Edible oils are essential components in a human being’s diet due to their rich flavor and nutritional values. In India, the use of sunflower (Helianthus annuus) oil has increased because of its unique qualities, such as having good polyunsaturated fatty acids and less mono-unsaturated fatty acids. Sunflower oil is a highly nutritious edible oil is likely to be adulterated with cheap oils like Palm oil due to the increase in demand. In this paper, the problem of sunflower oil adulteration with palm oil is studied using attenuated total reflection (ATR) mid-infrared spectroscopy. Using laboratory-made adulterated samples as a reference, ATR spectroscopy (wavenumber range 1781 to 915 cm⁻¹) combined with chemometrics was successful in detecting the presence of palm oil in sunflower oil and predicting the adulteration proportions. Applying principal component analysis (PCA), distinguishable clusters of pure and adulterated samples were realized. Multiple wave number ranges associated with various functional groups (1781–1680 cm⁻¹, 1490–919 cm⁻¹), and some ranges from PCA correlation loading plot were selected for further analysis. Classification of pure and adulterated samples was accomplished employing partial least square discrimination analysis and soft independent modeling of class analogy, with classification efficiency of 100 percent and 90 percent, respectively. For quantitative prediction, the partial least squares regression method on selected wavenumber range data was applied. The coefficient of determination (R²) for calibration and validation was observed to be 0.98. This study showed ATR-based Mid-infrared spectroscopy’s potential for the detection of palm oil adulteration in sunflower oil even at a minimum of 5% adulteration.

食用油因其丰富的风味和营养价值而成为人类饮食中必不可少的成分。在印度，向日葵 (Helianthus annuus) 油的使用有所增加，因为它具有独特的品质，例如具有良好的多不饱和脂肪酸和较少的单不饱和脂肪酸。葵花油是一种营养丰富的食用油，由于需求增加，很可能会掺入棕榈油等廉价油。在本文中，使用衰减全反射 (ATR) 中红外光谱研究了向日葵油掺入棕榈油的问题。使用实验室制造的掺假样品作为参考，ATR 光谱（波数范围 1781 至 915 cm ^-1) 结合化学计量学成功地检测了葵花油中棕榈油的存在并预测了掺假比例。应用主成分分析 (PCA)，实现了可区分的纯样品和掺假样品簇。与各种官能团相关的多个波数范围（1781-1680 cm ^-1 , 1490-919 cm ^-1)，并从 PCA 相关载荷图中选择了一些范围进行进一步分析。纯样本和掺假样本的分类采用偏最小二乘判别分析和类类比的软独立建模完成，分类效率分别为 100% 和 90%。对于定量预测，对选定的波数范围数据应用偏最小二乘回归方法。观察到校准和验证的决定系数 ( R ² ) 为 0.98。这项研究表明，基于 ATR 的中红外光谱在检测葵花油中棕榈油掺假方面的潜力，即使掺假至少为 5%。