-
对21#欧贝拉蛋黄酥(龙海市吉香园食品有限公司)食品包装袋样本进行10次重复实验,所得差分喇曼光谱图中出峰峰数、峰位相同,峰形基本一致(见图 1),说明该方法的重现性良好。
-
根据食品塑料包装袋样本的差分喇曼光谱图的特征峰,可将其分为两大类,第Ⅰ类样本在1059cm-1, 1125cm-1, 1289cm-1, 1429cm-1处存在特征峰,其主要成分为聚乙烯(polyethylene, PE)(见图 2),共8个样本; 第Ⅱ类在809cm-1, 841cm-1, 971cm-1, 1149cm-1, 1322cm-1, 1451cm-1处存在特征峰[13],其主要成分为聚丙烯(polypropylene, PP)(见图 3),共38个样本,分类表见表 1。
Table 1. Classification of 46 samples
categorymain
componentsample number Ⅰ PE 5#, 11#, 24#, 33#, 34#, 38#, 39#, 45# Ⅱ PP 1#, 2#, 3#, 4#, 6#, 7#, 8#, 9#, 10#, 12#, 13#,
14#, 15#, 16#, 17#, 18#, 19#, 20#, 21#, 22#,
23#, 25#, 26#, 27#, 28#, 29#, 30#, 31#, 32#,
35#, 36#, 37#, 40#, 41#, 42#, 43#, 44#对第Ⅰ类样本又可以根据其在1085cm-1附近有无特征峰进一步分类:a类在1085cm-1处没有特征峰(见图 4),共4个样本; b类在1085cm-1附近有特征峰(见图 5),共4个样本。1085cm-1是填料CaCO3的喇曼特征峰,CaCO3的喇曼特征峰有153cm-1, 280cm-1, 711cm-1, 1085cm-1。对第Ⅱ类样本根据其在458cm-1附近有无特征峰可将其分为两类:c类在458cm-1附近没有特征峰(见图 6),共14个样本; d类在458cm-1附近有特征峰(见图 7),共24个样本。458cm-1是填料BaSO4的喇曼特征峰,BaSO4的喇曼特征峰有458cm-1, 616cm-1, 1084cm-1, 1139cm-1[14]。
-
选取同一公司来源的样本的差分喇曼光谱图进行分析,发现这些样品的特征峰位置基本相近,说明同一来源的食品塑料包装袋的基本成分大致相同。但特征峰的峰数存在一定差异,说明同一品牌、不同系列的食品塑料包装袋样本所用原料成分可能有差异。如27#来源于玛氏食品有限公司的德芙巧克力包装袋样本(见图 8)在扫描范围内有13个特征峰,而12#来源于玛氏食品有限公司的脆香米牛奶巧克力包装袋样本(见图 9)在扫描范围内有14个特征峰,由此可将同一品牌不同系列的样本区分开来。
-
通过分析3#、30#、35#、42#样本的差分喇曼光谱图,发现35#小浣熊干脆面(见图 10)在扫描范围内有9个特征峰,其它3个样本特征峰峰数均为14。3#扬子江法式烤芙酥(见图 11)在1358cm-1附近有特征峰,其它3个样品在此没有特征峰。42#男梅牌梅子糖(见图 12)在2921cm-1附近有特征峰,其它3个样品在此没有特征峰。30#好丽友派(见图 13)在902cm-1附近出现特征峰而其它3个样本没有,因此仍然可以区分开来。
-
系统聚类又称为凝聚性层次聚类,主要是将每个个体先视为一类,根据个体之间的距离和相似性原则进行合并,直到所有个体都归为一个簇[15-17],该方法广泛应用于现代科学技术和人文社科等各个领域[18-22]。
采用组间联接法作为类间亲疏程度的度量方法,平方欧氏距离度量个体间距离,进行系统聚类分析,得到46组样品的树状聚类图(见图 14)。
根据图 14可知,当并类距离最小时,可将样品分为37类;当并类距离为3时,可将样品分为17类;当并类距离为5时,可将样品分为11类;阈值达到25时,所有样品并为一类。为确定分多少类是科学合理的,对聚类结果进行相关性分析。
针对不同层次的分类结果,分别计算各个类别的样本之间的显著性和Pearson相关系数,以聚类结果中的第1类33#、第2类34#、第4类11#、第5类5#和24#为例,结果如表 2所示。**表示在0.01水平(双侧)上显著相关。
Table 2. Correlation analysis results
5# 11# 24# 33# 34# 5# Pearson
correlation
coefficient1 0.848** 0.963** 0.482** 0.537** significance
(bilateral)— 0.000 0.000 0.000 0.000 11# Pearson
correlation
coefficient0.848** 1 0.847** 0.510** 0.526** significance
(bilateral)0.000 — 0.000 0.000 0.000 24# Pearson
correlation
coefficient0.963** 0.847** 1 0.519** 0.560** significance
(bilateral)0.000 0.000 — 0.000 0.000 33# Pearson
correlation
coefficient0.482** 0.510** 0.519** 1 0.953** significance
(bilateral)0.000 0.000 0.000 — 0.000 34# Pearson
correlation
coefficient0.537** 0.526** 0.560** 0.953** 1 significance
(bilateral)0.000 0.000 0.000 0.000 — 由表 2可知,各变量的显著性值均为0.000,表明参量显著性良好。33#样品与34#样品、5#样品与24#样品的Pearson相关系数分别为0.953, 0.963, 表明样品之间的相关性很高, 而11#样品与33#样品、11#样品与34#样品、11#样品与24#样品的Pearson相关系数分为别0.519, 0.560, 0.847,表明相关性较低。科学准确的聚类结果应是簇内具有较高的相关性。因此,通过不同组间距离下各变量的Pearson系数和相关性的差异,可以得出:当并类距离为3时,样品分为17类是最科学合理的。
基于差分喇曼光谱快速鉴别食品塑料包装袋
Rapid identification of food plastic packaging bags based on differential Raman spectroscopy
-
摘要: 为了快速鉴别案件现场残留的食品塑料包装袋样本,采用差分喇曼光谱法结合系统聚类分析方法,对46个不同来源、不同系列的食品塑料包装袋样本进行了分析检验。结果表明, 依据差分喇曼光谱图中特征峰的不同,可以有效区分不同来源的食品塑料包装袋样本;同时结合系统聚类分析法,可将46个样本分为17类。该方法不破坏检材、重现性好,实验结果较为理想,为微量物证检验提供了一定的参考依据。Abstract: In order to quickly identify the material evidence of food plastic packaging bags on the scene, 46 samples from different sources and different series of food plastic packaging bags were analyzed and tested by differential Raman spectroscopy and system cluster analysis. The results show that the samples of different sources of food plastic packaging bags can be effectively distinguished according to the different characteristic peaks in the differential Raman spectrum. And 46 samples can be divided into 17 categories by combining with the system cluster analysis. This method does not damage the sample, and the experimental results are ideal, which provides a certain reference for the examination of trace evidence.
-
Key words:
- spectroscopy /
- differential Raman /
- food plastic packaging /
- cluster analysis
-
Table 1. Classification of 46 samples
categorymain
componentsample number Ⅰ PE 5#, 11#, 24#, 33#, 34#, 38#, 39#, 45# Ⅱ PP 1#, 2#, 3#, 4#, 6#, 7#, 8#, 9#, 10#, 12#, 13#,
14#, 15#, 16#, 17#, 18#, 19#, 20#, 21#, 22#,
23#, 25#, 26#, 27#, 28#, 29#, 30#, 31#, 32#,
35#, 36#, 37#, 40#, 41#, 42#, 43#, 44#Table 2. Correlation analysis results
5# 11# 24# 33# 34# 5# Pearson
correlation
coefficient1 0.848** 0.963** 0.482** 0.537** significance
(bilateral)— 0.000 0.000 0.000 0.000 11# Pearson
correlation
coefficient0.848** 1 0.847** 0.510** 0.526** significance
(bilateral)0.000 — 0.000 0.000 0.000 24# Pearson
correlation
coefficient0.963** 0.847** 1 0.519** 0.560** significance
(bilateral)0.000 0.000 — 0.000 0.000 33# Pearson
correlation
coefficient0.482** 0.510** 0.519** 1 0.953** significance
(bilateral)0.000 0.000 0.000 — 0.000 34# Pearson
correlation
coefficient0.537** 0.526** 0.560** 0.953** 1 significance
(bilateral)0.000 0.000 0.000 0.000 — -
[1] ZHANG L, YIN J X. Rapid identification of plastic packaging materials for food by infrared spectroscopy [J]. Packaging and Food Machinery, 2016, 34(6): 65-67(in Chinese). [2] JIANG H, JU Ch Y, ZHANG B Y, et al. Analysis of various components of plastic bags by infrared spectroscopy combined with X-ray fluorescence spectroscopy [J]. Chemical Research and Application, 2019, 31(3): 391-396(in Chinese). [3] IBARRA V G, de QUIRÓS A R B, LOSADA P P, et al. Identification of intentionally and non-intentionally added substances in plastic packaging materials and their migration into food products[J]. Analytical and Bioanalytical Chemistry, 2018, 410(16): 3789-3803. [4] XIE P. Study on the migration detection of plasticizers in plastic packaging materials by liquid chromatography tandem mass spectrometry [D].Changchun: Jilin University, 2008: 103-105(in Chinese). [5] CHEN D W, LIAO X Y, WANG Zh L, et al. Solid phase microextraction gas chromatography-mass spectrometry for the determination of six phthalate esters in food plastic packaging bags [J]. Food Science, 2014, 35(6): 125-128(in Chinese). [6] WU R J, JIANG H. Inspection of plastic slippers by SEM/EDS[J]. Shanghai Plastics, 2017, 45(1): 18-22(in Chinese). [7] DONG X, RAO Zh F, YANG X Y, et al. Raman spectrometric detection of several plastics [J]. Plastics Industry, 2011, 42(6): 73-76(in Chinese). [8] FANG G, YIN L, LIU F, et al. Research on the application of diffe-rential Raman technology in fluorescence suppression [J]. Laser Technology, 2019, 43(3): 359-362(in Chinese). [9] MARSHALL S, COOPER J B. Quantitative Raman spectroscopy when the signal-to-noise is below the limit of quantitation due to fluorescence interference: Advantages of a moving window sequentially shifted excitation approach[J]. Applied Spectroscopy, 2016, 70(9): 1489-1501. [10] HU Y L, HUANG Y W, WANG R L, et al. Application of portable Raman spectrometer in food detection [J]. Food Industry Technology, 2017, 38(17): 319-323(in Chinese). [11] ZHOU H W, CAI Zh J, WU J H. Fluorescence suppression in Raman detection of illegal food additives [J]. Chinese Journal of Lasers, 2013, 40(9): 0915001 (in Chinese). [12] TANG Z J, ZHAMG M M, LU T J, et al. Application of portable differential Raman spectroscopy in amber identification [C]//Proceedings of China International Jewelry Academic Exchange Confe-rence (2019). Beijing: Jewelry and Jade Jewelry Management Center, Ministry of Land and Resources, 2019: 427-431(in Chinese). [13] SHAMEEM K M M, CHOUDHARI K S, BANKAPUR A, et al. A hybrid LIBS-Raman system combined with chemometrics: An efficient tool for plastic identification and sorting[J]. Analytical and Bioanalytical Chemistry, 2017, 409(13): 3299-3308. [14] MA X, JIANG H, YANG J Q, et al. A study on the inspection of plastic packing belt (rope) by Raman spectroscopy [J]. Shanghai Plastics, 2018, 46(4): 29-35(in Chinese). [15] ZHANG W T. SPSS advanced course of statistical analysis [M]. Beijing: Higher Education Press, 2012: 260-263(in Chinese). [16] HU L F. Five commonly used cluster analysis methods and their comparison [J]. Zhejiang Statistics, 2007, 26(4): 11-13(in Chin-ese). [17] CHU X L. Chemometrics and molecular light spectroscopy [M]. Beijing: Chemical Industry Press, 2011: 55-57(in Chinese). [18] TANG W J, YIN X D, YAN Q Zh, et al. Application of cluster analysis in distributed optical fiber vibration sensing system [J]. Laser Technology, 2015, 39(6): 854-857(in Chinese). [19] HE X L, WANG J F, LIU T F, et al. Raman spectroscopy combined with system clustering method to test the front bumper of automobile [J]. Journal of Light Scattering, 2018, 30(2): 168-173(in Chin-ese). [20] JIANG H, FAN Y, WANG J G, et al. A study on the inspection of rubber sole by XRF [J]. Infrared and Laser Engineering, 2017, 46(10): 326-331(in Chinese). [21] ZHANG L, MIN G Ch, TANG B, et al. Police situation analysis and application based on system clustering [J]. Police Technology, 2019, 35(1): 30-33(in Chinese). [22] YU X Ch, GUAN L, LI Z C, et al. A rapid recognition method of light fuel oil species by Raman spectroscopy based on fluorescence suppression and improved system cluster analysis of graphitized carbon black filter adsorption treatment [J]. Spectroscopy and Spectral Analysis, 2019, 39(3): 807-812(in Chinese).