Multivariate optimization of simple procedure for determination of Fe and Mg in cassava starch employing slurry sampling and FAAS

•A slurry sampling procedure was developed for cassava starch analysis.•Iron and magnesium were determined by fast sequential FAAS.•The analytical variables were optimized using a Box-Behnken design multivariate.•The procedure is simple, low-cost, precise, accurate and reliable.

A slurry sampling procedure has been developed for Fe and Mg determination in cassava starch using flame atomic absorption spectrometry. The optimization step was performed using a univariate methodology for 200 mg samples and a multivariate methodology, using the Box-Behnken design, for other variables, such as solvent (HNO3:HCl), final concentration (1.7 mol L−1) and time (26 min). This procedure allowed determination of iron and magnesium with detection limits of 1.01 and 3.36 mg kg−1, respectively. Precision, expressed as relative standard deviation (%RSD), was of 5.8 and 4.1% (n = 10) for Fe (17.8 mg kg−1) and Mg (64.5 mg kg−1), respectively. Accuracy was confirmed by analysis of a standard reference material for wheat flour (NIST 1567a), which had certified concentrations of 14.1 ± 0.5 mg kg−1 for Fe and 40 ± 2.0 mg kg−1 for Mg, and the concentrations found using proposed method were 13.7 ± 0.3 mg kg−1 for Fe and 40.8 ± 1.5 mg kg−1 for Mg. Comparison with concentrations obtained using closed vessel microwave digestion was also realized. The concentrations obtained varied between 7.85 and 17.8 mg kg−1 for Fe and 23.7-64.5 mg kg−1, for Mg. The simplicity, easily, speed and satisfactory analytical characteristics indicate that the proposed analytical procedure is a good alternative for the determination of Fe and Mg in cassava starch samples.