| 李婧妍，郭春锋，李琴，刘拉平.脂肪酸含量分析结合化学计量学鉴别
掺伪核桃油方法研究[J].中国油脂,2026,51(8):.[LI Jingyan,GUO Chunfeng, LI Qin, LIU Laping.Identification of adulterated walnut oil based on fatty acid content analysis and chemometrics methods[J].China Oils and Fats,2026,51(8):.] |
| 脂肪酸含量分析结合化学计量学鉴别
掺伪核桃油方法研究 |
| Identification of adulterated walnut oil based on fatty acid content analysis and chemometrics methods |
|
投稿时间：2025-11-24 修订日期：2026-03-26 录用日期：2026-01-16
出版日期：
|
| DOI：10.19902/j.cnki.zgyz.1003-7969.250499 |
| 中文关键词: 核桃油 掺伪鉴别 脂肪酸组成 化学计量学方法 |
| 英文关键词:walnut oil adulteration identification fatty acid composition chemometrics methods |
| 基金项目:陕西省重点研发计划项目（2023-YBNY-191）；秦创原“科学家+工程师”队伍建设项目（2024QCY-KXJ-084）；西安市农业技术攻关项目（2024JH-NYYB-0134）；陕西省市场监督管理局科技计划项目（2023KY29）；宁夏中央引导地方科技发展资金项目（2024FDR05071） |
|
|
| 摘要点击次数: 3 |
| 全文下载次数: 0 |
| 中文摘要: |
| 为给核桃油掺伪鉴别提供新的筛查与理论参考，基于脂肪酸含量分析结合化学计量学技术，建立了核桃油掺伪鉴别方法。采用气相色谱法测定核桃油及6种常见植物油(大豆油、花生油、菜籽油、玉米胚芽油、葵花籽油和橄榄油）的脂肪酸组成，运用聚类分析(HCA)和主成分分析（PCA）探究不同植物油中各脂肪酸分布规律，筛选出可能掺伪于核桃油中的植物油，并运用偏最小二乘判别分析（PLS-DA）和多元逐步线性回归分析构建核桃油掺伪鉴别的定性和定量模型。结果表明：6种植物油与核桃油的脂肪酸组成存在不同程度的差异;HCA 和 PCA确定葵花籽油和玉米胚芽油与核桃油脂肪酸组成最为接近，可拟作核桃油的掺伪油；所建PLS-DA模型可区分掺伪量在5%以上的掺伪油，并筛查出掺伪葵花籽油的特征脂肪酸（木蜡酸、亚麻酸、硬脂酸、油酸、棕榈酸和亚油酸）和掺伪玉米胚芽油的特征脂肪酸（亚油酸、亚麻酸、油酸、硬脂酸和棕榈酸）；基于多元逐步线性回归分析建立的核桃油掺伪葵花籽油和玉米胚芽油定量预测模型的决定系数（R2）均大于0.9；对2个定量模型验证表明，实际掺伪量（5%~50%）和预测掺伪量在95%置信区间内均显示出良好的线性关系，R2均超过0.98。综上，所建多元逐步线性回归分析模型可应用于核桃油中掺伪葵花籽油和玉米胚芽油的定量识别。 |
| 英文摘要: |
| To provide a new screening and theoretical reference for the adulteration identification of walnut oil, an adulteration identification method was established based on fatty acid content analysis combined with chemometric techniques. The fatty acid compositions of walnut oil and six common vegetable oils (soybean oil, peanut oil, rapeseed oil, maize oil, sunflower seed oil and olive oil) were determined using gas chromatography. Cluster analysis (HCA) and principal component analysis (PCA) were employed to explore the distribution patterns of various fatty acids in different vegetable oils and to screen for vegetable oils that could potentially adulterate walnut oil. Partial least squares discriminant analysis (PLS-DA) and multiple stepwise linear regression analysis were used to construct qualitative and quantitative models for walnut oil adulteration identification, respectively. The results showed that there were varying degrees of differences in the fatty acid composition among the six vegetable oils and walnut oil. HCA and PCA revealed that sunflower seed oil and maize oil exhibited the closest similarity to walnut oil in terms of fatty acid composition, suggesting they could be used as adulterants for walnut oil. The established PLS-DA model could distinguish adulterated oils with an adulteration level of 5% or more, and identified characteristic fatty acids for adulteration with sunflower seed oil (lignoceric acid, linolenic acid, stearic acid, oleic acid, palmitic acid, and linoleic acid) and for adulteration with maize oil (linoleic acid, linolenic acid, oleic acid, stearic acid, and palmitic acid). The determination coefficients (R2) of the quantitative prediction models for walnut oil adulterated with sunflower seed oil and maize oil, established based on multiple stepwise linear regression analysis, were both greater than 0.9. Validation of the two quantitative models showed a good linear relationship between the actual adulteration level (5%-50%) and the predicted adulteration level within the 95% confidence interval, with R2 values exceeding 0.98 for both models. In conclusion, the established multiple stepwise linear regression analysis models can be applied for the quantitative identification of sunflower seed oil and maize oil adulterated in walnut oil. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
