A fluorescence-based multivariate method for biodiesel quantification in undiluted diesel-biodiesel blends without sample preparation

In this work, excitation-emission matrices (EEMs) were used in association with parallel factor analysis (PARAFAC) to assess biodiesel content in undiluted diesel-biodiesel blends (DBBs) without pre-sample preparation. EEMs were decomposed using the PARAFAC (EEMs-PARAFAC), and the loading values of the PARAFAC component as a function of biodiesel content in the blends were used to build an analytical model to quantify the biodiesel content in DBBs. The proposed model presenting a limit of detection (LOD) and a limit of quantification (LOQ) of 2.5% and 11% w/w, respectively, successfully predicted the biodiesel content in the validation samples. The robustness of the model was confirmed by a close analysis of the root mean square error of prediction, standard error of prediction, relative standard deviation of prediction, and Bias. Therefore, an accurate and robust analytical model based on EEMs-PARAFAC was developed to quantify the biodiesel content in undiluted DBBs without sample preparation.