Studies have chemically identified plastic particles in bottled water, including water sold in glass bottles. The particles can arise from closures or processing as well as the bottle body. Counts vary with the sampled products, particle-size window, and analytical method. The reviewed findings establish occurrence and possible entry points, without supplying a universal concentration or a measured human health effect.12
What this page covers
This page compares two complementary designs. A process-tracing study sampled mineral-water production from groundwater to the finished bottle. A retail survey analyzed water sold in PET, recycled PET, and returnable glass. The first can investigate where particles enter a particular production system. The second describes what was found in selected finished products. Their different sampling designs prevent a direct material ranking across studies.12
The term bottled water describes the beverage and its full packaging system. A glass body can still use plastic sealing material. Finding polyethylene in glass-packaged water therefore does not mean the analyst identified glass as plastic. It raises a source question about closures or other contact materials, which must be investigated separately.12
What the studies found
Scroll horizontally for methods and measurement limits.
| Study and sample | Reported particles | Identification and size context | Interpretation |
|---|---|---|---|
| Process tracing: filled/capped reusable glass bottles1 | 317 ± 257 confirmed plastic particles/L | FPA-FTIR imaging; particles ≥11 µm | Source-specific result at studied bottling facilities, with blank-based thresholds. |
| Retail survey: single-use PET-packaged water2 | 5.09 ± 3.28 confirmed plastic particles/L | Raman confirmation of detected particles; predominant sizes 20-100 µm | Mean ± SD of pooled analytical samples after polymer-specific control subtraction. |
| Retail survey: returnable-glass-packaged water2 | 8.65 ± 5.39 confirmed plastic particles/L | Same retail-study Raman workflow | Plastic can enter glass packaging; no universal glass-versus-PET ranking follows. |
Tracing entry during bottling
The process study collected raw and treated water, bottle-washing solutions, cleaned bottles, and finished filled/capped bottles. It also sampled filled bottles before capping at one facility. A marked increase after closure application, together with chemical resemblance to polyethylene seal material, supported closure abrasion as an important source in the sampled systems.1
The workflow used infrared imaging, automated spectral classification, and manual review. It tested size detection and particle recovery rather than assuming that filter pore size alone defined the observed range. Procedural blanks were present, and separate detection thresholds were calculated for different sample types. A value below such a threshold means the method could not distinguish a count confidently from its background. It does not establish that every particle was absent.1
The study also identified nonplastic material and discussed contamination associated with the sampling apparatus. This distinction is important because a particle seen on a filter can be cellulose, mineral material, or another substance. Only the chemically assigned plastic group belongs in the microplastic concentration used here.1
What the retail survey measured
The retail survey pooled bottle contents by product and market into analytical samples. Analysts located particles across the filter and used Raman spectra to assess each detected particle’s identity. They subtracted control findings according to the polymer identified, rather than counting every observed object as plastic. Fragments predominated, and the most common observed size classes were above the smallest particles detectable by Raman in principle.2
PET appeared prominently in PET-packaged samples. Polyethylene, including material with additive-related signals, was prominent in the recycled-PET and glass groups, consistent with possible closure contributions. That chemistry can suggest a source but does not by itself isolate where a particle entered the water. The retail design did not follow every bottling stage as the process study did.2
Pooling also changes interpretation. An analytical mean over pooled samples is not a guarantee for a particular bottle purchased later. Variability can reflect product, packaging, processing, or market differences. The paper’s recycled-PET mean differs slightly between the abstract and results text, so this page does not repeat a precise recycled-PET concentration until that discrepancy is resolved.2
Why the concentrations differ
The process and retail studies sampled different systems and used different identification workflows. Their size coverage, recovery, filter handling, spectral classification, and blank correction also differ. A lower concentration in one study cannot demonstrate that its packaging material inherently sheds less across all uses. It may reflect real product differences, analytical differences, or both.12
Neither study provides a complete nanoplastic count. An instrument’s ability to identify a small reference particle is not equivalent to complete capture, detection, and chemical assignment of every environmental particle at that size. Counts per liter must stay attached to the stated observation window and should not be combined with smaller-particle estimates from another technique as though they formed a single validated measurement.12
Regulatory status
The EU decision adopts a measurement methodology for water intended for human consumption, while California’s supplied methods describe analysis in treated drinking water. Their scope, size coverage and quality-control requirements matter when comparing these papers.345 The supplied California handbook copy is dated before its adoption meeting; its proposed monitoring calendar is not presented as the verified current timetable.6 These documents are not used here as a health-based particle ceiling or as proof that a particular bottled-water product falls within a public-water monitoring program. Read the scope notes.
What remains uncertain
These studies do not establish an internal human dose, a clinical effect, or a safe or unsafe count for bottled water. Their most direct contribution is evidence of occurrence and, in the process study, a supported entry mechanism in selected facilities. Broader representative sampling and compatible validated methods are needed to determine how well a given concentration describes other products, locations, and particle sizes.12
References
- From the Well to the Bottle: Identifying Sources of Microplastics in Mineral WaterResearch article
- Characterization of microplastics in water bottled in different packaging by Raman spectroscopyResearch article
- Commission Delegated Decision (EU) 2024/1441: methodology to measure microplastics in water intended for human consumptionLegal instrument
- SWB-MP1-rev1: Infrared spectroscopy of microplastic particles in drinking waterOfficial method
- SWB-MP2-rev1: Raman spectroscopy of microplastic particles in drinking waterOfficial method
- California microplastics drinking-water policy handbook, supplied August 9 versionPolicy document