Cobalt determination in serum and urine by electrothermal atomic absorption spectrometry.

Cobalt determinations in biological fluids are of great interest in biological or toxicological research programs. Cobalturia is often chosen as an indicator for a biological monitoring program in occupational exposure to cobalt dusts. The method described here derives from the IUPAC reference method for nickel determination. It enables cobaltemia and cobalturia to be measured in small samples (1 mL). The mean usual values for cobalt in biological fluids are very low (2.7 nmol L-1 for serum and 6.7 nmol L-1 for urine), and therefore, thus require an analytical procedure with preconcentration and extraction. The sample is mineralized by wet acid digestion. After digestion, inorganic cobalt is extracted in form of ammonium pyrrolidine dithiocarbamate complex into isobutyl methyl ketone and measured in the organic layer by electrothermal atomic absorption spectrometry. The analytical parameters are described in detail. The extraction output is about 99%. The detection limits are 1.93 and 1.89 nmol L-1 for serum and urine, respectively. Sensitivity (expressed as the concentration that gives a 0.044 absorbance) is 3.4 nmol L-1 for serum and 3.3 nmol L-1 for urine. Within-run precision ranged between 3.9 and 2.5% (coefficients of variation) for serum and 4.2 and 1.1% for urine, at 87 and 136 nmol L-1 levels, respectively. Between-run precision ranged between 4.3 and 3.3% (coefficients of variation) for serum and 4.2 and 2.3% for urine, at 87 and 136 nmol L-1 levels, respectively. At very low concentration, 5.7 nmol L-1 for serum and 2.5 nmol L-1 for urine, the between-run precision is, respectively, 19.5 and 28%. Linearity is effective between 0 and 272 nmol L-1. Interferences and matrix effects are negligible for urine, serum, or plasma samples without hemoglobin. The method is easily applicable for routine determinations.