Oßmann · Water research 2018 · cross-sectional laboratory analytical study · n=32 water samples

Small-sized microplastics and pigmented particles in bottled mineral water.

Cited 989 times in the scientific literature.

Level 5 - mechanism / opinion, no new human data

Level 5 by design analogy (laboratory analytical testing of commercial water samples, no human subjects)

PubMed 29803096 · doi:10.1016/j.watres.2018.05.027 · record verified 2026-08-26

What was done

Researchers analyzed 32 bottled mineral water samples from single-use poly(ethylene terephthalate) (PET), reusable PET, and glass bottles for contamination with microplastics, pigments, and chemical additives. Using aluminum-coated polycarbonate membrane filters and micro-Raman spectroscopy, the study quantified and chemically identified particles down to a minimum size of 1 µm.

What was found

Microplastics were detected across all packaging types, with average concentrations ranging from 2,649 ± 2,857 particles per liter in single-use PET bottles to 6,292 ± 10,521 particles per liter in glass bottles. Plastic bottles predominantly contained PET, whereas glass bottles contained polymers such as polyethylene and styrene-butadiene-copolymer. Reusable paper-labeled bottles contained high levels of pigment particles matching label inks (195,047 ± 330,810 particles per liter in glass; 23,594 ± 25,518 particles per liter in reusable PET). Reusable PET bottles also contained an average of 708 ± 1,024 particles per liter of the plastic additive Tris(2,4-di-tert-butylphenyl)phosphite. Over 90% of all detected microplastics and pigment particles were smaller than 5 µm.

Why it matters

This study pushed the analytical detection limit for bottled water particles down to 1 µm, revealing that the vast majority of microplastic and pigment contamination exists below the 5 µm threshold previously unexamined by earlier surveys. It also demonstrates that glass bottles and bottle-cleaning/labeling processes are major contamination vectors alongside plastic packaging.

Limits

The dataset was limited to 32 water samples, and particle counts showed extremely high standard deviations (often exceeding the mean), reflecting substantial heterogeneity. Biological absorption, toxicity, and human exposure implications were not directly tested.

Cited by