Fast Determination of Biodiesel Content in Commercial Diesel/Biodiesel Blends by Using Digital Images and Multivariate Calibration.

A new simple, rapid and inexpensive analytical method was developed to determine the biodiesel percentage in biodiesel/diesel blends through simple digital images of samples obtained by scanning with a commercial scanner. Soybean biodiesel and petroleum diesel samples were acquired from refineries currently in operation. There were prepared several mixtures within the range 1.5 to 12.0% of biodiesel in diesel oil, using the same procedure as is done in industry. The analytical signals were images recorded with a scanner. This data was decomposed with different color systems: RGB, HSV, HLS, CMYK and Grayscale. Chemometrics models based on color signals obtained from different mixtures of biodiesel/diesel were built. The quantification by using partial least squares (PLS) resulted in a RMSEP value for biodiesel of 0.9% (w/w); this load approximately 10-times smaller than the corresponding calibration range, with a correlation of 0.96 between predicted and reference values.

Highly sensitive quantitation of pesticides in fruit juice samples by modeling four-way data gathered with high-performance liquid chromatography with fluorescence excitation-emission detection.

A study regarding the acquisition and analytical utilization of four-way data acquired by monitoring excitation-emission fluorescence matrices at different elution time points in a fast HPLC procedure is presented. The data were modeled with three well-known algorithms: PARAFAC, U-PLS/RTL and MCR-ALS, the latter conveniently adapted to model third-order data. The second-order advantage was exploited when analyzing samples containing uncalibrated components. The best results were furnished with the algorithm U-PLS/RTL. This fact is indicative of both no peak time shifts occurrence among samples and high colinearity among spectra. Besides, this latent-variable structured algorithm is capable of better handle the need of achieving high sensitivity for the analysis of one of the analytes. In addition, a significant enhancement in both predictions and analytical figures of merit was observed for carbendazim, thiabendazole, fuberidazole, carbofuran, carbaryl and 1-naphtol, when going from second- to third-order data. LODs obtained were ranged between 0.02 and 2.4µgL(-1).

Modeling excitation-emission fluorescence matrices with pattern recognition algorithms for classification of Argentine white wines according grape variety.

This paper reports the modeling of excitation-emission matrices for classification of Argentinean white wines according to the grape variety employing chemometric tools for pattern recognition. The discriminative power of the data was first investigated using Principal Component Analysis (PCA) and Parallel Factor Analysis (PARAFAC). The score plots showed strong overlapping between classes. A forty-one samples set was partitioned into training and test sets by the Kennard-Stone algorithm. The algorithms evaluated were SIMCA, N- and U-PLS-DA and SPA-LDA. The fit of the implemented models was assessed by mean of accuracy, sensitivity and specificity. These models were then used to assign the type of grape of the wines corresponding to the twenty samples test set. The best results were obtained for U-PLS-DA and SPA-LDA with 76% and 80% accuracy.