Monitoramento da hidrogenação industrial do óleo de soja empregando cromatografia gasosa, espectroscopia de infravermelho médio e análise multivariada

Abstract

The oils and fats industry routinely get a large amount of data during its operational routines. One of the main parameters used to control the hydrogenation process is the iodine value (IV) which quantifies the relative unsaturation degree of fatty acids. The traditional method for determining the iodine value (AOCS Cd 1b-87) involves several stages of preparation, resulting in a running time of around 30 to 45 minutes. Generally, the industrial hydrogenation process takes place in batches, so the reactor remains stationary with the oil and reagents until complete the IV analysis, which is critical to release the hydrogenated fat for the next steps. Thus, the combination of fast instrumental techniques with multivariate analysis methods is impportant to extract the most relevant information about the chemical composition of samples, allowing an online process monitoring. In this context, this work had two main objectives. First, monitoring was performed in the industrial process of soybean oil hydrogenation using self-organizing maps (SOM) in combination with fatty acid composition data obtained by gas chromatography with flame ionization detector (GC-FID). Subsequently, through the mid-infrared spectroscopy with attenuated reflectance (FTIR-ATR), was developed a regression model by partial least squares (PLS) for the rapid determination of the IV. Through the SOM it was possible monitoring the saturation and isomerization reactions which occur during the hydrogenation. It was also observed that the trans isomers are more stable and less sensible to the hydrogenation that the cis isomers. According to the maps, the variation of monounsaturated fatty acids is mainly related with an increase in C18:1, the variation of polyunsaturated with reduction in C18: 2 (cis), saturated fatty acids with the increase in C18:0 and trans fatty acids an increase in C18:1 (trans). The IV was the main chemical parameter responsible for segmentation of the samples as a function of hydrogenation time, with a significant reduction in average of 125.75 to 60.04 g I2.100 g-1. Thus, the SOM was efficient for the recognition and characterization of the different stages of the industrial hydrogenation process. According to performance shown by figures of merit, the proposed PLS model showed a good correlation between the reference values and the predicted values of IV. The model adequately captured the chemical nature of soybean oil hydrogenation process on industrial reactor and confirming the absorption bands more correlated to IV. The set of results shows that the multivariate analysis in tandem with instrumental methods can be used as a fast and reliable alternative to monitoring the industrial soybean oil hydrogenation process and determination of IV with minimal sample preparation and without the generation of chemical residues. Thus, collecting samples of the industrial process in progress and get results in real time can correct operational deviations and prevent production downtime.