Prediction of diesel fuel properties by vibrational spectroscopy using multivariate analysis

Partial least squares regression ( PLS ) calibration models based on Fourier transform infrared ( FTIR-ATR ) and Raman spectra ( FT-Raman ) were applied to the rapid and accurate simultaneous determination of the main properties of diesel fuels. Training sets were composed of over ninety commercial diesel fuel samples. The methods use baseline-uncorrected, raw FTIR-ATR and FT-Raman spectra. Two spectral regions were studied: full spectral region and “fingerprint” region. The models were validated using the cross-validation process. Based on the correlation coefficient and root mean square error of cross validation ( RMSECV ) values the both developed calibration models, PLS/FTIR-ATR and PLS/FT-Raman, were very accurate and comparable with standard testing methods. The following diesel fuel properties may be confidently estimated: cetane number, cetane index, density, viscosity, distillation temperatures at 10% (T10), 50% (T50) and 90% (T90) recovery, as well as the contents of total aromatics and polycyclic aromatic hydrocarbons. As compared to the “fingerprint” spectral region, the PLS/FTIR-ATR model using full spectral region displayed slightly better performances with the most of the correlation coefficient values above 0.98.