近红外光谱方法预测生物柴油主要成分

Prediction of Main Chemical Compositions in Biodiesel by Near Infrared Spectroscopy

采用近红外光谱快速测定法对生物柴油的成分(脂肪酸甲酯、单甘酯、二甘酯、三甘酯和甘油)进行了研究。采用气相色谱方法获得其成分的基础数据,通过偏最小二乘方法与近红外光谱数据进行回归运算,分别建立以文冠果油生物柴油为例的单原料油校正模型及多种原料油生物柴油的混合校正模型,并以花椒油生物柴油为例考察了校正模型的适用性。结果表明:通过偏最小二乘方法可以建立适合多种原料油生物柴油的通用校正模型。对于新型生物柴油,向校正集中添加10个以上样本,扩充校正模型后,便可较为准确地测定这类新生物柴油样本的成分含量。此方法分析速度快、成本低、操作便捷、重复性好,适合于生物柴油生产过程的中间控制分析。

A rapid method was established for the determination of the main chemical compositions in biodiesel （including fatty acid methyl esters, monoglycerides, diglyceride, triglyceride and glycerol） by the combination of near infrared spectra （NIR） and chemometrics. The reference values were obtained by the gas chromatographic method, which was correlated with the NIR spectra data by partial least squares regression method. The calibration models for multi-feedstock biodiesel and a single feedstock biodiesel were established respectively. The applicability of the calibration models was examined with the zanthoxylum oil biodiesel as an example. The results showed that a universal calibration model suitable for various biodiesel from different sources can be established by partial least-squares method. For this biodiesel manufactured from other fatty oils, the same calibration model can be extended by adding 10 more new samples. The calibration model so obtained will provide more accurate results. The NIR method is suggested as a rapid method for the quality control in the process of biodiesel production.