Mixture Design Optimization of an Analytical Procedure for Iron Extraction and Determination From Cassava Leaves by Slurry Sampling Flame Atomic Absorption Spectrometry

This paper applies statistical simplex-centroid design to mixture modeling for optimization of the liquid phase composition of cassava slurry leaves in the development of an analytical procedure for iron determination using flame atomic absorption spectrometry (FAAS). This procedure is based on a slurry formation from powdered cassava leaves and a liquid mixture composed of HNO3, HCl, and H2O2 after an ultrasonication process. A quadratic model fitted to the analytical response shows the existence of an experimental region, characterized by low proportions of H2O2, for which higher responses are obtained. The proposed procedure allows the determination of iron in the cassava leaves with a detection limit of 1.1 mu g g(-1). The precision expressed as relative standard deviation (%RSD, n = 10) was 1.5% for iron concentrations of 25 mu g g(-1). The developed procedure was validated by the comparison of results obtained from the application of the digestion procedure and the analysis of certified reference materials: Apple leaves (NIST 1515). Results found were in agreement with the certified values. The proposed method was applied for the determination of iron in four samples of cassava leaves acquired in markets of Feira de Santana City, Brazil. The concentration of iron found in the cassava leaves varied from 250.8 +/- 0.7 to 283.4 +/- 0.7 mu g g(-1).