Environmental Fate of Silver Nanoparticles in Boreal Lake Ecosystems

Abstract

Silver nanoparticles (AgNPs) are currently the most commonly used nanoparticles in consumer products, yet their environmental fate in natural waters is poorly understood. In the present study, we investigated the persistence, transformations and distribution of polyvinylpyrrolidone (PVP) and citrate (CT) coated AgNPs in boreal lake mesocosms dosed either with a 6-week chronic regimen or a one-time pulse treatment at environmentally relevant dosing levels. In the chronic treatments, total Ag (TAg) concentrations reached ∼40% of target concentrations by the end of the experiment, and in the pulsed mesocosms, TAg dissipated slowly, with a half-life of ∼20 days. Sediments and periphyton on the mesocosm walls were an important sink for Ag. We found little effect of AgNP loading and surface coating on the persistence of TAg. There were also no differences between treatments in the degree of agglomeration of AgNPs, as indicated by the accumulation and distribution of Ag in the particulate and colloidal fractions. The low ionic strength and relatively high dissolved organic carbon concentrations in the lake water likely contributed to the relative stability of AgNP in the water column. The low concentrations of dissolved Ag (<1 μg L(-1)) in the size fraction <3 kDaA reflect the importance of natural ligands in controlling the concentrations of Ag released by dissolution of AgNPs. Overall, these data indicate that AgNPs are relatively stable in the tested lake environment and appear to result in quantities of highly toxic ionic Ag(+) that are below our limit of detection.

Carbon Nanotube Integrative Sampler (CNIS) for passive sampling of nanosilver in the aquatic environment

“…In the 386 two mesocosms in which CNIS were deployed, the concentrations of total silver (TAg) steadily 387 increased over the course of the study as CT-AgNPs and PVP-AgNPs were added and reached 388 final concentrations that were approximately 60% of the predicted nominal concentration (Figure 389 3). Deviation from the estimated nominal concentration might be explained by losses in the water 390 column through sedimentation and adsorption to the mesocosm walls, as discussed previously by 391 Furtado et al (2015). The estimated TWA concentrations for silver measured in the TAg and the colloidal Ag 398 (i.e.…”

“…Water samples were collected over the 52 day mesocosm 385 study and three size fractions were analyzed, as previously reported by Furtado et al (2015). In the 386 two mesocosms in which CNIS were deployed, the concentrations of total silver (TAg) steadily 387 increased over the course of the study as CT-AgNPs and PVP-AgNPs were added and reached 388 final concentrations that were approximately 60% of the predicted nominal concentration (Figure 389 3).…”

“…0.22 µm filtrate) fraction over each deployment period were calculated by integrating the total 399 area under the accumulation curves over various periods corresponding to the CNT deployments 400 (Table 3). The levels of dAg in the <3 kDa filtrate were not included in this analysis because levels 401 of dAg in the mesocosms were very low, as previously reported by Furtado et al (2015). 402…”

“…Lake 239 is an oligotrophic, dimictic lake. A total of 12 mesocosms were installed in a 191 shallow embayment on the south side of the lake, as described by Furtado et al (2015). to collect and analyze these samples are described in detail by Furtado et al (2015).…”

“…A total of 12 mesocosms were installed in a 191 shallow embayment on the south side of the lake, as described by Furtado et al (2015). to collect and analyze these samples are described in detail by Furtado et al (2015). Briefly, a 215 volume of water from each mesocosm was collected at 30 cm below the surface and passed 216 through a 35 µm screen to remove macroplankton.…”