Detecting microplastics in drinking water requires specialized laboratory equipment, not a simple at-home test. Particles range from 5 millimeters down to 100 nanometers, and most are invisible to the naked eye. The methods used to find and identify them vary in cost, speed, and the smallest particle size they can catch, but they all share a basic workflow: collect a water sample, filter it, remove organic debris, and then analyze what’s left using some form of chemical identification.
What Counts as a Microplastic
Microplastics are generally defined as solid plastic particles smaller than 5 millimeters. The European Chemicals Agency proposed a regulatory definition that includes any solid polymer-containing particle with all dimensions between 1 nanometer and 5 millimeters, plus fibers up to 15 millimeters long. Below 100 nanometers, particles are classified as nanoplastics. There’s also an in-between category sometimes called “submicroplastics” for particles smaller than 1 micrometer but larger than 100 nanometers.
This size range matters because no single detection method covers all of it. The technique you’d use to find a 2-millimeter fiber is completely different from the one needed to spot a 100-nanometer fragment, and some of the most concerning particles are the smallest ones.
Preparing a Water Sample
Before any analysis, the water sample needs to be filtered and cleaned. Researchers typically pass water through filters with pore sizes around 20 micrometers, which balances catching small particles against practical filtration time. For detecting even smaller particles, filters with pores as fine as 0.45 micrometers are used, though filtering large volumes through these takes considerably longer.
Once filtered, the material on the filter often contains organic matter like algae, bacteria, and other biological debris that can interfere with detection. This organic material is digested away using hydrogen peroxide or enzyme treatments, leaving behind only the synthetic particles. Contamination control is critical throughout: lab air itself contains airborne microplastic fibers, so researchers filter their rinse water, use glass or metal equipment instead of plastic, and run blank samples alongside every test to account for any particles introduced during the process.
Infrared Spectroscopy for Particles Above 10 Micrometers
Micro-FTIR (Fourier Transform Infrared) spectroscopy is one of the two workhorse methods for microplastic detection. It works by shining infrared light onto a particle and measuring which wavelengths the material absorbs. Every type of plastic has a unique absorption pattern, like a chemical fingerprint, that identifies whether the particle is polyethylene, polypropylene, polystyrene, or something else entirely.
The technique is generally preferred for particles larger than 10 micrometers. It can map relatively large filter surfaces quickly, scanning thousands of particles in a single automated run. It also identifies non-plastic materials like cellulose and protein-based fibers, which helps avoid false positives. The practical lower limit, however, sits around 50 micrometers. Below that size, the spectral quality drops sharply because very thin samples don’t absorb enough infrared light to produce a clear reading. While the theoretical limit is around 6.6 micrometers, most particles below 50 micrometers get missed in practice. Relying on FTIR alone can significantly undercount the total microplastic load in a water sample, since the smallest particles tend to be the most abundant.
Raman Spectroscopy for Smaller Particles
Micro-Raman spectroscopy fills in where FTIR drops off. Instead of measuring absorbed light, it measures how laser light scatters off a particle’s molecular bonds. This gives it finer spatial resolution, capable of identifying particles down to about 1 micrometer, with some setups reaching 0.4 micrometers.
In a head-to-head comparison published in Heliyon, Raman spectroscopy detected significantly more microplastics in drinking water samples than FTIR, primarily because of its ability to catch particles below 10 micrometers. That size fraction turned out to be the most common in the samples tested. The tradeoff is speed. Raman analysis is considerably more time-consuming because each particle needs individual laser targeting. For a thorough analysis across all size ranges, many labs now use both methods: FTIR to efficiently scan the larger particles and Raman to capture the small ones.
Measuring Total Plastic Mass
Spectroscopy-based methods count individual particles and identify their polymer type, but they don’t directly measure how much plastic is present by weight. That’s where pyrolysis combined with gas chromatography and mass spectrometry comes in. This technique heats a sample until the plastics thermally decompose, then identifies the breakdown products. Each polymer type produces characteristic marker compounds: polyethylene produces a compound called 1-dodecene, polypropylene produces 2,4-dimethyl-1-heptene, and polystyrene produces fragments of its styrene building blocks.
By calibrating against known quantities of each plastic, this method provides a mass concentration, telling you not just that polyethylene is present but exactly how many micrograms per liter. The limitation is that it destroys the sample in the process, so you can’t go back and examine individual particles afterward. It also can’t tell you anything about particle count, size, or shape. It’s best used alongside spectroscopic methods to get both pieces of the picture.
Detecting Nanoplastics Below 1 Micrometer
Particles smaller than 1 micrometer fall below the reach of standard Raman and FTIR instruments. Detecting these nanoplastics requires more advanced techniques. One of the most promising is stimulated Raman scattering (SRS) microscopy, which amplifies the faint light signals from tiny particles by a factor of more than 100 million compared to conventional Raman. A 2024 study in the Proceedings of the National Academy of Sciences demonstrated that SRS microscopy can detect individual polystyrene nanoplastics as small as 100 nanometers, with theoretical sensitivity down to 60 nanometers.
This technique allowed researchers to chemically image individual nanoplastic particles in bottled water for the first time, revealing that the number of plastic particles per liter was far higher than previous estimates that could only count down to the micrometer range. SRS microscopy remains a research tool, though, available only in specialized optics laboratories.
Fluorescent Staining as a Screening Tool
For faster, lower-cost screening, researchers sometimes use a fluorescent dye called Nile Red. This dye binds preferentially to non-polar materials like plastics and glows under ultraviolet or blue light, making plastic particles visible against a dark background. Because the dye’s fluorescence changes color depending on the chemical polarity of the surface it attaches to, it can partially distinguish plastics from natural particles like sand or shell fragments.
The brightness of the fluorescence depends on using solvents that cause the plastic to swell slightly, allowing the dye to penetrate the polymer surface. A typical protocol involves staining with Nile Red dissolved in hexane for 30 minutes at 40°C, followed by a second staining step in an ethanol-acetone mixture. After rinsing away excess dye, the filter is examined under a fluorescence microscope.
Nile Red staining is useful for getting a rough count quickly and cheaply, but it has significant accuracy limitations. Recovery rates in spiked experiments have ranged from as low as 3% to 37%, with high variability between replicates. It also can’t identify which type of plastic a particle is made of. Most researchers treat it as a preliminary screening step before confirming results with spectroscopy.
AI-Assisted Identification
One of the biggest bottlenecks in microplastic analysis is the sheer manual labor involved. A trained analyst might spend hours examining a single filter under a microscope, deciding which particles look like plastic and which don’t. Machine learning is increasingly being applied to automate this process. AI systems can analyze microscopic images and spectral data to classify particles by size, shape, color, and polymer type.
Deep learning models trained on datasets of tens of thousands of Raman spectra can identify common polymers like polyethylene, polypropylene, and polystyrene with high accuracy. Image recognition algorithms using convolutional neural networks can sort through microscope images of filter surfaces, distinguishing plastic fragments from mineral grains or biological material. Portable systems combining FTIR with onboard machine learning have also been developed for field use, enabling near-real-time particle identification outside a traditional lab. These tools reduce human error and make it feasible to process the large sample volumes needed for drinking water monitoring.
What’s Available for Home Testing
There are currently no validated consumer-grade test kits for microplastics in drinking water. The detection methods described above require equipment costing tens of thousands to hundreds of thousands of dollars, operated by trained scientists in contamination-controlled environments. Some commercial laboratories will analyze water samples you send in, but costs typically run several hundred dollars per sample, and results depend heavily on which method the lab uses and what size range they target.
If your concern is reducing microplastic exposure rather than measuring it precisely, the more practical route is filtration. Reverse osmosis systems and activated carbon block filters have been shown to remove a significant portion of microplastics from tap water. The particles you’d catch on a home filter, however, would still require lab analysis to confirm they’re actually plastic rather than mineral scale, rust, or sediment.