Abstract
An inductively coupled plasma optical emission direct-reading polychromator (ICP-OES) has been evaluated for multiple trace element analysis of foodstuffs. Major trace elements (copper, iron, manganese and zinc) and trace elements (cadmium, lead, nickel, chromium, arsenic and tin) have been considered. For each element, random bias has been exemplified for routine practice, in both sulphuric and hydrochloric acid standard solutions, by the standard deviations for instrument noise, repeatability and reproducibility of measurement. Systematic bias originating from direct interferences has been categorised and relevant corrections have been calculated.To avoid the worst excesses of these interferences and to concentrate the trace elements to levels that permit measurement, a method involving isolation by chelation, extraction, oxidation and solubilisation in fixed acid solution has been used. Six of the elements are amenable to this treatment: copper, manganese, zinc, nickel, cadmium and lead. When this treatment is applied to digests of a selection of foodstuffs no element satisfies the criterion of accuracy required for quantitative analysis in dietary surveys of foodstuffs. The results display comparable confidence intervals to those obtained by flame atomic-absorption spectrometric measurement, however, for the first four of the elements listed.

This publication has 0 references indexed in Scilit: