Determination of arsenic species in human urine using HPLC with on-line photooxidation or microwave-assisted oxidation combined with flow-injection HG-AAS

Abstract

An improved analytical procedure is presented for the separation and simultaneous determination of hydride-forming (toxic) and not hydride-forming (non-toxic) arsenic species in human urine. Separation was performed by cation-exchange chromatography using a new solid phase type based on the continuous bed chromatography (CBC) technology. This column permits by a factor of 4 higher flow rates than conventional columns resulting in a drastical reduction of retention times without any loss of resolution. Using this type of column, arsenobetaine (AsBet), arsenocholine (AsChol), and dimethylarsinic acid (DMA) were separated from the more toxic arsenic species arsenous acid (As(III)), arsenic acid (As(V)), and methylarsonic acid (MA) within only 4 min. The HPLC system was coupled via a flow injection system and either a UV or a microwave (MW) reactor to the HG-AAS instrument. UV photolysis and MW digestion were used to transform AsBet and AsChol to hydride-forming species and to make them accessible to HG-AAS. UV photolysis turned out to be more suitable for this application than MW digestion, because the latter technique led to peak broadening and poorer performance. The described procedure was applied to the determination of arsenic species in urine samples of non-occupationally exposed persons before and 12 h after seafood consumption. Detection limits were about 1 μg/L for each arsenic species. After consumption, the AsBet and DMA excretion increased by at least a factor of 150 for AsBet and by a factor of 6 for DMA, respectively, while the excretion of the other species did not increase significantly. This invalidates the use of total urinary arsenic as well as total hydride-forming arsenic as an indicator for exposure to inorganic arsenic.