Bovine liver sample preparation and micro-homogeneity study for Cu and Zn determination by solid sampling electrothermal atomic absorption spectrometry

This work describes a systematic study for the bovine liver sample preparation for Cu and Zn determination by solid sampling electrothermal. atomic absorption spectrometry. The main parameters investigated were sample drying, grinding process, particle size, sample size, microsample homogeneity, and their relationship with the precision and accuracy of the method. A bovine liver sample was prepared using different drying procedures: (1) freeze drying, and (2) drying in a household microwave oven followed by drying in a stove at 60 degrees C until constant mass. Ball and cryogenic mills were used for grinding. Less sensitive wavelengths for Cu (216.5 nm) and Zn (307.6 nm), and Zeeman-based three-field background correction for Cu were used to diminish the sensitivities. The pyrolysis and atomization temperatures adopted were 1000 degrees C and 23 00 degrees C for Cu, and 700 degrees C and 1700 degrees C for Zn, respectively. For both elements, it was possible to calibrate the spectrometer with aqueous solutions. The use of 250 mu g of W+200 mu g of Rh as permanent chemical modifier was imperative for Zn. Under these conditions, the characteristic mass and detection limit were 1.4 ng and 1.6 ng for Cu, and 2.8 ng and 1.3 ng for Zn, respectively. The results showed good agreement (95% confidence level) for homogeneity of the entire material (> 200 mg) when the sample was dried in microwave/stove and ground in a cryogenic mill. The microsample homogeneity study showed that Zn is more dependent on the sample pretreatment than Cu. The bovine liver sample prepared in microwave/stove and ground in a cryogenic mill presented results with the lowest relative standard deviation for Cu than Zn. Good accuracy and precision were observed for bovine liver masses higher than 40 Vg for Cu and 30 mu g for Zn. The concentrations of Cu and Zn in the prepared bovine liver sample were 223 ing kg(-1) and 128 mg kg(-1), respectively. The relative standard deviations were lower than 6% (n = 5). The accuracy of the entire procedure was checked with bovine liver from NIST (1577b) and determination of Cu and Zn using flame atomic absorption spectrometry after microwave-assisted sample digestion. (c) 2005 Elsevier B.V All rights reserved.