菜籽油品质多样性与红外光谱特征的研究Quality diversity and Fourier transform infrared spectrum characteristic of rapeseed oil

Abstract

为对菜籽油的品质监控和快速检测提供科学依据，测定49种菜籽油样品的质量指标和脂肪酸组成，分析其品质多样性，并考察其品质特征与傅里叶变换衰减全反射红外光谱的相关性。结果表明：49种菜籽油样品过氧化值和酸值的变异系数分别达63.20%和152.90%，所含主要脂肪酸中棕榈酸、油酸和亚油酸含量的变异系数均未超过21.38%，而花生一烯酸含量的变异系数高达127.09%；通过分类分析发现，油样的过氧化值、亚油酸含量和不饱和脂肪酸含量与其产地、加工工艺、品种、上架期和质量等级有关；采用化学计量学方法构建油样红外光谱的共有模式，以相关系数法分析各光谱的相似度介于0.68～0.98之间，并筛选出10个共有峰。以共有峰的峰高均值为自变量，构建脂肪酸(棕榈酸、油酸、亚油酸、花生一烯酸、不饱和脂肪酸和饱和脂肪酸)含量的多元线性回归模型均达显著水平(p＜0.05)，但过氧化值和酸值的回归模型不显著(p＞0.05)。通过对30种热处理油样的验证实验发现，回归模型能准确预测其不饱和脂肪酸含量，预测值的相对误差介于-11.15%～10.95%之间。综上，红外光谱法可初步用于市售菜籽油的脂肪酸组成的快速分析。 To provide scientific basis for quality monitoring and rapid detection of rapeseed oil, fortynine rapeseed oil samples were collected to determine their quality index and fatty acid composition, their quality diversity was analyzed, as well as the relationship between quality characteristics and attenuated total reflectance-Fourier transform infrared spectrum(ATR-FTIR). The results showed that the coefficients of variation (CV) for the peroxide value (PV) and acid value (AV) of the oil samples were 63.20% and 152.90%, respectively. The CV of the main fatty acids composition of rapeseed oils, palmitic acid, oleic acid and linoleic acid was not more than 21.38%. By comparison, the CV of eicosenoic acid content reached 127.09%. Further classification analysis indicated that the PV, linoleic acid content and unsaturated fatty acid content of the oil samples were associated with their producing areas，processing technologies, rapeseed varieties, shelf periods and quality grades. The ATR-FTIR common model of rapeseed oils was established by chemometrics. The similarity value of each spectrum was analyzed by correlation coefficient method, ranging from 0.68 to 0.98, and 10 common peaks were screened out. Taking the peak height mean value of the common peak as the independent variable, the multiple linear regression models of PV, AV, and main fatty acid content were constructed. The fatty acid models, involving palmitic acid, oleic acid, linoleic acid, eicosenoic acid, unsaturated fatty acid and saturated fatty acid, were all significant (p0.05). The results of verification test performed with 30 heat-treated samples indicated that the model could accurately predict the content of unsaturated fatty acid with a relative error range of -11.15%-10.95%. In conclusion, Fourier transform infrared spectroscopy can be preliminarily used for the rapid analysis of fatty acid composition of the commercially available rapeseed oil.

Keywords

• 菜籽油；品质；傅里叶红外光谱；多元线性回归模型
• rapeseed oil; quality; fourier transform infrared spectrum; multiple linear regression model