Design experiments to detect and quantify soybean oil in extra virgin olive oil using portable Raman spectroscopy

In this work, a full factorial design 23 for the optimization of experiments was performed using a portable Raman spectrometer to find ideal instrumental conditions (laser power, analysis accessory, and luminosity) in the detection and quantification of adulterant, soybean oil, in extra virgin olive oil (EVOO). The ratio between the intensities of vibrational bands of 1440 and 1266 cm-1, characteristics of oleic and linoleic acids, respectively, were evaluated as a response measure. It was found that the configuration of the accessories has no significant influence on the analytical response. For quantification of soybean oil in EVOO, an analytical curve was constructed with six concentrations of the binary mixture containing EVOO with soybean oil (100/0; 80/20; 60/40; 40/60; 20/80; and 0/100 % v/v) and the model was validated from use of five samples adulterated with known concentrations (90/10; 70/30; 50/50; 30/70; and 10/90 % v/v) employing the intensity (I) ratio of the I1266/ I1440 and I1656/I1440. The best regression method, using ordinary least squares (OLS), was through the ratio of the intensity of I1266/I1440, which presented higher linearity with R2c and R2p of 0.9769 and 0.9450, respectively; and Root Mean Square Error of Calibration (RMSEC) and Root Mean Square Error of Prediction (RMSEP) of 1.445 and 1.975 % (v/v), respectively.

Automated summary

In this study, a full factorial design was carried out to optimize the detection and quantification of adulterated soybean oil in extra virgin olive oil using a portable Raman spectrometer. The ratios between vibrational bands were evaluated as response measures, and the best regression model using ordinary least squares was determined for quantification. By constructing an analytical curve and validating the model, it was found that the intensity ratio of I1266/I1440 presented the highest linearity and accuracy for determining adulterant concentrations.