Carbon filters may inadvertently transport contaminants through water treatment

Activated carbon and biochar have become standard tools in water treatment around the world, trusted to pull pollutants out of drinking water and wastewater. But a new study suggests these materials may be doing something their users did not anticipate: they may be creating a hidden pathway for contaminants to slip through treatment systems and into the water that emerges on the other side.

Researchers at The University of Manchester tested activated carbon alongside three types of biochar—made from hardwood, wheat straw, and corn straw—to see what happens when these materials break down during use. They used lead as a model contaminant and ran it through a series of filtration stages: sand filters, membranes with pores of 0.45 micrometers, and even finer membranes at 0.02 micrometers. What they found was troubling. Extremely small particles of the carbon materials themselves were escaping downstream, and they were carrying the contaminants stuck to their surfaces with them.

The scale of this problem became clear when the team measured where the lead ended up after sand filtration. Between 84.5 and 90.2 percent of the total lead detected was not floating freely in the water as dissolved metal. Instead, it was bound to residual particles of carbon that had broken away from the original material. This matters because it means the contaminant had not actually been removed—it had simply changed form and was still mobile within the treatment system.

Ziheng Wang, the lead author of the study published in the journal Biochar, explained the core issue: adsorption, the process these materials rely on, does not guarantee permanent removal. "If very small carbon particles remain mobile, the contaminants attached to them may also continue moving through the treatment system," Wang said. The particles in question were extraordinarily small—clustering around 100 to 200 nanometers, 0.5 to 1 micrometer, and some apparently as large as 5 micrometers. Even after passing through a 0.45 micrometer membrane filter, lead remained associated with particles in the 0.02 to 0.45 micrometer range, demonstrating that conventional filtration was not catching everything.

This finding exposes a blind spot in how water treatment is currently evaluated. Standard analytical procedures typically filter water samples before measuring dissolved metals, which means the particles carrying contaminants are removed before measurement even begins. The result is an incomplete picture of what is actually in the treated water. A sample might appear clean when measured this way, even though contaminants are still present, just bound to particles that were filtered out during analysis rather than during treatment.

The researchers emphasize that their work was conducted under controlled laboratory conditions. Real water treatment systems operate in far more complex environments, where pH, ionic strength, natural organic matter, and dissolved ions could all affect how stable these particles remain and how readily they transport contaminants. Biochar, which is attracting interest as a potentially cheaper alternative to activated carbon because it can be made from biomass, showed particularly high particle-bound lead levels—as much as 90.2 percent after sand filtration.

The path forward requires rethinking how water treatment performance is measured and potentially how the materials themselves are prepared and deployed. The researchers suggest that future work should evaluate practical interventions: pre-washing adsorbents before use, mechanically stabilizing them, improving coagulation and flocculation processes, and combining sand and membrane filtration in new ways. Most urgently, water treatment systems need to be assessed using both dissolved contaminants and particle-bound contaminants together, not one or the other. Until that happens, treatment that appears successful on paper may be leaving contaminants behind in forms that current testing methods simply do not see.