The remainder of the thawed urine sample was filtered through a 0.2 μm Nylon filter (Whatman GmbH, Dassel, Germany) into a 250 μL polypropylene crimp vial (Agilent Technologies). This filtered sample was analysed directly by anion-exchange HPLC/ICPMS. Additionally, a portion (90 μL) of the filtered sample was removed from the HPLC vial and 10 μL of H2O2 were added, to convert any trivalent- and thio-arsenicals to their pentavalent and/or oxygenated forms, and the mixture was allowed to stand for at least two hours at a temperature > 23°C before analysis by anion-exchange HPLC/ICPMS.
The anion-exchange HPLC conditions (identical for both non-oxidised and oxidised urine samples) were: PRP-X100 column (4.6 mm × 150 mm, 5 μm particles; Hamilton Company, Reno USA) at 40°C with a mobile phase of 20 mM aqueous phosphoric acid adjusted with aqueous ammonia to pH 6 at a flow rate of 1 mL min−1. Injection volume was 20 μL. A carbon source (1% CO2 in argon) was introduced directly to the plasma, as previously described for selenium,14 (link) to provide a 4-5-fold increase in sensitivity. The CO2 was introduced via the T-piece of the high matrix sample introduction kit and the optional gas was set to 0.17 L min−1. Under these chromatographic conditions, As(III) elutes near the void volume, very close to AB and most other cationic arsenic species. This void-volume peak was assigned as AB + As(III) in the non-oxidised sample, and as AB in the oxidised sample (Fig. 1), based on the premise that AB is the only arsenic cation found in significant quantities in urine (see below).15 The total iAs content [As(III) + As(V)] was obtained from the As(V) peak in the oxidised sample. For all HPLC runs, peaks were quantified against the respective standard. Calibration was usually performed in the range 0.10 to 20.0 μg As L−1 (six-point calibration curve); limit of detection was 0.1 μg As L−1 for iAs [As(V) peak], MA, DMA and AB, and the intra-assay coefficient of variation was better than 5 % for all species.
The premise that AB was essentially the only cationic arsenic species in the urine samples was tested by performing cation-exchange HPLC/ICPMS on 188 samples that had shown a significant peak at the void volume during anion-exchange HPLC/ICPMS of the oxidized samples. A Zorbax 300-SCX column (4.6 mm × 150 mm, 5 μm particles; Agilent Technologies) at 30°C was used with a mobile phase of 10 mM pyridine at pH 2.3 (adjusted with formic acid) at a flow rate of 1.5 mL min−1. The injection volume was 10 μL. ICPMS was used as a detector with the settings described above for anion-exchange HPLC/ICPMS.