Evidence for TiO2 nanoparticle transfer in a hard-rock aquifer.

Water flow and TiO2 nanoparticle (NP) transfer in a fractured hard-rock aquifer were studied in a tracer test experiment at a pilot site in Brittany, France. Results from the Br tracer test show that the schist aquifer can be represented by a two-layer medium comprising i) fractures with low longitudinal dispersivity in which water and solute transport is relatively fast, and ii) a network of small fissures with high longitudinal dispersivity in which transport is slower. Although a large amount of NPs was retained within the aquifer, a significant TiO2 concentration was measured in a well 15m downstream of the NP injection well, clearly confirming the potential for TiO2 NPs to be transported in groundwater. The Ti concentration profile in the downstream well was modelled using a two-layer medium approach. The delay used for the TiO2 NPs simulation compared to the Br concentration profiles in the downstream well indicate that the aggregated TiO2 NPs interacted with the rock. Unlike Br, NPs do not penetrate the entire pore network during transfer because of electrostatic interactions between NP aggregates and the rock and also to the aggregate size and the hydrodynamic conditions, especially where the porosity is very low; NPs with a weak negative charge can be attached onto the rock surface, and more particularly onto the positively charged iron oxyhydroxides coating the main pathways due to natural denitrification. Nevertheless, TiO2 NPs are mobile and transfer within fracture and fissure media. Any modification of the aquifer's chemical conditions is likely to impact the groundwater pH and, the nitrate content and the denitrification process, and thus affect NP aggregation and attachment.