Fermentação alcoólica extrativa com remoção de etanol por arraste com CO2 e recuperação por absorção

Abstract

Ethanol removal by CO2 stripping during alcoholic fermentation is one way of overcoming the problem of inhibition by the ethanol. However, the lack of efficient methods to recover the removed ethanol still makes the use of stripping unviable. In this work, an innovative process for ethanol production was developed based on the integration of the extractive fermentation with CO2 stripping and with the recovery of ethanol by absorption. Firstly, the ethanol removal from hydro alcoholic solutions by CO2 stripping was evaluated in order to obtain a model to describe this process. Subsequently, techniques based on the combination of Fourier transform mid-infrared (FT-MIR) spectroscopy and partial least-squares (PLS) regression were developed for monitoring of the ethanol absorption and the alcoholic fermentation. In sequence, gas absorption was evaluated as a technique for the recovery of ethanol from the gas mixture generated by CO2 stripping. Water, monoethylene glycol (MEG), and diethylene glycol were evaluated in terms of their performance in ethanol absorption in order to choose the most appropriate absorbent. Subsequent assays using the best absorbent were conducted to investigate the influence of the initial MEG volume in the absorber, the recirculation volumetric flow rate of solvent, and the use of two absorbers connected in series on the absorption process. A modeling procedure was developed based on mass balance equations for the species involved (ethanol, water, CO2, and absorbent), and incorporating stripping and absorption kinetics, and vapor-liquid equilibrium concepts. A conventional fed-batch fermentation was performed in order to estimate kinetic parameters of growth for the yeast Saccharomyces cerevisiae. Lastly, extractive fermentations without CO2 recirculation (using commercial CO2) and with CO2 recirculation (using only the CO2 produced in the fermentation), integrated to ethanol recovery by absorption, were carried out using different feed substrate concentrations. The proposed CO2 stripping model considering the removal of ethanol and water was able to accurately describe the process behavior. The PLS calibration models presented excellent performance in monitoring both, absorption and fermentation. In the fermentation, the FT-MIR/PLS technique was able to provide accurate monitoring of the major components involved in the process (sucrose, glucose, fructose, ethanol, glycerol, and yeast cells). With respect to the evaluation of the absorption process, the MEG was selected as the most appropriate absorbent. The modeling developed was able to accurately describe the absorption behavior. The use of two absorbers, each with 0.80 L of MEG, enabled recovery of up to 93.1% of the ethanol from the stripping gas mixture. The integrated process of extractive fermentation with CO2 stripping and recovery by absorption without CO2 recirculation provided conversion of 300 g L-1 substrate feed, ethanol total production of 135.2 g L-1 and recovery of 91,6 % of the ethanol from the stripping gas mixture. The modelling proposed for both extractive fermentation and absorption recovery adequately described the two stages. In the integrated process with CO2 recirculation, it was possible to consume 280 g L-1 of substrate feed, producing an amount of 126 g L-1 of ethanol (57% higher than in conventional fermentation). The recovery efficiency of ethanol was 98.3%, similar to that found for distilleries considering the loss of ethanol removed by the CO2 produced in the fermenters. The alternative process proposed for ethanol production is promising and could offer improvements for the conventional process, reducing volume of vinasse and energy consumption in the distillation and, consequently, reducing overall costs.