In this work, the assessment of butter and margarine using 1H nuclear magnetic resonance (NMR) spectroscopy and principal component analysis (PCA) was achieved successfully. The first two PCs that account for more than 90% of the total variation in the spectral dataset were used in constructing the PCA score plot. 40 samples of butter from 8 brands and 35 margarine samples from 7 brands were studied. To imitate adulterated butter with margarine, a selected butter brand was mixed with different mass ratios of a selected margarine brand (5, 10, 20 and 30%). All samples in this study were dissolved in CDCl3 and analyzed in a similar manner. Again, successful detection of margarine adulterated butter was achieved. Quantitative determination of adulteration was not possible in the current study. In all PCA parts through this study, data corresponding to the chemical shift period (2.5–1.0 ppm) produced the best assessment results. Hence, this part from all spectra was used throughout this study.

中文翻译：

---

使用 1H NMR 和 PCA 研究人造黄油掺入黄油

在这项工作中，成功实现了使用 1 H 核磁共振 (NMR) 光谱和主成分分析 (PCA) 评估黄油和人造黄油。占光谱数据集中总变异 90% 以上的前两个 PC 用于构建 PCA 得分图。研究了来自 8 个品牌的 40 个黄油样品和来自 7 个品牌的 35 个人造黄油样品。为了用人造黄油模仿掺假黄油，将选定的黄油品牌与选定的人造黄油品牌的不同质量比（5%、10%、20% 和 30%）混合。本研究中的所有样品都溶解在 CDCl3 中并以类似的方式进行分析。再次成功检测出人造黄油掺假黄油。在当前的研究中不可能对掺假进行定量测定。在本研究的所有 PCA 部分中，对应于化学位移周期 (2.5–1.0 ppm) 的数据产生了最好的评估结果。因此，在整个研究中使用了来自所有光谱的这部分。