面，各品牌乳粉同品种样品内相似性较高，不同品种样品间差异性较大，以及二维相关拉曼光谱具有更高的样品分辨率。据此，可实现牛、羊乳粉的分类鉴别，具有分析速度快、可量化分析等多种优势。
　　关键词：乳粉；拉曼光谱；二维相关；相似度
　　中图分类号：TS252.7 文献标识码：A DOI：10.16465/j.gste.cn431252ts.20220325
　　Characterization and Discriminant Analysis of Milk Powder Based on Multidimensional Raman Spectroscopy
　　Gu Ying， Zhao Shutong， Xiong Lanping， Ji Na， Hong Hao， Zhang Zhengyong
　　（ School of Management Science and Engineering， Nanjing University of Finance and Economics， Nanjing， Jiangsu 210023 ）
　　Abstract： To meet the needs of milk powder quality control， multidimensional Raman spectroscopy characterization and quantitative discriminant analysis methods with potential identification of application value were developed in this work. Taking bovine and goat milk powder as the research object， the traditional Raman spectral information of the sample was obtained， and the three-dimensional Raman spectroscopy was constructed by using the two-dimensional correlation spectrum method. The feature spectral information of the research object was characterized from the multidimensional Raman spectroscopy. The differences between bovine and goat milk powder samples were quantitatively analyzed by using the similarity analysis algorithm， and the idea of characteristic spectral analysis of the research object was put forward. The results showed that at the level of spectral analysis， the traditional Raman spectra and two-dimensional correlation Raman spectra of bovine and goat milk powder contained rich sample chemical characteristic information and intuitive visual characteristics. At the level of similarity quantitative analysis， the similarities of milk powder of different brands within the same variety of samples are high， the differences between different varieties are large， and two-dimensional correlation Raman spectroscopy has higher sample resolution. Therefore， the classification and identification of bovine and goat milk powder can be realized， which has many advantages， such as fast analysis speed， quantifiable analysis.
　　Key words： milk powder， Raman spectroscopy， two-dimensional correlation， similarity
　　乳制品作橄费者日常食品购买的重要组成部分，尤其是满足婴幼儿营养需求的重要食品来源，其质量安全问题一直深受我国市场主管部门和大众的关注，然而近年来，国内外乳制品相关质量安全问题仍时见报道，为此，国家“十四五规划纲要”明确提出要加强和改进食品安全监管制度，完善食品安全法律法规和标准体系，深入实施食品安全战略，着力提高食品安全检测技术[1]。针对乳制品质量安全风险，目前管控和研究策略主要包括：① 建立健全国家乳制品相关产品标准、检测标准；② 研究开发各类先进检测方法。如李涛等[2]研究建立了一种基于超高效液相色谱-质谱的乳粉中16种喹诺酮类药物检测方法；Mendes等[3]运用拉曼光谱结合主成分分析、偏最小二乘回归法进行了生牛奶中乳清的检测；Li等[4]研究构建了一种基于表面增强拉曼光谱的侧向流免疫传感器，可实现牛乳中黏菌素的快速识别。

nlc202208251108

转载注明来源:https://www.xzbu.com/1/view-15438340.htm