The work was carried out by Edgars Kuka and colleagues at the Laboratory of Wood Degradation and Protection of the Latvian State Institute of Wood Chemistry in Riga, funded by the Latvian Council of Science. Their motivation stems from a well-recognised gap in microplastics research: while methods exist to quantify microplastic shedding from textiles during washing, airborne microplastics from waste facilities, and tyre abrasion, most plastic products that degrade outdoors have never been assessed as microplastic sources. Building materials in particular have received little attention, despite reports that plastic sheets, paints, fibre-reinforced materials, insulation, and composites can emit significant quantities of particles. Wood plastic composites, a rapidly growing market segment, had shown indications of potential release in earlier work, but the amounts had never been measured.

The core problem the team set out to solve is that established weathering standards, such as ISO 4892, ASTM G154, and EN 927, and the commercial devices built around them, are designed to evaluate changes in material properties like colour, gloss, and surface integrity. They were never intended to capture the particles that break away from a degrading surface. Previous attempts to estimate microplastic release indirectly, for example through surface roughness measurements, cannot account for processes such as volatile product formation and chemicrystallisation, which embrittle the surface without producing particles. Direct gravimetric approaches, meanwhile, are confounded by the fact that photodegradation also generates volatile and water-soluble products that escape as mass loss without ever becoming particles.

The new process design is built on the QUV Accelerated Weathering Tester, a widely used commercial device. The researchers fitted the chamber with a cascade of sieves with mesh sizes of 4000, 500, 150, 75, and 20 micrometres, through which all run-off water from the spray cycles is directed, along with a water recirculation loop of roughly 25 litres topped up as needed. Each artificial weathering cycle lasts one week and comprises 140 hours of ultraviolet irradiation from UVA-340 lamps, which mimic sunlight in the 295 to 365 nanometre range, and 4 hours of water spray at a flow rate of 6 to 7 litres per minute, with the chamber held at 60 degrees Celsius. Specimens with a total exposed area of 0.23 square metres were subjected to eight such cycles over two months. Particles larger than 20 micrometres are retained on the sieves and weighed after drying, with results expressed as mass of collected microplastics per square metre of exposed surface. For the fraction below 20 micrometres, a portion of the circulating water was concentrated by evaporation and the dried solids analysed by pyrolysis gas chromatography-mass spectrometry.

The model materials were recycled polypropylene supplied by a local recycler, and a wood plastic composite made from 60 weight percent of the same polypropylene and 40 weight percent pine wood particles of 400 to 1000 micrometres, compounded without additives on a two-roll mill and compression moulded into 1-millimetre-thick sheets. Two identical composite batches were tested to gauge reproducibility. The results were unambiguous: the composite began shedding measurable particles during the second weekly cycle, exceeding 1 gram per square metre cumulatively, and reached approximately 4.4 grams per square metre after four cycles and 9.4 grams after eight. The recycled polypropylene, by contrast, released only about 0.3 grams per square metre over the full test. The difference between the two composite batches never exceeded 7 percent, which the authors describe as very good repeatability for the tested period.

Microscopic examination of the weathered surfaces explains this counterintuitive outcome. Scanning electron microscopy revealed intense cracking in regions where wood particles lay beneath the polymer layer. The researchers attribute this to repeated swelling and shrinking of the hygroscopic wood particles as moisture fluctuates between UV irradiation and water spray phases, generating internal stresses in the surrounding polymer matrix. When these stresses exceed the strength of the polymer, microcracks form and propagate, eventually releasing particles. Weak adhesion between the nonpolar polymer and the polar wood, a consequence of chemical incompatibility, likely aggravates the effect. In the pure polypropylene, degradation followed a slower route: microvoids appeared after the first cycle, followed by microcracking driven by chemicrystallisation, in which chain scission during photodegradation increases crystallinity and embrittles the surface. The crack networks formed hierarchical patterns, with wide deep cracks enclosing segments of about 100 micrometres, within which finer networks created segments of roughly 10 micrometres.

The collected particles themselves were overwhelmingly small: 99 percent were below 500 micrometres, with most in the 75 to 500 micrometre range, and the proportion in the 20 to 75 micrometre fraction increased as weathering progressed. Pyrolysis GC-MS confirmed the particles were polypropylene, showing a strong peak for the indicator ion 2,4-dimethyl-1-heptene. Notably, the pyrograms of the concentrated run-off water showed no such indicator, suggesting that no microplastics smaller than 20 micrometres passed through the finest sieve during the test period, although photodegradation products such as ketones, acids, and aldehydes were detected in the water. The authors caution that a 5-micrometre protective filter in the system may have removed some particles in the 5 to 20 micrometre range, and that the setup, as currently designed, cannot reliably detect very small releases, with blank runs showing weighing fluctuations of around plus or minus 0.05 grams per square metre.

A key finding with implications well beyond the laboratory concerns the difference between real and artificial microplastics. The team also prepared reference particles by cryogenic milling of the same polypropylene, a common approach for generating test materials in toxicological studies, and exposed some of them to ultraviolet radiation. Chemically, the weathered composite particles showed elevated hydroxyl, carbonyl, and carbon-oxygen indices characteristic of photo-oxidation, but their carbonyl index remained fairly stable at 0.7 to 1.0 across cycles, whereas the milled particles irradiated for 1000 hours reached 2.7. The authors interpret this as evidence that particles detach from a surface once a critical level of degradation is reached, so prolonged irradiation of milled particles overshoots the chemistry of genuinely released material. For this particular plastic, 500 hours of ultraviolet exposure brought artificial particles closest to the real ones. Morphologically, the real particles were fragments with aspect ratios between 1.0 and 3.5, broadly comparable to the milled particles, but their surfaces were pitted with micro-cavities, likely formed when water spray dissolved and washed away degradation products, giving them a large surface area and presumably high adsorption capacity. The milled particles, never exposed to water, were noticeably smoother.

The study also demonstrates why simple mass-loss measurements cannot substitute for direct particle collection. After eight cycles, the composite sheets had lost 66.4 grams per square metre and the polypropylene 36.8 grams, yet microplastic release accounted for only a tiny fraction of these figures. The remainder is explained by volatile compounds such as carbon dioxide, water, ketones, acids, esters, and aldehydes formed during photodegradation, along with leaching of low-molecular substances. In earlier work by the same group, a quarter of the mass of an irradiated polypropylene powder disappeared purely through volatile formation. Gravimetric approaches therefore work for abrasion-dominated sources like tyres but are unsuitable for weathering.

The authors are careful to spell out the limitations. The tests are accelerated and simplified, including only ultraviolet radiation and water spray, the two stressors considered most severe, and omit factors such as wind, temperature cycling, air pollution, and geographic variation, so the results cannot be translated directly into annual release rates under natural conditions. The composite contained no functional additives beyond those already in the recycled polymer, meaning real commercial products with stabilisers could behave very differently. The method was validated on only two related materials, and the researchers state that further testing with other plastic-based materials is needed to establish its broader applicability. They also note that the smallest particles, below 20 micrometres and potentially down to the nanoscale, remain analytically challenging, an area where methodologies are still being developed and validated worldwide.

Even so, the implications are considerable. A reproducible, standardised way to rank materials by microplastic release under identical conditions could inform future legislative limits on emissions from building products, decking, furniture, toys, and single-use items, much as existing weathering standards govern durability claims today. Because the method collects the particles it quantifies, it also supplies authentic weathered microplastics, formed under realistic combinations of light and moisture rather than by milling or chemical synthesis, for use in environmental and toxicological research. Given mounting evidence linking microplastic exposure to cardiovascular, intestinal, pulmonary, and inflammatory conditions, and given that weathered particles differ chemically from pristine ones in ways that affect their behaviour, the availability of genuinely representative test material may prove as valuable as the measurement itself. The Latvian team’s sieve-and-recirculation retrofit turns a routine durability instrument into a window on one of the least quantified pathways of plastic pollution.