Novel DGT binding layer for organomineral colloids to identify their role in trace metal mobility in soil

Natural soil colloids can play a substantial role in the mobility and bioavailability of trace metals. However, the understanding of their role as carrier is hampered by sampling artefacts of these colloids in traditional soil solution sampling methods, including changes in the in situ soil environment by extraction or centrifugation. To overcome this, a method based on the Diffusive Gradients in Thin films (DGT) technique is being developed to enable in situ sampling of colloids in undisturbed soil. Spatially resolved analysis of a DGT binding layer with laser ablation-ICP-MS will allow high-resolution mapping of free and colloid-associated metal distributions in soil. Different new types of DGT binding gels were synthesised and tested for their ability to accumulate organomineral colloids of iron (Fe) oxides associated with natural organic matter (NOM). These Fe-NOM colloids are expected to adsorb onto metal oxide binding gels used for anionic species in DGT, via interactions with negatively charged carboxylic and phenolic hydroxyl reactive groups of the NOM. Colloids experience slow diffusion due to their size and might be outcompeted for sorption by oxyanions, e.g. phosphate (PO₄). Therefore, a high colloid binding capacity is required to ensure sufficient colloid detection on the binding gel. Sorption tests showed that Fe-NOM colloids can be accumulated by the in situ precipitated zirconium oxide (ZrO₂) binding gel based on the ZrOCl₂ precursor that is currently being used in PO₄ DGTs. With respect to this binding layer, the increase of the ZrO₂ concentration was found to have the most remarkable effect on the general binding capacity, as the total PO₄ sorption capacity increased linearly with increasing Zr content in the gel. A novel approach, the in situ precipitation of metal oxides from Zr, titanium (Ti) and niobium (Nb) chloride and n-butoxide precursors instead of ZrOCl₂, did not significantly enhance either the general capacity or the affinity for Fe-NOM colloids of binding gels. In addition, the Fe-NOM colloid sorption was not significantly enhanced by adaptation of the agarose derivative-crosslinked polyacrylamide (APA) hydrogel by mixing the non-ionic adsorbent polyvinylpyrrolidone (PVP) within the matrix. Moreover, hydrogels based on agarose instead of APA did not promote sorption of larger-sized Fe oxide colloids (50 nm) to a significant extent, despite the larger pore sizes in agarose compared to APA hydrogels. Finally, the in situ precipitated 0.1 M ZrO₂ binding gel showed a linear uptake of small Fe-NOM colloids in time and concentration. This gel is, therefore, a promising DGT binding layer for high-resolution imaging of NOM-based Fe oxide colloids with associated trace metals in soil.