Otimização de um reator eletroquímico de placas paralelas para remoção do corante reativo azul 5g de efluente sintético

Abstract

Environmental impacts have increased due to the growth of industrial activities. The textile industry stands out due to the fact that it produces high levels of effluents which contain organic pollutants that are difficult to remove, such as dyes. Physical processes along with chemical processes have been used for color removal of these effluents. Therefore, the main contribution of this work was to optimize an electrochemical treatment system for the removal of reactive blue 5G dye from synthetic textile effluent based on the variable hydraulic retention time (HRT) and the electric current density (j). The eletrochemical reactor was adapted to parallel plates using iron electrodes. For this, sets were built with three electrodes each. The electrode which acts as the anode was set up in the center of the other two electrodes (cathode). Tests were carried out based on a Central Composite Rotatable Design (CCRD) with a 22 factorial design, with four factorial trials, four axial points, plus four replications in the central point. The effect of the two variables, HRT and j, on the response variable was evaluated. The color removal percentage on the evaluated conditions ranged from 56.13% to 98.95%. Both parameters, including their interaction, were statistically significant, só it was possible to find a quadratic model of the color removal efficiency as a function of HRT and j. The model was proven valid through the analysis of variance (ANOVA), allowing the generation or response surface and contour curves, as well as enabling the localization of optimal regions for the parameters studied. It was found that a good removal would be reached in the hydraulic retention time range from 20 to 30 min and electric current density range from 12.09 to 20.18 mA.cm-2. Validation of the results was carried out using 20 minutes to hydraulic retention time and 15 mA.cm-2 to electric current density, reaching a mean color removal of 97.18%. Thus, the error associated with model predictions was less than 3%.