Monitoring uranium (U) in soft freshwater environments is critical during and after uranium mine decommissioning, particularly where ecosystems are sensitive to small changes in water chemistry. This is especially relevant for the surface waters adjacent to the Ranger uranium mine in Kakadu National Park, including Magela Creek, which flows close to the mine and has the potential to be impacted during rehabilitation. These waters are typically characterised by extremely low ionic strength. Diffusive Gradients in Thin Films (DGT) samplers are attractive for this purpose because they provide time-weighted, in situ measurements of the labile and potentially bioavailable fraction of U. However, their performance in extremely low ionic strength waters remains insufficiently understood.
This study investigated the suitability of DGTs equipped with Chelex-100 or Metsorb binding layers for measuring U in soft tropical freshwater systems, with emphasis on the effects of U speciation, electrostatic interactions, and material variability on measurement accuracy. In laboratory experiments, U accumulated linearly with deployment time, however, DGT-derived concentrations consistently overestimated dissolved U at low ionic strength, irrespective of binding resin. This overestimation was attributed to Donnan potentials, whereby electrostatic partitioning enriches cationic uranyl species within negatively charged polyacrylamide-based diffusive gels relative to the bulk solution.
The magnitude of this effect varied between diffusion layer chemistries and between production batches, reflecting the absence of charge control in the manufacturing process. Agarose cross-linked polyacrylamide gels showed stronger overestimation, whereas agarose-only gels reduced electrostatic interference, but with lower analytical sensitivity. Validation of DGT-derived concentrations was further complicated by adsorption of U to container walls and freshly precipitated metal oxides, resulting in uncertain reference dissolved concentrations. These findings demonstrate that, unless diffusion layer charge is controlled or corrected for, each batch of DGT diffusive gel must be independently characterised, and improved protocols for dissolved U measurements are required to support reliable application of DGTs in soft freshwater systems.