Direct determination of bismuth in urine samples by electrothermal atomic absorption spectrometry: study of chemical modifiers

In this work the performance of some chemical modifiers (no modifier, Pd + Mg, Ir in solution, Ru, Ir or Zr independently as permanent modifier and Ir + Rh permanent) was investigated for the direct determination of Bi in human urine by electrothermal atomic absorption spectrometry (ETAAS). The urine samples were diluted directly in the autosampler cups with one-fold dilution with nitric acid 1% v/v. The obtaining of the permanent modifiers (500 mug of iridium, ruthenium and zirconium independently and 250 mug of Ir plus 250 mug of rhodium) was described elsewhere. Without modifier at all temperatures the signal was negative with a high background. Using Pd + Mg the best temperatures were of 1000 and 2500degreesC, respectively, for pyrolysis and atomization, but the sensitivity was half that when using Ir + Rh. Using ruthenium, iridium or zirconium independently no peak was obtained but a plateau that did not return to baseline even in 15 s of atomization. With iridium applied in solution together with the sample the best pyrolysis and atomization temperatures were of 900 and 2500degreesC, respectively, but the sensitivity was analogous with that obtained using Pd + Mg. Using Ir + Rh, the optimum pyrolysis and atomization temperatures were lower (900 and 1400degreesC), but in this case the best sensitivity was obtained with a characteristic mass of 29.2 pg (recommended of 30 pg for the manufacturer for the metal in aqueous solutions). With this modifier, the obtained peak was very symmetrical, returning to the baseline in 3 s with very low background. The use of low temperatures and a low time of atomization coupled to the use of permanent modifer will certainly extend the lifetime of the graphite tube. Aqueous calibration curves and standard addition were parallel with slopes of 0.042 and 0.044 for aqueous and standard addition, respectively. For this reason, the calibration was carried out with aqueous solutions. The r values of calibration curves were higher than 0.99. Spiked urine samples with 5.0, 10.0, 15.0, 20.0, 25.0 and 30.0 mg l(-1), analysed on different days, presented a recovery between 80.0 and 100.0%. The limit of detection was 50 pg. The lifetime of the tube was more than 500 cycles of atomization.