Desenvolvimento do método potenciométrico para determinação de fluoretos em amostras de banho electrolítico

Abstract

The composition of the electrolytic bath is a key criterion in the control of electrolytic cells or either in maximizing their efficiency for the production of aluminum. However, as the process consumes huge amounts of energy it is also important to lower the cell temperature, as a way to improve your performance. This is achieved through the addition of additives such as aluminum trifluoride. The use of additives also causes less energy to be lost to the and a lower resistance of the electrolytic bath to the passage of the electric current, hence the concentrations of all additives should be monitored at regular intervals. For this purpose the X-ray diffraction method is used which confronts the volumetric method. The present work aims to extend and validate the potentiometric method in the determination of excess of aluminum trifluoride, present in the electrolytic bath, determinable by the method of X-ray diffraction that confronts the volumetric method under production conditions. The study was developed with a view to replacing the volumetric method as a way to improve working conditions as the latter method uses thorium nitrate (a substance radioactive) as a titrant in the analysis of the electrolytic bath. For the validation of the potentiometric method, the determination of the relative errors was made and the comparison of the results thus obtained with the results obtained by the volumetric method, using the test t paired. Statistically it was found that at a 95% confidence level, there is a difference significant difference between the results. However, given that the relative errors obtained by the method potentiometric are between 0.46 and 4.66%, this method can be used for bath analysis electrolytic. The accuracy of the volumetric and potentiometric methods was verified through the analysis of standards (electrolytic bath), provided by Mozal's spectroscopy laboratory. It was found, from relative errors, that it was very good for all samples with relative errors ranging between 0.28 - 3.51 and 0.46 - 4.66% respectively. In terms of accuracy, the calculated% RSD values ​​demonstrate that it was very good for all samples analyzed, ranging from 0.86 - 2.13% for volumetry (except sample A134 with 14.55%) and 0.32 - 2.20 for potentiometry.