我国仿真饰品中有害元素含量测定与特征分析

Identification and characteristic analysis of the hazardous pollutant elements in artificially imitated jewelries made in China

仿真饰品与人体皮肤直接接触,其安全性与消费者健康密切相关。对我国制造的162个仿真饰品中的总铅、总镉、总汞、总砷、总铬和6价铬质量比进行了检测,采用四分位稳健统计法对6种有害元素的质量比进行了统计描述。利用本征矢量投影技术对试验数据进行降维处理,使样本点在新变量PC1、PC2空间的整体分布趋势及特征得到清晰表达,并对造成异常样品点的影响因素进行了分析。进一步因子分析结果表明,我国仿真饰品中有害元素的分布特征可基本确定为两个因子:Pb和Hg。因子分析的结果对解释影响样本点差异的实际变量方面更具实际意义,即仿真饰品样本点的异常值可归为两类:受Pb因子影响和受Hg因子影响。

The present paper intends to propose our research on how to identify and analyze the characteristic features of the hazardous elements in the imitated or artificial jewelries made in China. To oar knowledge, when the imitated jewelries are getting in direct touch with human skin, they are likely to bring about a lot of potential hazards to the carriers＇ health due to the heavy metal pollutants. To make clear the above said hazardous elements, we have collected 162 such jewelry samples made in China from July to November of 2012 and analyzed the content of 6 main hazard elements, that is, the content rate of Pb, Cd, Hg, As, Cr, and the Cr6 ＋ in each sample via the inductively coupled plasma-atomic emission spectrometry （ICP- AES）. Generally speaking, the imitated jewelries made in China can basically reach the standardized demands of the country and come to the demanded level. However, when the production standard process is used to test the actuality of the production process, it can be found that the results of the test fail to follow the demanded data distribution graph. When describing the pollutant content variation ranges by using the interquantile robust statistics, the highest content rate of Cr（T） is 3442 mg/kg among the six elements, whereas the medium rate of Pb（T） is 53.9 mg/kg, which turned to be higher than the other five elements. If Considered from the point of view of the coefficient matrix, the highest correlation data can be found between Pb （T） and Cd（T）, which may account for the intention to improve the ornament flexibility. When doing the principal component analysis （PCA） and factor analysis （FA） the distribution features of the sam- ples can clearly be found in the new PC1- PC2 ordinate generated by the orthogonal projection, which may account for the outliers in the samples under discussion. Most of the outliers were found in the ear acupuncture or necklace, in which Cr（T） content rate proves to be most obviously higher than in other samples. However, in some other outliers, the Cd（T） and Pb（T） content rate are rather seriously higher than in others. In order to get the latent variables which may most distinguishably affect the analysis accuracy, we have investigated the data profiles via factor analysis. Two latent variables were responsible for the sample distribution, that is to say, the Pb（T） factor （FA1） and Hg（T） factor （FA2） have been found in correspondence with PC1 and PC2 respectively. The results from the above factor analysis tend to more practical than from the principal component analysis.