Multivariate optimization of an aqueous two-phase extraction for determination of cadmium and manganese in food sample

Abstract This study proposes a methodology of determination of cadmium and manganese in food samples using an aqueous two-phase system without extracting agent with F AAS detection. In order to choose the best system for preconcentration, an experimental design with Doehlert matrix was employed, allowing the determination of the system composed by PEO400 + K2HPO4 + H2O with the best extraction efficiency. This same study verified that pH does not influence the extraction percentage. Still in the optimization stage, the best conditions were tie-line length = 53.3%w; mass ratio between bottom and top phases = 0.029; and extraction time = 18hs. The method reached values of LOD (0.45 µg.Kg−1 for Cd(II) and 6.4 µg.Kg−1 for Mn(II)), LOQ (1.5 µg.Kg−1 for Cd(II) and 21 µg.Kg−1 for Mn(II)) and enrichment factor (16 for Cd(II) and 45 for Mn(II)) suitable to determine both analytes in food samples. The intra and inter-day relative standard deviation ranged from 3.28 to 4.37%, and from 5.05 to 6.60% for Cd(II) and Mn(II), respectively. The analysis of the NIST 1575a certified sample had a recovery of 98.2% for cadmium, and NIST 1515 presented recovery of 99.2% for manganese, showing high accuracy. The method was applied to samples of green tea and soluble coffee, and the recovery tests obtained results ranging from 94.0 to 110%. Therefore, the proposed methodology is efficient and without using an extractor, being a highly accurate and environmentally safe alternative for the determination of trace levels of metal in food samples.