The starting material was extraordinary by any environmental standard. Iron, lead, and zinc were present at 25,695, 13,800, and 29,500 milligrams per kilogram respectively, while cadmium reached 1,140 milligrams per kilogram, mercury 54.6, copper 759, antimony 63.9, arsenic 159, and selenium 64.1. Compared with Italian regulatory thresholds for industrial land, cadmium exceeded its limit by a factor of 76, mercury by 11, lead by 14, zinc by 20, arsenic by 3, and selenium by 4. X-ray diffraction revealed why the soil was so contaminated and so difficult to treat: the metals were not merely adsorbed onto particle surfaces but existed as discrete mineral phases, including zincite (ZnO), cerussite (PbCO3), sphalerite (ZnS), hemimorphite, and iron-bearing pyroxenes such as ferrosilite. This mineralogical fingerprint reflects the site’s history of zinc and lead production, since elements like cadmium, mercury, arsenic, and selenium are classic impurities in those ores.

The first treatment step exploited a well-known principle of contaminated-site science: heavy metals concentrate in fine particles. Silt and clay fractions offer enormous specific surface area, giving metals vastly more adsorption sites than coarse sand or gravel. When the researchers sieved the soil at a 0.075-millimeter threshold, the finest fraction, just 21 percent of the sample mass, carried contaminant concentrations two to seven times higher than the coarser material. That meant 79 percent of the soil could be recovered as a cleaner coarse fraction, with concentrations of antimony, lead, mercury, and zinc roughly 1.5 times lower than the original soil, and arsenic, cadmium, copper, chromium, and tin about two times lower. In the finest particles below 10 micrometers, total metal concentrations reached a staggering 236 grams per kilogram, three to seven times the bulk soil value.

Next came the most visually striking step: high-intensity magnetic separation. Using a laboratory-scale rare-earth roll separator equipped with a neodymium-boron-iron magnet operating at 2 tesla, the team pulled a metal-rich magnetic fraction, 15 percent of the treated mass, out of the coarse soil. Removal efficiencies for the combined physical steps ranged from 73 to 93 percent for lead, arsenic, and antimony, with zinc at 70 percent. The physics here is subtle, because lead and zinc minerals are normally diamagnetic and should be invisible to a magnetic field. The trick is an indirect mechanism: these diamagnetic metal-bearing minerals are intimately associated with paramagnetic iron oxides and ferromagnetic impurities that act as magnetic carriers. Optical microscopy confirmed this, revealing dark, iron-rich inclusions embedded within the light brown diamagnetic mineral matrix of captured particles. Mercury and selenium, by contrast, resisted magnetic capture, with removals of only 14 and 16 percent, suggesting they were bound in forms unassociated with magnetic phases.

The final chemical stage targeted what physical separation could not reach. The nonmagnetic fraction was washed with EDTA solutions ranging from 15 to 75 millimolar, and a control test without EDTA confirmed that any metal release was driven by the chelating agent rather than by water contact alone. EDTA works by forming highly stable, water-soluble complexes with metal cations, prying contaminants loose from exchangeable, carbonate-bound, and even some oxidizable pools. Cadmium responded most dramatically, dropping from 232 to 48.6 milligrams per kilogram, roughly a fivefold reduction, while mercury, selenium, arsenic, copper, lead, and zinc fell by factors of two to two and a half. Notably, EDTA’s selectivity worked in the soil’s favor: iron was extracted only weakly, which preserves the soil’s physical structure, a major advantage over aggressive mineral acids that dissolve the soil matrix and leave it needing extensive neutralization.

The concentration data also revealed a practical ceiling. Metal removal improved as EDTA concentration rose, but only up to about 45 millimolar for cadmium, lead, copper, and zinc, after which concentrations plateaued. The researchers interpret this as depletion of the readily extractable pools: once the water-soluble, exchangeable, and carbonate-bound fractions are gone, additional chelator provides negligible benefit because the remaining metals are locked in less accessible forms. Arsenic was particularly resistant, requiring 60 millimolar EDTA for even a twofold reduction, likely because it exists as oxyanions for which EDTA has low affinity and can only be mobilized indirectly by dissolving its iron oxide carriers. Even at 75 millimolar, cadmium, mercury, and zinc remained 3, 2.5, and 1.8 times above Italian industrial thresholds, a reminder that extremely contaminated starting material limits what any single-reagent treatment can achieve.

What sets this study apart from most remediation work is that the researchers did not stop at removal percentages. Instead, they ran a full human health risk assessment using United States Environmental Protection Agency methodologies, calculating hazard quotients and hazard indexes for ingestion, dermal contact, and inhalation across three exposure scenarios: outdoor workers, residential adults, and residential children. In the untreated soil, cadmium and thallium were the dominant noncarcinogenic threats, with cadmium’s hazard index reaching 169.4 for children, driven by their higher ingestion rates of 200 milligrams per day versus 100 for adults and their far lower body weight of 15 kilograms versus 70. Arsenic, a known human carcinogen, pushed carcinogenic risk above the EPA threshold of 10 to the minus 4 for residential adults and children through ingestion. Perhaps most alarming was the lead assessment: the estimated fetal blood lead concentration for pregnant women exposed to the soil was 51 micrograms per deciliter, five times the EPA’s level of concern of 10 micrograms per deciliter, a level associated with irreversible neurodevelopmental harm.

Each treatment step translated measurably into risk reduction, and the risk framework exposed which steps mattered most for whom. Mechanical screening alone halved cadmium concentrations but left hazard indexes above unity in every scenario, though a 23 percent thallium reduction was enough to protect outdoor workers. Magnetic separation proved transformative: it cut most contaminant concentrations by 73 to 93 percent, rendered arsenic and antimony risks negligible for children, brought childhood arsenic carcinogenic risk into the tolerable range, and pulled the fetal blood lead estimate down to 11 micrograms per deciliter, nearly hitting the target. Chemical extraction then finished the job for adults. With 30 millimolar EDTA, cadmium risks fell below unity for outdoor workers; with 45 millimolar, all hazard indexes dropped below one for residential adults, and just 15 millimolar sufficed to meet the fetal blood lead goal of 10 micrograms per deciliter and to eliminate mercury risk for children.

The residual problem is stark and specific. Even after the full treatment train at 45 millimolar EDTA, cadmium’s hazard index for children remained at 7.2 and thallium’s at 2.7, both well above the safety threshold of one. Children’s hand-to-mouth behavior, higher gastrointestinal absorption, and small body mass amplify their dose to a degree that the achieved concentration reductions cannot fully offset. The authors are candid about the implication: the remediated soil does not meet the stringent requirements for residential land use without further measures, whether additional treatment steps, combined reagent schemes, or complementary management controls. They also note that extending the treatment with more extraction cycles or alternative reagents would sharply increase operational complexity and cost, potentially undermining the feasibility of scaling up to full-site remediation. Future work, they suggest, could incorporate bioaccessibility assays to refine estimates of the fraction of metal actually absorbed by the human body.

Still, the study’s overall message is one of pragmatic optimism. By combining established, mining-derived physical separation techniques with a well-understood chelating wash, all evaluated within a quantitative risk framework, the researchers demonstrated a treatment pathway that renders severely contaminated smelter soil safe for industrial and commercial redevelopment, protecting workers and adult residents. The approach is cost-effective, uses readily available equipment, preserves soil physicochemical properties, and even concentrates the worst contamination into small disposal volumes: the fine fraction and the magnetic concentrate together represent only a portion of the original mass. The remaining challenge, protecting the most vulnerable population, children, from persistent cadmium and thallium exposure, defines the next frontier for soil remediation science, and this study provides both the measurement tools and the honest risk accounting needed to pursue it.

Subject of Research: Integrated physical-chemical remediation and human health risk assessment of multi-metal contaminated industrial soil

Article Title: Integrated physical-chemical remediation and human health risk assessment of multi-metal contaminated industrial soils

Article References: Marsich, L., Ferluga, A., Cozzarini, L., & Bevilacqua, P. (2026). Integrated physical-chemical remediation and human health risk assessment of multi-metal contaminated industrial soils. Environmental Science and Pollution Research, 33(28), 14241-14265. https://doi.org/10.1007/s11356-026-38177-x

Image Credits: AI Generated

DOI: 10.1007/s11356-026-38177-x

Keywords: soil remediation, heavy metals, EDTA extraction, magnetic separation, mechanical screening, health risk assessment, cadmium, thallium, industrial contamination, smelting site, lead exposure, Environmental Science and Pollution Research