OTIMIZAÇÃO E VALIDAÇÃO DE METODOLOGIAS ANALÍTICAS ALTERNATIVAS PARA A DETERMINAÇÃO DEMn,Zn,Mg E Fe EM AÇÚCAR MASCAVO UTILIZANDO A ESPECTROMETRIA DE ABSORÇÃO ATÔMICA EM CHAMA (FAAS)

Abstract

The brown sugar is a result of evaporation and concentration of the juice of freshly extracted sugar cane to the crystallization point of sucrose. This is characterized by high concentration of sugars, mineral content and trace amounts of vitamins, giving it a high nutritional and medicinal value. Nevertheless, there is still little research focused on identification and quantification of essential minerals that food. This scarcity can be related to difficulties in the treatment of the sample, which requires a pre-treatment involving the destruction of most of the organic matter, which is usually done by standard sample preparation methods such as digestion by dry or digestion by wet. The objective of this work was to optimize and validate sample preparation alternative analytical methodologies for the determination of Zn, Mn, Mg and Fe in brown sugar by FAAS using the Fenton reaction and UV photo-oxidation and acid digestion assisted by ultrasound. To optimize the Fenton reaction as sample preparation for determination of Mn was used a factorial design 22, that were evaluated concentrations H2O2 and Fe. For the optimization of UV photo-oxidation reaction for the determination of Zn and Mg was applied to a factorial design 24, which evaluated the concentration of H2O2, the solution pH, the concentration of the buffer solution and the UV irradiation time, using as the absorbance response at 400 nm of brown sugar solutions. And for acid digestion assisted by ultrasound as sample preparation for the determination of Fe a univariate optimization which evaluated the weight of the sample and the volume of aqua regia was applied. To confirm the reliability of the developed methods was performed to validate the same, which evaluated the linearity parameters, limits of detection (LD) and quantification (LQ), precision and accuracy, from the application of different statistical tests. The used analytical curves were external standard for all metals analyzed, as they showed good linearity in the concentration range investigated, with values of Fregression >> Fcritical and no lack of linear model fit. The limits of quantification (LQ) for all metals are below the first point of the calibration curve. The accuracy of the sample preparation methods proposed was evaluated by applying a comparison of test-F of the variances weighted the results obtained by the application of alternative methods proposed with the variances weighted the results of the methods adopted as reference (digestion dry and humid digestion). The Fenton reaction has been compared with the two methods adopted as reference, and the reaction UV photo-oxidation and acid digestion assisted by ultrasound were compared only with the digestion dry. The accuracy of the Fenton reaction was also assessed by the standard deviation (% RSD) obtained under repeatable conditions. The Fcalculated values for Mn analysis from the Fenton reaction of Mg from the UV photo-oxidation reaction and Fe from the acid digestion were assisted by ultrasound below the Fcritical p > 0.05 confirming no significant differences in accuracy between the comparative methods. The value for Zn Fcalculated from the UV photo-oxidation reaction was superior to Fcritical p < 0.05 suggesting that there are significant differences in accuracy. To verify that the photo-oxidation UV would be a more accurate method than the digestion dry was applied again the comparison of test-F of the variances weighted tail. The results showed that the UV photo-oxidation is lower than the standard weighted digestion dry deviations, indicating that it presents better accuracy for determination of Zn brown sugar samples adopted as the reference method. The value of RSD (%) for Mn analysis from the Fenton reaction was lower than recommended by the AOAC (RSD (%) < 7.3). The accuracy of the alternative sample preparation methods proposed was evaluated by comparison with the reference methods (digestion dry and humid digestion), applying a paired test-t and linear regression analysis. Again, the Fenton reaction was compared with the two methods adopted as a reference and UV photo-oxidation reaction and the acid digestion assisted by ultrasound were compared by digestion with only dry. The accuracy of the Fenton reaction was also assessed by standard addition by recovery experiments. The results for the paired test-t showed no significant differences between the methods compared, because tcalculated < tcritical with p > 0.05. Moreover, the high Fregression values obtained in the linear regression analysis confirmed that linear models for Mn, Zn, Mg and Fe are suitable, and the content of these metals exhibit a linear correlation between the sample preparation methods evaluated. The Mn analysis of recovery rates from the Fenton reaction (90.7 to 97.5%) were within the limits recommended by the AOAC (80-110%). These results indicate that the sample preparation methodologies proposals presented adequate accuracy. Thus, it is suggested that the methodologies developed can be applied in the quality control of brown sugar with advantages over conventional sample preparation methods such as simplicity, speed, low cost and reduction of toxic waste in addition to offering greater security to the analyst.