Abstract

Despite the numerous studies about the sorption of dissolved organic matter (DOM) onto nanoparticles,the extrapolation of laboratory results to environmental conditions is currently impossible. Indeed, thecomplex dynamics of DOM under variable environmental conditions are not completely reproducibleunder control conditions. In this study, we propose a different approach by exploring a method forexposing nanoparticles to realistic environmental conditions in natural river water by using dialysismembranes as passive reactors. Inside this reactor, the complexity and the temporal variability of a largenumber of environmental parameters (DOM structure and composition, temperature, inorganic ions, pH,etc.) are reproduced, while colloidal and particulate interferences remain separated. To verify thisassumption, we determined the concentration of the water components and nanoparticles (n-TiO2,20–50nm) inside and outside the reactor before and after exposure to river water. In river water, more than 90%of the n-TiO2nanoparticles remained inside the reactor while DOM retained its molecular composition/characteristics after passing through the membrane (DOC, fluorescence EEM, and FT-ICR MS). For mostelements and anions, the concentrations inside and outside the reactor did not differ, indicating a goodpermeability for inorganic constituents (IC, ICP-OES); however, the concentrations of Al, Fe, Mn, and nitratewere lower. Membrane fouling, in terms of pore size distribution, was investigated using NMR relaxometryand AFM in fluid mode; no significant reduction in pore size was observed under the applied conditionsduring seven days of exposure. Finally, ATR-FTIR and CHNS analysis of n-TiO2before and after exposureto the river water revealed that sorption of DOM occurred under field conditions. Therefore, we coulddemonstrate the validity and the potential of this method.

Download