Comparison of atomic absorption and fluorescence spectroscopic methods for the routine determination of urinary arsenic.

Objective: Interpretation of urinary arsenic measurements is sometimes difficult because of the absorption of seafood that contains trimethylated arsenic forms (arsenobetaine and arsenocholine). The objective of this study was to develop a rapid and robust technique for the measurement of the sum of inorganic arsenic metabolites.

Methods: Measurement of arsenic was performed in urine after hydride generation in acid medium. Using atomic fluorescence spectrometry (AFS) as the detection system, we developed a rapid (one determination in less than 2 min) technique using 50 microl urine without pre-treatment. Standardisation was done externally with a mixed standard solution containing inorganic trivalent arsenic (As(i) (III)), inorganic pentavalent arsenic (As(i) (V)), monomethylarsonic acid (MMA) and dimethylarsinic acid (DMA) (15/15/12.5/57.5).

Results: Samples distributed in the frame of an international comparison programme were used to assess accuracy of the AFS procedure, which gives a linear response up to 50 microg/l and proves more precise [coefficient of variation (CV)< 4% at 5 microg/l] and sensitive than the atomic absorption spectroscopy (AAS) technique using a quartz cell. An additional adaptation that allows the detection of non-directly reducible arsenic forms has also been validated for samples with high arsenic concentrations.

Conclusions: The present study demonstrates the superiority of AFS over atomic absorption spectrometry (AAS) in arsenic determination and the interest of online mineralisation prior AFS detection for the determination of arsenic concentration in urine.