Modeling the characteristics and quantification of adulterants in gasoline using FTIR spectroscopy and chemometric calibrations

Abstract

Abstract The criminal act of fuel (gasoline) adulteration still remains a global worry due to its environmental, health and economic effect. Current methods for the detection of fuel adulteration have not been effective in most developing countries due to the associated cost of implementation. Therefore, there is the need for a fast, reliable and cheaper approach for screening of adulterants in fuel. This study combined FTIR analyses with Chemometric (multivariate) techniques for qualitative and quantitative determination of four possible adulterants: kerosene, diesel, naphtha and premix in gasoline. Synthetic admixtures prepared by mixing the gasoline with varying proportions of the adulterants were obtained and used for the model calibration. Soft Independent Modeling Class Analogy (SIMCA) classification and Partial Least Square (PLS) regression methods were the Chemometric techniques employed. The SIMCA classification model developed predicted the type of adulterant present at an error rate of 6.25% for Kerosene and naphtha, and 12.5% for premix. However, no prediction error was recorded for classifying samples contaminated with diesel. The PLS regression model was able to predict the concentrations of adulterant with prediction errors lower than 5% for all adulterants considered. Applying the models to commercial gasoline samples collected from a Metropolis in Ghana revealed 7% gasoline adulteration with kerosene (4%), premix (2%) or diesel (1%). No adulteration with naphtha was detected. The FTIR-Chemometric approach proved a fast and cheaper method for detection of adulteration which can be adopted by quality assurance and monitoring laboratories for forensic screening of gasoline in Ghana