The main objective of this article is to document and demonstrate Os contamination of natural waters stored in commonly used plastic bottles, which have been acid washed. This observation is a direct result of development of reliable chemical separation and mass spectrometric procedures that have improved detection limits. In the 1990s, a number of attempts were made to determine the osmium concentration and isotope composition of seawater. Its low abundance in seawater and possible multiple oxidation states frustrated these efforts to precisely measure its concentration and isotope composition. Koide et al. (1996) used isotope dilution resonance ionization mass spectrometry to obtain the osmium concentration and isotope composition of seawater. A 25 L sample seawater was spiked with 190 Os tracer, and Os was separated using anion-exchange chromatography and purified using distillation. Sharma et al. (1997) reduced seawater and tracer Os by bubbling SO 2 (g) and then coprecipitating Os wiiron oxyhydroxide in 4-10 L samples. Osmium was oxidized in a sealed glass tube (Carius tube) and extracted using CCl 4 , purified, and measured using negative thermal ionization mass spectrometry (N-TIMS). These techniques required handling of large volume samples (4 to 25 kg). Three new techniques attempted to use smaller volume samples (50 g to 1.5 kg) and to equilibrate tracer and water Os by oxidizing it to a common oxidation state (= OsO 4 ). These techniques indicated that seawater Os concentration was 4-to 5-fold higher than determined by the previous studies. Levasseur et al. (1998)
AbstractA low blank-high yield procedure for the accurate determination of seawater osmium concentration and isotope composition has been developed. The improvement in the detection limit has revealed a subtle temporal increase in the concentration of samples obtained during the GEOTRACES expeditions that were stored in high density polyethylene bottles. This increase in Os concentration is accompanied by significant decrease in the 187 Os/
188Os ratio of the water indicating substantial contamination of waters in two out of three bottles that were investigated; the third bottle, which is not contaminated, shows evidence of Os removal. In comparison, analyses of another aliquot of water stored in a Teflon bottle show no contamination. Samples collected in low density polyethylene bottles during SAFe expedition have also been contaminated. Bottle cleaning with HNO 3 has directly/indirectly contributed to the bottle blank. We also find that a snow-melt sample stored in a Teflon bottle has remained uncontaminated for more than a year, despite having an Os concentration that is ~14 times lower than seawater. We conclude that the low level samples, such as seawater, need to be stored in Teflon bottles for accurate and precise estimate of Os concentration and isotope composition, although polypropylene bottles may be an acceptable alternative. A re-evaluation of previously published data and data obtained in the present study still indicates ...