No abstract

A three-step development, optimization and validation strategy is described for gas chromatography (GC) fingerprints of Brazilian commercial diesel fuel. A suitable GC-flame ionization detection (FID) system was selected to assay a complex matrix such as diesel. The next step was to improve acceptable chromatographic resolution with reduced analysis time, which is recommended for routine applications. Full three-level factorial designs were performed to improve flow rate, oven ramps, injection volume and split ratio in the GC system. Finally, several validation parameters were performed. The GC fingerprinting can be coupled with pattern recognition and multivariate regressions analyses to determine fuel quality and fuel physicochemical parameters. This strategy can also be applied to develop fingerprints for quality control of other fuel types.

In the present study, hierarchical cluster analysis was used to select 150 S500 diesel fuel samples from an initial set of 1320 samples assayed through official standards according to ANP Brazilian Regulation No. 50/2013. Four physicochemical properties were analyzed, namely, relative density, distillation temperatures (T10%, T50%, and T85%), flash point, and cetane number. Selected samples were also analyzed by gas chromatography with flame ionization detection (GC-FID), a very common technique used for fuel quality control due to its convenience, accuracy, simplicity, and possible association of the chromatographic profiles with multivariate analyses. PLS regression models were obtained aiming at predicting the four physicochemical properties of the diesel fuel samples. From a maximum chromatographic analysis time of 108 min, regression models with unbiased predictions and good prediction capability for all properties were obtained, with average relative errors lower than 6%.