Validação de método eletrodo íon seletivo para determinação de fluoretos em água potável

Abstract

The present work had as aim the adequation and validation of an analytical method of determination of fluoride concentration in drinking water, using ion selective electrode, to be applied in a pharmaceutical industry. The analytical method was developed based on the method proposed by APHA 1999 with adjustments for its optimization and viability. The determination is performed using an ion selective electrode (ISE) meter, which is connected to a ion selective electrode, measuring the potential signal (mV), a calibration curve prepared with standard solution is applied as a reference, so the sample concentration can be be determined. Working Interval was determined between 10–190 % of the 1.5 mg L-1 specification. Selectivity results showed a recovery of 100 % for Fe3+ and 98.7 % for Al3+, the main conflicting metals in the quantification. The Buffer Effect Study concluded that the angular coefficients are equal to the 5 % significance level, in this way the buffer isnt an interference in the analysis. The angular coefficients obtained in The Matrix Effect Study were equal to the 5 % significance level, so the sample quantification doesnt suffer matrix interference. Linearity showed a correlation coefficient (R) of 0.9996 and a coefficient of determination (R2) of 0.9991. The Cochran test showed that the variances are homoscedastic, with a confidence level of 95 %. No outliers were found by the Grubbs test in the results. The combination of these results demonstrate the linearity of the method in the working range. The applied method met the acceptance criteria for Precision (Repeatability and Reproducibility). During the execution of Accuracy the results showed recoveries and relative standard deviation within the acceptable range. The Robustness was evaluated by the the signal stabilization time, wich had a variation in the reading of the first sample (DPR of 3.93%). The result may be related to the signal stabilization time generated by the first sample, since all other samples remained stable. It concludes that the developed and evaluated method meets the proposed specifications for the parameters and can be considered validated and can be use for determination of fluoride in drinking water.