The impact of soil properties on the chemistry and ecotoxicity of copper oxide nanoparticles

Abstract

Copper is applied since decades in agriculture as fungicide and fertilizer, but can accumulate in top soils and cause negative effects to soil biota. Copper-based nanoparticles as ingredient could lead to more efficient applications and reduce the amount of applied copper, but at the same time raise novel risks towards the environment. The behaviour and the resulting risk of nanoparticles is strongly depending on their interaction with the environment. In the present thesis, we studied the chemical behaviour and/or the resulting ecotoxicity of a model substance, uncoated copper oxide nanoparticles (CuO-NP, <50 nm), in different artificial test media (soil porewater, fresh- and seawater), field soils and artificial OECD soils. The impact of the media on size, surface characteristics and dissolution of CuO-NP under the influence of humic acid was studied. The size and dissolution of CuO-NP in the pore water of four field soils was also investigated, one sandy soil and three loamy soils of varying pH and organic content were used. The impact of CuO-NP as well as CuCl2 on survival, reproduction and body weight of a model soil organism, the springtail Folsomia candida, was investigated in these field soils, but also in artificial OECD soils containing either kaolin or montmorillonite as representatives for the most common clay minerals in soils. For the OECD soils, also catalase activity, metallothionein level and moulting frequency of springtails were investigated. For all soils, mainly field-realistic Cu concentrations were applied. In media of a high divalent cation fraction but low ionic strength, particle size and dissolution were remarkably higher, but the addition of humic acid roughly aligned these parameters for all media. The behaviour of CuO-NP in artificial media was clearly impacted by the interaction between humic acid, the ratio of dissolved ions of different valency and their competition for interaction sites on the NP surface. The particle size in the pore water of field soils was independent of soil properties, but increased with increasing test concentrations. However, toxic effects occurred only in loamy soils and mainly at the lowest NP concentration of 5 mg Cu/kg, with reproduction and body weight being reduced most by 61 or 28%, respectively, in a loamy-acidic soil. The effects were ascribed to the small size of CuO-NP at low concentrations providing more reactive sites and the higher content of clay minerals in loamy soils chemically interacting with CuO-NP, possibly supported by low soil pH and curbed by high SOM. In artificial OECD soils containing montmorillonite and CuO-NP, survival, reproduction and body weight of springtails were reduced by up to 52, 33 and 48%, respectively, at soil concentrations of 1-3 mg Cu/kg, while at higher concentrations no negative effects occurred. Further, the catalase activity was depleted by up to 62% at 10 mg Cu/kg or lower. The metallothionein levels and the Cu content of exuviae shed by springtails was higher in these soils than for kaolin soils and/or CuCl2 treated soils. We assume that CuO-NP-montmorillonite associations can provide and shuttle electrons for the formation of reactive oxygen species on the NP surface through Fe and Cu in a non-linear dose-response relationship. Furthermore, they may attach stronger to springtail gut cells and impact excretion mechanisms. Overall, the thesis underlines the strong impact of environmental substances such as ion solutions, organic matter and clay minerals on the behaviour and toxicity of CuO-NP. In the soil environment, a negative impact on springtails can occur at field-realistic Cu concentrations in loamy soils, which are of high relevance for agriculture, containing Fe-rich reactive clays. Therefore, we suggest a focus on these soils and a denser testing at lower, field-realistic Cu concentrations for the risk assessment of agrochemicals containing Cu-based nanoparticles.