Researchers validate low-cost filter device for sampling microplastics in wells

Groundwater may be the last great blind spot in microplastic research, and a team of hydrogeologists in the United States believes the reason is largely economic. Scientists at the University of Memphis and Saint Louis University have now built and validated a modular, in-line filtration apparatus that costs roughly 120 US dollars to assemble from hardware-store parts, a fraction of the price of commercial systems that can run into the thousands. In laboratory spike tests, the device recovered an average of 96 percent of added microplastic particles when paired with a peristaltic pump, and it successfully filtered approximately 500 liters of groundwater per well across three field deployments in the Memphis, Tennessee, metropolitan area without clogging or mechanical failure. The work, published in the open-access journal Microplastics and Nanoplastics, addresses what the authors describe as a critical gap in subsurface monitoring.

The scale of the blind spot is striking. Microplastics, generally defined as plastic particles smaller than 5 millimeters, have been documented in oceans, rivers, soils, air, wildlife, and even the human body since they were first flagged as an emerging contaminant in 2004. Yet a systematic review covering 2017 to 2023 identified only 201 published studies on microplastics in groundwater worldwide, and just 24 of those reported actual field concentrations. Most of the existing work is concentrated in Asia and Europe and targets shallow alluvial or karst aquifers. This scarcity persists despite groundwater supplying drinking water to nearly half the global population and more than 40 percent of irrigation water worldwide.

Part of the problem is methodological. Because microplastic concentrations in low-turbidity groundwater are expected to be low, researchers now recommend extracting large volumes, typically more than 100 liters per sample, to obtain statistically meaningful counts. Small-volume approaches using bailers, amber glass bottles, buckets, or flasks, which have dominated the literature, generally capture less than 10 liters and may substantially underestimate particle abundances. Pumping hundreds of liters, however, demands in-line filtration systems, and those have historically been expensive. Commercial filter holders alone can cost several hundred to several thousand dollars. Individual filter elements from industrial suppliers have been quoted at 330 to 450 dollars each, and a portable pump-filtration system marketed for coastal microplastic sampling is priced near 8,800 dollars, with additional filter plates costing about 400 dollars. For laboratories in resource-limited regions, those prices are a genuine barrier to entry, and they help explain why groundwater remains among the least studied compartments for plastic pollution.

The new apparatus tackles the cost problem through commodity components. Each filtration module consists of a stainless-steel filter disk with a 7-centimeter inner diameter, sealed between two silicone gaskets and compressed between iron floor flanges bolted together in a crisscross pattern to at least 6.8 newton-meters of torque. A coarse stainless-steel backing mesh with roughly 4.5-millimeter openings reinforces each disk against deformation under high flow. The modules thread together in series using standard 2-inch National Pipe Thread fittings, arranged from coarse to fine mesh, so that water passes sequentially through 270-micrometer, 110-micrometer, and approximately 30-micrometer screens. This cascade enables size fractionation, a feature that helps researchers characterize the particle-size distribution of contamination and reduces clogging by distributing retained solids across stages. Notably, silicone gaskets are the only polymeric components in the entire assembly, and they serve purely as sealing elements, never contacting the sample stream. Inlet connections offer two options: a recommended copper compression fitting for metal tubing, or a hose barb for plastic tubing when a particular pump configuration requires it.

Contamination control, a persistent concern in microplastic science, shaped nearly every aspect of the protocol. All procedures took place in a laboratory dedicated exclusively to microplastic work, with restricted access and personnel wearing bright green, 100 percent cotton lab coats to make any shed fibers visually conspicuous. Reagents, including deionized water and detergent, were vacuum-filtered through 1.6-micrometer glass fiber filters before use. Glassware and apparatus components were rinsed sequentially with filtered detergent, filtered deionized water, and compressed air passed through a 30-micrometer screen, then stored wrapped in aluminum foil with openings facing downward. Work surfaces and components were routinely inspected under 365- and 395-nanometer ultraviolet illumination to catch fluorescent particles invisible under ordinary light, and procedural blanks were run throughout to track airborne deposition.

To validate recovery, the team ran triplicate spike tests, each introducing a standardized mixture of 50 particles into 5 liters of filtered deionized water: 24 fibers of varying sizes, 20 fragments roughly 230 to 250 micrometers wide and about 10 micrometers thick, three polyurethane foam pieces, and three low-density polyethylene film pieces. The particles were prepared in-house from household and hardware materials, a deliberate choice given that commercial reference microplastics can themselves cost hundreds to thousands of dollars, and the heterogeneity was intended to mimic real environmental diversity. With a peristaltic pump operating at roughly 2.5 liters per minute, average recovery reached 96 percent. Fibers, foams, and films were recovered at 98 to 100 percent, while fragments, the smallest and flattest particles, lagged at 75 percent, most retained on the two finest meshes.

Results with a modified stainless-steel submersible centrifugal pump were more variable, ranging from 22 percent for foams to 100 percent for small fibers. Critically, however, when the researchers rinsed and examined the pump housing itself, they found substantial particle retention, primarily within the narrow annular space between the inner and outer housings. Correcting for those pump-retained particles raised the apparatus’s own recovery to 76 percent, and residuals inside the filtration modules were limited to 7 percent and consisted only of fragments. The comparison underscores that recovery outcomes in microplastic sampling depend as much on pump geometry and internal flow pathways as on the filter itself. The centrifugal pump pushes about 6 liters per minute through multiple intake openings into a constricted one-millimeter channel, creating far more opportunity for particle capture than the smooth, unobstructed bore of peristaltic tubing. Interestingly, trials filtering 10 liters instead of 5 showed that larger volumes helped flush retained foams and films free, improving recovery.

Field performance was tested in three monitoring wells, ranging from 11.7 to 39 meters in total depth, screened in unconsolidated sediment aquifers of the Mississippi embayment regional aquifer system. Using the modified submersible pump, the team filtered approximately 500 liters per well at flow rates of 2.6 to 3.4 liters per minute, consistent with published high-volume protocols. The apparatus remained structurally stable throughout extended pumping, with no deformation of flanges, gaskets, or meshes. Minor leakage occurred at higher pump speeds, at less than 1 percent to roughly 1.3 percent of total flow, and was eliminated by improving gasket alignment and applying uniform torque. In the deeper confined well, intermittent flow reductions traced to pressure buildup and possible gas entrapment from groundwater degassing were resolved by briefly reducing flow and venting the outlet. Recovered filters showed textbook size fractionation, with coarse sediments concentrated on upper stages and progressively finer material on lower meshes, and preliminary microscopy under visible and ultraviolet light revealed features visually consistent with anthropogenic fibers and fragments, though formal particle identification and polymer characterization were beyond the scope of the study.

The system’s flexibility extends beyond aquifers. In a preliminary surface-water trial, the apparatus filtered about 50 liters of river water with turbidity exceeding 20 nephelometric turbidity units without clogging or leakage, again producing clear fractionation across the mesh stages. The authors also document practical refinements for future users: sonication of metal components in filtered deionized water between deployments effectively dislodges trapped particles, routine drying or replacement guards against surface rust on non-stainless parts, and operators should avoid over-applying anti-seize lubricant, which can trap particles at threaded joints. For deeper wells or degassing groundwaters, lower flow rates and a small bleed valve are recommended to prevent pressure buildup.

The broader significance is one of access. By demonstrating that a device assembled from pipe fittings, flanges, bolts, and stainless-steel mesh can match the retention performance of systems costing tens of times more, the researchers have removed one of the main excuses for the field’s neglect of subsurface environments. An affordable, scalable, contamination-minimizing platform that yields size-fractionated samples from 500-liter volumes could finally allow groundwater, the water source for billions of people, to be monitored for plastic pollution with the same rigor long applied to oceans and rivers. Whether particles below roughly 30 micrometers, and the polymers they are made of, can be characterized with this hardware will be the next question, but the entry barrier to asking it has just dropped dramatically.