Rational biodiesel production by screening different long-term stored plant oils via nuclear magnetic resonance spectroscopic measurement

Abstract

Screening the raw materials for biodiesel production has normally concentrated on the triglyceride (TAG) content and fatty acid profile without considering the oxidation stability, a crucial restriction factor for commercial life of biodiesel. The traditional screening methods had complex procedures by characterizing the raw material’ and the prepared biodiesel’ properties, which was not applicable for large sets of raw materials. These factors propelled the development of a simple and efficient screening method for potential biodiesel feedstock. Herein, we compared two kinds of spectroscopic measurements, the three-dimensional fluorescence (3D-FL) and high field nuclear magnetic resonance (NMR), for screening special and differential raw materials by investigating 19 plant oil species with one-year storage for biodiesel production. Among these samples, 13 kinds of plant oils contained TAGs, and the other 6 were free of TAGs. The results showed that 3D-FL measurements improperly assigned citronella oil as a TAG-containing plant oil and determined limited information about their chemical composition, suggesting unsatisfactory screening performance. In contrast, 1H NMR measurements screened 3 kinds of differential oils out of those 19 plant oil species, including coconut oil (Cocos nucifera L.), cotton seed oil (Gossypium barbadense L.), palm oil (Elaeis guineensis). Hierarchical clustering analysis (HCA) and partial least square discriminant analysis (PLS-DA) presented that the statistically significant NMR spectral bins for their differential compositions were signals at 3.6 ppm (–OCH3), 2.31 ppm (–OCH2–COOH), 4.29 ppm (TAG skeleton), and 5.27 ppm (TAG skeleton), representing the oxidation products, unsaturated fatty acids, and TAGs. Finally, the converted methyl and ethyl esters (biodiesel) from those three differential plant oil samples were subjected to Rancimat tests. They showed quite different oxidation stabilities from Rancimat tests, which was relevant to the composition differences in the NMR analysis. Our results suggested that the rapid, simple, and reliable NMR measurement combined with HCA and PLS-DA can serve as an efficient tool for potential biodiesel feedstock screening.