Determinação de esteróis totais em óleos vegetais usando Colorimetria por Imagem Digital

Abstract

Sterols are components present in vegetable oils and studied mainly for their property of inhibiting the absorption of cholesterol. Its determination can be performed using chromatographic methods or based on the Liebermann-Buchard colorimetric reaction, followed by spectrophotometric analysis. On the other hand, mobile devices have shown to be efficient in colorimetric analysis, enabling the development of portable and lower cost methods. Thus, the objective of this work was to develop a method for determining total sterols in vegetable oils through analytical signals obtained with a smartphone camera. For this, a Samsung Galaxy J5 smartphone with a 13MP camera was used, and the analyzes were performed using the Pipette Color Picker application, with a distance of 25 cm from the mobile device to the sample, and ambient lighting. Cholesterol calibration standard solutions in chloroform were submitted to colorimetric reaction, according to the Libermann-Burchard method and analyzed by the proposed method and by UV-Vis spectrophotometry. The quality of the adjustments was evaluated by the R² values in different color scales, the best fit being obtained using the RGB scale, with an average R² of 0.90 for B. The influence of the distance from the device to the sample (15 cm and 25 cm), lighting conditions (ambient and artificial fluorescent) and application (Pipette Color Picker and Color Grab) used were evaluated to predict the value of sterols in an olive oil sample. No significant effects of lighting variation were observed, however, a better analysis condition, with lower relative error, was observed when using the Pipette application at 25 cm away under ambient lighting, showing the importance of standardization of analysis conditions. Two pre-treatment conditions of the analytical signal were also tested, and when converting the value obtained from RGB to a value equivalent to the absorbance, better results of linearity of the method were obtained, with R² of 0.96 using the color component B, and standard error close to that found for the spectrophotometric method. The tests performed showed satisfactory linearity, detection limit, precision and accuracy. The relative standard deviation values obtained from the repeatability tests were less than 14.0% and less than 10.1% in terms of intermediate precision. The apparent recovery obtained in the accuracy tests performed was in the range of 91.6 and 107.9%. The precision and accuracy also evaluated by the analysis of vegetable oil samples showed results compatible with those found in the validation tests that used standard solutions. The variation in the days of analysis, batch of reagents and preparation of new stock and standard solutions without interfering in the results produced indicated that the method can be considered robust. Therefore, mobile devices can be used to replace the spectrophotometer in the determination of sterols in vegetable oils through the Liebermann-Burchard reaction, allowing greater access to the analysis, through devices widely used in everyday life, of easy mobility and low cost when compared to instrumental methods of chemical analysis.