William Shotyk, Michael Krachler, and Bin Chen measured antimony by ICP-sector-field mass spectrometry in Canadian and European bottled waters and compared PET, glass, polypropylene, and source water. The limit of detection was 0.03 ng/L. Units below are nanograms per litre as printed. Antimony species in the water were not measured. Table 1 uses letter codes. Trade names are not repeated here. The companion commentary in Canadian Chemical News is a separate source page.
Key numbers
Pristine Ontario groundwater is 2.2 ± 1.2 ng/L (n = 34) in the abstract and in most of the article. One results line prints 2.2 ± 1.4 ng/L (n = 34). Twelve Canadian natural waters in PET had an abstract mean of 156 ± 86 ng/L and a body range of 112 to 375 ng/L (n = 21). Three deionized waters in PET had an abstract mean of 162 ± 30 ng/L. The body gives the Ontario deionized PET bottles as 134 to 195 ng/L. An Ontario natural water in polypropylene was 8.2 ± 0.9 ng/L in the abstract and 8.2 ± 0.7 ng/L (n = 7) in the body.
Table 1, commercially bottled mineral water, ng/L:
| Code | Source | Glass | PET, October 2005 | PET, July 2005 |
|---|---|---|---|---|
| A | 3.8 ± 0.9 (n = 5) | 11.5 ± 4.4 (n = 6) | 359 ± 54 (n = 6) | 626 ± 15 (n = 3) |
| B | not available | 84.5 ± 10.2 (n = 6) | 255 ± 20 (n = 6) | |
| C | not available | 26.4 ± 3.1 (n = 6) | 301 ± 43 (n = 6) |
Other German PET brands were 253 to 546 ng/L (n = 5). Three bottles in unidentified recyclable plastic were 32 to 60 ng/L (n = 3). Across 35 European PET brands from 11 countries the median was 343 ng/L. The lowest value, 6 ng/L, was a sample from Poland. Pristine groundwater placed in German PET contained 50 ± 17 ng/L (n = 2) after 37 days in the refrigerator in the introduction, and 59 ± 17 ng/L (n = 2) after 37 days in the refrigerator in the results text. An earlier room-temperature experiment reached 566 ng/L (n = 1) after six months. Both 37-day figures are printed, and both are kept.
Neutron activation of the polymers found 397 mg/kg Sb in a PET water bottle (code A) and 351 mg/kg in a PET cola bottle. Glass bottles contained 7.6 and 10.1 mg/kg Sb. The median lead concentration the authors report for the European PET set is 5 ng/L. SLRS-4 measured 224 ± 12 ng/L Sb (n = 17) against a certified 230 ± 40 ng/L.
Methods (brief)
Clean laboratory, ICP-SFMS, LOD 0.03 ng/L. Leach tests used acid-cleaned containers and PET bottles. Brand identity is coded.
Evidence fitness
The measurements support antimony leaching from PET relative to glass, polypropylene, and source water, in ng/L, without speciation. They do not support a brand ranking.
Limitations
Abstract and body disagree on the groundwater standard deviation (1.2 versus 1.4) and on the polypropylene standard deviation (0.9 versus 0.7). The deionized-water abstract mean and the body range are both printed and are not the same summary. The 37-day refrigerator leach prints as 50 ± 17 ng/L in the introduction and as 59 ± 17 ng/L in the results.
Related evidence
- Bottled drinking water (purified, spring, artesian)
- Mineral water
- Antimony
- Lead
- Think before you drink: antimony contamination from Arctic snow to bottled water
Update history
No substantive edit history is available in this build. The full commit record is available in git.