Twelve colorless PET bottled waters bought in Monterrey, Mexico, were analysed for total antimony by hydride generation atomic fluorescence spectrometry, with ion-exchange preconcentration for the water and microwave acid digestion for the plastic. Antimony in the water as purchased was reported below the EU drinking-water value of 5 μg/L, while the PET itself contained on the order of 73 to 111 mg/kg. A factorial migration test in ultrapure water found temperature to be the strongest factor, with the highest release at 75 °C, pH 7, and 5 days. Under that worst-case condition the twelve bottles released 1.1 to 18.5 μg/L, and the modelled chronic daily intake for one coded bottle exceeded the USEPA reference value of 400 ng/kg/day. Individual brand names are not used here; bottles are identified only by the paper’s PET codes.
Key numbers
| Analyte | Matrix | Value | Unit | Statistic | N | Where |
|---|---|---|---|---|---|---|
| Sb | drinking water regulatory limit | 5 | μg/L | regulatory_limit | p.PDF p.1 (printed 511) | |
| Sb | drinking water regulatory limit | 6 | μg/L | regulatory_limit | p.PDF p.2 (printed 512) | |
| Sb | drinking water regulatory limit | 40 | mg Sb/kg | regulatory_limit | p.PDF p.2 (printed 512) | |
| Sb | drinking water regulatory limit | 400 | ng/kg/day | regulatory_limit | p.PDF p.1 (printed 511) | |
| Sb | water, HG-AFS before preconcentration | 0.113 | μg/L | lod | p.PDF p.4 (printed 514) | |
| Sb | water after 20-fold Dowex preconcentration | 5.65 | ng/L | lod | p.PDF p.4 (printed 514) | |
| Sb | PET bottled drinking water, as purchased | 0.28-2.30 | μg/L | range | 12 brands | p.PDF p.1 (printed 511) |
| Sb | PET bottled drinking water, as purchased | 1.05 | μg/L | mean | 12 | p.PDF p.4 (printed 514) |
| Sb | PET bottle material | 73.0-111.3 | mg/kg | range | 12 | Table 2 p.PDF p.1 (printed 511) |
| Sb | PET bottle material | 87.6 ± 3.9 | mg/kg | mean | 12 | Table 2 p.PDF p.4 (printed 514) |
| Sb | bottled water PET-1, as purchased | 0.28 ± 0.06 | μg/L | mean | 3 | Table 2 p.PDF p.4 (printed 514) |
| Sb | PET bottle PET-1 | 81.2 ± 2.9 | mg/kg | mean | 3 | Table 2 p.PDF p.4 (printed 514) |
Methods (brief)
HG-AFS (AF-640A), stibine generation with 0.7% NaBH4 and 10% HCl after KI/ascorbic acid pre-reduction. Water: 100 mL through 1.5 g Dowex 1X8-100, eluted with 5 mL 2.0 M HNO3. PET: about 0.1 g, 5 mL HNO3:HCl (4:1), 180 °C for 15 min, made to 50 mL. Migration: four 5 × 4 cm pieces (about 2.1 g) in 100 mL, pH 3 or 7, 25 or 75 °C, 5 or 15 days, duplicated. Spike recoveries 91.0-118.0% for PET (50 mg/kg spike) and 96.5-101.5% for water (400 and 800 ng/L).
Detection and reporting: HG-AFS before preconcentration: LOD 0.113 μg/L, LOQ 0.377 μg/L. After 20-fold preconcentration: LOD 5.65 ng/L, LOQ 19.01 ng/L.
Basis: Total Sb. Bottled-water concentrations are in the water as purchased. PET content is mg/kg plastic. Migration tests are release into ultrapure or deionized water, not total content.
Evidence fitness
This source can support: Total Sb in Mexican PET bottled water as purchased; Total Sb in the PET plastic; Sb migration into water under stated temperature, pH, and time; Comparison with EU 5 μg/L, USEPA 6 μg/L, and USEPA RfD 400 ng/kg/day as cited by the paper.
It cannot support: Sb(III) versus Sb(V) speciation in the waters; Ordinary room-temperature storage kinetics for every brand (Fig. 3 values are 75 °C); Brand-identified concentrations; A certification limit.
Limitations
Table 2 includes PET-4 at 0.07 ± 0.01 μg/L, which sits below the abstract and results range of 0.28-2.30 μg/L; both are transcribed. Fig. 3 brand-by-brand leaching values are not printed as a table, so only the stated range 1.1-18.5 μg/L and the PET-2 mean 5.1 ± 0.2 μg/L are entered. One results sentence gives CDI in μg/kg/day while Table 4 and the abstract use ng/kg/day; table units were used. Reported Sb is total after pre-reduction, not Sb(III) versus Sb(V) in the samples. Worst-case migration used ultrapure water at 75 °C, not the commercial water under ordinary storage.
Related evidence
Update history
No substantive edit history is available in this build. The full commit record is available in git.