Determination of antimony (III), antimony (V), selenium (IV) and selenium (VI) in natural waters by hydride generation atomic absorption spectrophotometry combined with a cold trap

Antimony and selenium in water samples are reduced to stibine and hydrogen selenide, respectivery, by an addition of Na[BH₄]. The stibine and hydrogen selenide generated are once collected by a liquid-nitrogen cold-trap, and then led, by heating the trap, into a quartz furnace for atomic absorption spectrometry. Sb(III) is selectively reduced by Na[BH₄] at pH 6, but Sb(V) not. The latter is reducible under highly acidic condition including KI as a prereducting reagent. According to the fact, Sb(III) is determined through the reduction in 1.6 M potassium citrate buffer of pH 6, and total antimony Sb(III+V) is determined in 12 M HCl/2 M KI solution of pH 1, Sb(V) is then determined by the difference. Similarly, Se(IV) is determined by the reduction with Na[BH₄] under highly acidic condition, and total selenium Se(IV+VI) determined through successive reductions in KBr/HCl solution and in Na[BH₄] solution, and Se(VI) determined from the difference. Detection limits of this method are: 16 ppt for the total antimony: 1 ppt for Sb(III); 2 ppt for Se(IV); and 2 ppt for the total selenium. The relative standard deviation is within a few percent for both antimony and selenium determinations. The reproducibility could be improved by removing water by using an auxiliary dry-ice/2-propanol trap. The use of EDTA as a masking reagent is effective for selenium determination. Sb(V) ranging 70–460 ppt, Se(IV) 3–30 ppt and Se(VI) 3–60 ppt were actually observed in natural waters, but Sb(III) was not.