Recognition of important chemicals in Engine oils affecting the value of oil viscosity indices using HS-SPME-GC-FID and PLS as a multivariate calibration technique

Viscosity index is one of the important physicochemical properties of engine oils covered under the American Standards for Testing and Materials (ASTM) [1]. In this study, we focus on the combination of gas chromatography with Chemometric techniques to relate the viscosity index of engine oils to the important retention times of their chemical compositions. Extracting of compounds from engine oils was done using graphene/polyvinylchloride nanocomposite coated SPME fiber [2] under headspace mode. The effective parameters on HS-SPME extraction of chemical compounds were investigated by one-variable-at-a-time method with optimized circumstances ( extraction temperature, 140 0C; extraction time, 30 min; desorption temperature, 230 0C; desorption time, 5 min; and stirring rate, 500rpm). Many preprocessing methods were done prior to partial least squares (PLS) technique, including asymmetric reweighted penalized least square (air-PLS), Savitzky-Golay filter and correlation optimized warping (COW). The results of the regression coefficient plot from PLS technique showed that the retention times of 12.78, 12.79, 12.97, 15.29, 15.65, 15.99, 16.91, 17.67 minutes have great positive effects on the value of viscosity indexes, and addition of such improver compounds in relation to these retention times can increase the value of viscosity index. Furthermore, the constructed model shows acceptable figure of merits (RMSEC= 2.35, R2= 0.99, Q2= 0.994 RMSECV= 1.95) which indicate the good performance and robustness of PLS model. The results in this work suggest that following the development of a robust model, it is possible to accurately determine which retention times are responsible for increasing the values of viscosity index of engine oils, using HS-SPME-GC-FID-PLS strategy. Moreover, the results demonstrated that the combination of HS-SPME-GC/FID with Chemometrics techniques is an appropriate and low-cost method for analysis of engine oils and evaluation of their viscosity index.