A growing environmental concern, nanoplastics have been found in water around the world—from the deep oceans to Arctic ice.
New University of Wisconsin-Madison research, published in the August 2025 issue of the journal Environmental Science and Technology, could make these tiny plastic particles easier to detect and monitor. That’s important, in part, because nanoplastics can easily make their way into our bodies and adversely impact our health.
Haoran Wei and Mohan Qin, both assistant professors of civil and environmental engineering at UW-Madison, developed an algorithm that reduces noise from the membranes used in Raman spectroscopic analysis of nanoplastics. In Raman spectroscopy, researchers shine a laser onto a material and measure shifts in light frequency as photons interact with molecular bonds. Because each molecule scatters light differently, researchers can identify materials by tracking energy gained or lost in those interactions.
However, since nanoplastics measure between 1 and 1,000 nanometers—smaller than a human hair—they’re so tiny that they push the limits of Raman detection, Wei says.
“The laser spot size we use is relatively large—about several micrometers in size,” Wei says. “If the plastic particle is only a few hundred nanometers across, it generates a small signal. At that point you can get a lot of interference from the membrane itself, because it produces slight signal returns.“
Those signal returns can show up as false positives when researchers are scanning a water sample for nanoplastics. To address this, Wei’s and Qin’s teams used experimental data sets to determine signal thresholds at different points along the light spectrum. Wei says the thresholds are customizable.
The teams tested the algorithm on water samples from Lake Mendota, in Madison, Wisconsin, and from Lake Michigan. The algorithm was 93.5% accurate in identifying nanoplastics and more than 90% accurate in rejecting non-nanoplastic interference.
With a cleaner picture, the researchers can count the pixels in a scanned image to measure nanoplastic concentrations in a sample. “It works very well for nanoplastic identification,” Wei says. “We cleaned up almost all of the interference across the nanoplastics signals. We achieved almost perfect quantification calibration curves because there were no false positives across the membrane surface.”
While the algorithm currently only works with single-band Raman spectroscopy, which can be less selective if two polymers share common features, Wei plans to refine the algorithm to work with multi-band Raman spectroscopy, also.
He hopes the method will be a step toward improving how we monitor nanoplastic levels in our water. “We don’t have a standard method for measuring nanoplastics in water,” Wei says. “But there’s been a lot of focus on this emerging contaminant in the last few years, because it’s potentially a bio-accumulator and we don’t know how toxic it may be. This project was for Great Lakes water, but it could potentially be extended to drinking water, wastewater, or groundwater.”
Featured image caption: Ziyan Wu, a PhD student in Assistant Professor of Civil and Environmental Engineering Mohan Qin’s lab, shows a boat’s sequential filtration system while collecting samples from Lake Michigan. The sequential filtration system helps capture contaminants like microplastics and “forever chemicals” in water samples.