多元素分析结合化学计量学方法鉴别富平山羊乳

Identification of Fuping goat milk by multi-element analysis combined with chemometrics

中文关键词:  山羊乳  多元素分析  化学计量学  正交偏最小二乘判别分析法  电感耦合等离子体质谱法  产地鉴别

英文关键词:goat milk  multiple elements analysis  chemometrics  orthogonal partial least squares discriminant analysis  inductively coupled plasma-mass spectrometry  geographical origin identification

中文摘要:

      目的 采用微波消解与电感耦合等离子体质谱(inductively coupled plasma-mass spectrometry, ICP-MS)法结合化学计量学方法实现富平山羊乳的产地鉴别。方法 从陕西省内的宝鸡市、汉中市、咸阳市泾阳县、咸阳市乾县、渭南市、西安市、延安市7个地区共采集70个羊乳样品, 冻干成乳粉, 经消解后, 采用ICP-MS对消解液进行检测, 通过主成分分析(principal component analysis, PCA)与正交偏最小二乘判别分析(orthogonal partial least squares-discrimination analysis, OPLS-DA)对不同产地羊乳进行区分。结果 对7个产地的羊乳粉中42种元素进行检测, K、Ca、Fe、Zn、Si、Na的含量较高, 其次是B、Ti、Mg、Al、Cu、Ba, 其他元素含量相对较低, 尤其是Pr、Nd、Sm、Eu、Gd、Tb、Dy、Er、Tm、Yb、Lu、Ho、Y等稀土元素含量甚微。按距离将7个产地分为I(宝鸡市、汉中市、渭南市、延安市)、II(咸阳市泾阳县、咸阳市乾县、渭南市、西安市)两组, 分别采用OPLS-DA模型分析, 准确率分别为95.12%、87.80%, 其中渭南市富平县判别准确率均为100%。对于组I, Ba、Mn、Si、Zn、Se、Na、B元素含量差异为产地区分的主要因素, 对于组II, Se、B、Ni元素对产地区分贡献较大。地理距离是影响羊乳产地鉴别的重要因素。将6个渭南市富平县羊乳样品数据分别导入组I、组II模型进行验证, 判别准确率均为100%。结论 本研究建立的多元素分析结合化学计量学方法可实现小尺度(50 km以上)范围内富平山羊乳的快速、准确识别。

英文摘要:

      Objective To achieve the geographical discrimination of Fuping goat milk using microwave digestion combined with inductively coupled plasma mass spectrometry (ICP-MS) and chemometric methods. Methods A total of 70 goat milk samples were collected from 7 regions in Shaanxi Province, including Baoji, Hanzhong, Jingyang County and Qian County (both in Xianyang), Weinan, Xi’an and Yan’an. They were lyophilized to milk powder, and after microwave digestion, the digested solution was analyzed using ICP-MS. Principal component analysis (PCA) and orthogonal partial least squares-discrimination analysis (OPLS-DA) were employed to discriminate goat milk from different origins. Results A total of 42 kinds of elements in the milk powder from 7 different origins were detected. The concentrations of K, Ca, Fe, Zn, Si and Na were relatively high, followed by B, Ti, Mg, Al, Cu and Ba, while other elements were found in lower concentrations, particularly the rare earth elements such as Pr, Nd, Sm, Eu, Gd, Tb, Dy, Er, Tm, Yb, Lu, Ho and Y, which were present in negligible amounts. The 7 origins were divided into two groups based on distance: Group I (Baoji; Hanzhong; Weinan; Yan’an) and Group II (Jingyang County, Qian County of Xianyang; Weinan; Xi’an). OPLS-DA model was applied to both groups, with accuracy rates of 95.12% and 87.80%, respectively. Among them, the accuracy of samples from Fuping County-Weinan City was 100%. In Group I, the content differences of Ba, Mn, Si, Zn, Se, Na and B were the primary factors for origin discrimination. While in Group II, Se, B and Ni elements contributed significantly to the origin discrimination. Geographical distance was the most important factor to distinguish goat milk from different origins. While the data of 6 goat milk samples from Fuping County-Weinan City were imported into both Group I and Group II models for validation, the discrimination accuracy was 100% for both groups. Conclusion The established multi-element analysis combined with chemometrics could achieve rapid and accurate identification of Fuping goat milk from small-scale regions (>50 km).