Application of Multivariate Curve Resolution for in-line monitoring of the acid esterification of Jatropha oil
Raw materials with high content of free fatty acids (FFAs), like Jatropha curcas oil (JCO), may produce a high amount of soap when the biodiesel is produced from triglycerides by transesterification reaction with short chain alcohols in the presence of a basic catalyst. Prior to transesterification reaction, high acidic vegetable oils require a preliminary step of acid esterification to convert FFAs into esters in the presence of both alcohol and acid catalysts. However, during the acid esterification (primary reaction) secondary reactions may take place: (1) acid transesterification of triglycerides (2) triglyceride hydrolysis and (3) alkylation of sulphuric acid with ethanol. The study of the esterification reaction is not simple; the temperature and the concentration of reactants determine the presence of two phases but their global behavior depend of the type of oil. The purpose of this work is to evaluate the capacity of methods using NIR spectroscopy and multivariate curve resolution alternating least squares (MCR-ALS), to study all reactions involved in the esterification process of the JCO. Sulphuric acid alkylation in the presence of ethanol and the esterification reactions of the oleic acid and the soybean oil have been considered as useful pattern of the reactions for assessing the capacities of the NIR-MCR-ALS methodology. First, the raw JCO was submitted to a degumming process to remove eliminate the phospholipids contained in the oil. The esterification reactions were carried out in a Mettler Toledo OptiMaxTM 1001 glass reactor with 1000 mL, using ethanol and oil in molar ratio of 16:1, 1% w/w H2SO4 as a catalyst, temperature of 75°C and reaction time of 120 minutes. The pattern reactions were carried out separately but using the same conditions of the esterification reactions of JCO. All reactions were monitored in line by using a fiber-optic transflectance probe (Solvias) connected to an ABB Bomen FTIR (FTLA 2000-160) spectrometer. The spectra were acquired within the interval between 4000 cm-1 and 12000 cm-1, with a 1.0 mm optical path, average of 100 scans, spectral resolution of 8 cm-1 using the kinetic mode of the equipment (every 60 seconds).
Spectroscopic and chemometric analysis of the data, evidenced the complexity of the quantitative resolution of the studied reactions because the spectral variations always have been associated to absorption bands assigned to stretching vibration of the –OH and –CH3 and –CH2 functional groups in different chemical environments and their total concentration did not vary during the process. However, from the analysis of the spectra and concentration profiles recovered by MCR-ALS analysis in each experiment, we concluded that analytical methods based on NIR with MCR-ALS are a quick and efficient strategy to evaluate the feasibility of the esterification process of oil with high content in FFA for posterior biodiesel production.