Direct cadmium determination in sediment samples by flame atomic absorption spectrometry using multivariate calibration procedures

Abstract

A direct solid sampling procedure by flame atomic absorption spectrometry (DSA-FAAS) has been developed for the determination of cadmium in marine and river sediment samples. Multivariate calibration using the partial least squares (PLS) regression procedure was applied to calibration using dry sediment reference samples. Test-samples were introduced into the flame as a dry aerosol using a T-quartz cell set between the burner and the optical path. Background signals were always low and a characteristic mass of 0.21 ng Cd was found and an absolute limit of detection of 0.25 ng (3s), or 0.25 µg g⁻¹ if a sample mass of 1 mg was used. Using the proposed multivariate calibration procedure the results were considered satisfactory regarding both agreement with the results from reference samples (from 96.8 to 101.7%) and precision (relative standard deviation from 5.5 to 7.0%). With this procedure errors occurring when conventional univariate calibration was used (up to 64%) were reduced to less than 4%.