Rapid and sensitive method for detecting adulterants in gasoline using ultra-fast gas chromatography and Partial Least Square Discriminant Analysis

• Autor(a) principal: Nespeca, Maurílio Gustavo [UNESP]
• Data de Publicação: 2018
• Outros Autores: Munhoz, João Fernando Villarrubia Lopes [UNESP], Flumignan, Danilo Luiz, de Oliveira, José Eduardo [UNESP]
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.1016/j.fuel.2017.11.032; http://hdl.handle.net/11449/170375
• topic: Fuel quality control; Gasoline adulteration; Multivariate filters; Partial Least Square Discriminant Analysis; Selectivity ratio; Ultra-fast gas chromatography

In the last years, the Brazilian Fuel Quality Monitoring Program drastically reduced the number of analyzed fuel samples as consequence of the current economic crisis in the country. The impoverishment of the monitoring program may lead to an increase in cases of gasoline adulteration, nonetheless, it also strengthens the search for faster and less costly methodologies for the fuel quality monitoring. Thus, this study aimed the development of a rapid analytical method to detect the adulteration of gasoline with organic solvents through ultra-fast gas chromatography with flame ionization detector (UFGC-FID) associated with the supervised pattern recognition method, Partial Least Square Discriminant Analysis (PLS-DA). The sample set consisted of 171 Brazilian common gasoline (i.e., with ethanol in its composition) and 171 adulterated gasoline prepared in laboratory using 19 different solvents in the concentration range of 2–10% (v/v). The chromatographic method required only 2.85 min and the chromatograms presented 125 peaks on average. The PLS-DA model was developed with 3 latent variables and provided correlation coefficients close to 0.99 and correct discrimination of 100% of calibration and validation samples. Therefore, the developed UFGC-FID/PLS-DA method provided a sensitive, fast and automated alternative method for the detection of adulterants in the monitoring of gasoline quality.