Combination of FTIR, liquid-liquid aqueous partition and chemometric methods in the identification of tampered gasoline

Author(s)

Carneiro, Renato Lajarim ; Coan, Vitor Paulin

• Document type: Journal article
• Source: BRAZILIAN JOURNAL OF ANALYTICAL CHEMISTRY; v. 2, n. 7, p. 8-pg., 2012-01-01.

Abstract

Due to dependence on the use of gasoline, traders see an opportunity to increase their profit by adding solvents such as ethanol, kerosene, turpentine and diesel. Attenuated total reflectance with Fourier transform infrared spectroscopy (ATR-FTIR) is a fast and nondestructive technique and requires little sample. In this work, the applicability of ATR-FTIR for detection of adulteration of Brazilian gasolines using liquid-liquid aqueous separation and chemometric tools was evaluated. Gasolines from 15 gas stations from the cities of Tiete, Ibate and Sao Carlos (State of Sao Paulo) were acquired. Three data sets were generated: pure gasolines, polar phase and aqueous phase. The partition procedure aids in the identification of possible adulterants, since ethanol, present in Brazilian gasoline, presents strong peaks in the IR spectra. By IR spectra analysis methanol was found in the polar fraction of a sample, a novelty in the "gasoline tampering procedure". The same partition procedure was applied on samples spiked with methanol (MTH), ethanol (ETH), turpentine spirit (TUS) and toluene (TOL) from 0 to 25% yielding twelve datasets. It was found that identification of adulterated gasolines can be performed using spectra without partition but the quality of PCA separation is increased when pure, nonpolar and polar fraction spectra are concatenated, due the increase of chemical information. PLS models presented standard errors of cross-validation of 1.36%, 4.10%, 1.92% and 2.34% for MTH, ETH, TUS and TOL, respectively. The tampered sample presented 8.93% of methanol, using the developed model. (AU)

• FAPESP's process: 10/16520-5 - Application of calibration chemometrics methods and multivariate curve resolution in Raman spectroscopy for qualitative, quantitative and polymorphism analyses in medicines
• Grantee: Renato Lajarim Carneiro
• Support Opportunities: Regular Research Grants