Fast automated determination of toxicologically relevant arsenic in urine by flow injection-hydride generation atomic absorption spectrometry

Lower results for arsenic in urine are encountered when potassium iodide-ascorbic acid is used as a pre-reductant in flow injection-hydride generation atomic absorption spectrometry (FI-HGAAS). For the determination of toxicologically relevant arsenic in urine, i.e. the sum of inorganic As(III) and As(V) and its organic metabolites (monomethylarsonate and dimethylarsinate) has been elaborated a simple analytical procedure: ten-fold diluted urine is treated with l-cysteine (1%, m/v) in 0.03 mol 1−1 hydrochloric acid medium for 1 h, and then a 500 μl aliquot is directly injected into an automated, commercially available FI-HGAAS system. The effects of most critical instrumental and chemical parameters have been studied and optimum conditions are recommended, comprising of relatively low sample and carrier acidities, 0.03 and 0.01 mol 1−1 HCl, respectively, higher sodium tetrahydroborate levels (0.5%, m/v), lower flow rates of reagents than that typically recommended for this system to ensure longer reduction and residence times in the gas-liquid separator, and integrated absorbance measurements for eliminating the residual kinetic effects in hydride generation and transport of individual arsenic species. The recoveries of spiked As(III), As(V), MMA and DMA in urine are 100%, 100%, 96% and 95%, respectively. This procedure has been validated by analysis of several certified reference materials and collaborative study samples, the results being in good agreement with the certified values. The limit of detection (3σ) is 0.1 μg 1−1 in 1 : 10 diluted urine, corresponding to 1 μg 1−1 of As in undiluted urine. The relative standard deviation is 2% and 7% (within-run and between-run in three successive days, respectively) at 3 μg 1−1 (1.5 ng) As level. The sampling frequency is up to 50–55 measurements per hour.