In controlled laboratory tests, scientists at the VSB Technical University of Ostrava showed that swarms of magnetic microrobots can extract up to 94% of polystyrene microplastics from water and 81% from moist soil within one hour. Published in NPG Asia Materials on August 1, 2026, the results point to a possible approach to microplastic pollution, though the technology remains far from real-world deployment.
The system uses tiny structures made from layered MXene Ti₃C₂Tₓ microparticles and nickel nanoparticles, each about 500 nanometres in size. Rotating magnetic fields propel and steer the microrobots, so they need no onboard motors or batteries. They also require no electronics: external magnetic fields synchronize their movement and guide them through water or soil samples. In tests, the robots reached speeds of 22.1 micrometres per second under a 5-millitesla, 3-hertz field.
How the Robots Work
The MXene surfaces act as powerful adsorbents, binding microplastic particles on contact. The robots' mobility lets them seek out and capture dispersed contaminants instead of passively waiting for pollutants to pass through a filter. Once they collect the microplastics, the robots can be retrieved magnetically, concentrating the captured material for removal. This separates the plastic, rather than destroying it: the collected fraction still needs to be handled or processed.
In soil, the microrobots moved through water-filled microenvironments, helping dislodge and recover trapped plastic fragments. The approach builds on earlier research into magnetic materials for water purification and applies the concept to soil remediation, where traditional filtration is ineffective. The broader Advanced Nanorobots and Multiscale Robotics Laboratory at VSB-Technical University of Ostrava is led by Professor Martin Pumera. Its work spans the design and synthesis of micro- and nanorobotic systems, including their propulsion and functionalization, with applications beyond pollution control.
Lab Results and Limitations
Researchers tested the robots on 1-micrometre polystyrene (PS) particles and polyethylene terephthalate (PET) fragments up to 150 micrometres. After 60 minutes, active microrobots removed 94.0% of PS and 89.2% of PET from water, compared with 81.1% and 74.4% for static material. In moist soil, the removal rates were 80.6% for PS and 72.2% for PET, versus 70.7% and 52.5% without movement. The largest improvement, nearly 20 percentage points, came with PET in soil.
These figures come from controlled laboratory conditions using prepared suspensions and Petri dish soil samples. Researchers have not yet tested the system in natural water bodies, open soils or environments with complex contaminant mixtures. Rivers and real soils introduce variables: organic matter and minerals can affect performance, as can roots and fluctuating moisture. The energy and material requirements for large-scale use also remain unknown. Removal percentages from a prepared sample do not directly show how much plastic could be taken out of a river or field, where particle concentrations, sizes and sampling conditions differ.
Microplastics are generally defined as plastic particles smaller than 5 millimetres. They can be manufactured at small sizes or form as larger plastic products break down. Other pathways into the environment include tyre wear and synthetic textile fibres, along with industrial pellet losses, wastewater and stormwater. The OECD has estimated that about 22 million tonnes of plastic leaked into the environment in 2019, including roughly 2.6 million tonnes of microplastics. Those global mass estimates measure a different scale from the laboratory removal rates reported in the study.
Safety and Environmental Concerns
One unresolved issue is nickel, which gives the robots their magnetic response but could release metal ions if the robots are not fully recovered. The authors warn that incomplete retrieval may cause secondary contamination. Before the technology is used in open environments, developers will need safer magnetic materials and thorough toxicological studies. Strong recovery protocols will also be impressive. For now, the most plausible applications are in closed systems, including industrial effluents or treatment plants, where containment is easier to guarantee.
Policy Context and Next Steps
This research comes as the European Union steps up efforts to curb microplastic emissions. The Zero Pollution Action Plan targets a 30% reduction by 2030, while new REACH restrictions on intentionally added synthetic particles took effect in October 2023. Regulation 2025/2365, effective from December 2025, imposes stricter controls on plastic pellet losses, especially for facilities handling over five tonnes annually.
The scale of pollution puts cleanup research alongside efforts to prevent plastic waste. The United Nations Environment Programme estimates that around 11 million metric tonnes of plastic enter the ocean each year and projects annual inputs of 23 million to 37 million tonnes by 2040 without stronger action. Its 2023 Turning off the Tap assessment concluded that policy and market changes using existing technologies could cut plastic pollution by about 80% by 2040. The assessment also points to the continuing need to reduce plastic use and improve reuse, while strengthening recycling and waste management.
The Ostrava team's technology is not ready for environmental deployment. Its next major challenge is to demonstrate effectiveness, safety and economic viability outside the lab. The university's Technology Transfer Centre provides an institutional route toward potential partnerships or applications, but practical tests must establish whether the robots can be reliably recovered and the nickel-bearing material contained. Their performance in complex real-world conditions also remains to be tested.
The full study is available in NPG Asia Materials.
