Research highlights: improved understanding of ecological impacts resulting from nanomaterial-based in situ remediation

Abstract

Nanomaterials are currently being used for in situ remediation of soils and groundwater. However, the continued use of currently implemented nanomaterials and the systematic development of more effective and ecologically benign materials require a more complete understanding of their ecological impact, which should include the transport through the subsurface, acute, chronic and long term effects of exposure, and the role of nanomaterial characteristics (e.g., composition, surface coating). In the current highlight, three articles that examine different aspects of nanoscale zero-valent iron (nZVI) transport, reactivity or exposure to model organisms are summarily reported, which advance the development of more sustainable remediation approaches. The first study examines the role of a model biofilm on the transport of different Pd-doped nZVI species through granulated media, and also the associated nanomaterial toxicity to the forming and sessile bacteria. The second study examines the multigenerational reproductive impacts of C. elegans resulting from nZVI exposure. Lastly, the resulting products of nZVI reactivity with U(VI) species at environmentally relevant molar ratios are examined, and a thorough analysis of the resulting products are reported, which provide valuable data for predicting the consequential role nZVI remediation will have on the ecosystem at and near contaminated sites.

Nano impact

The continued implementation and refinement of more effective nanomaterial based technologies for environmental remediation will occur in part through an improved understanding of the ecological risks associated with different iterations of the delivery method, nanomaterial composition, and long term effects of use. Although nZVI based strategies are already in use, continued work to understand their impact on organisms directly exposed at the contaminated sites and the byproducts of use is necessary for further implementation of current practices and to guide the development of future technologies.

Introduction

Nanoscale zero valent iron (nZVI) is ubiquitously used throughout the world for remediation of contaminated soil, groundwater and wastewater. However, concerns about the impacts of different nZVI to ecosystems have remained, and recent reviews outlining the concerns and knowledge gaps in the field have been published (Lefevre et al., Sci. Total Environ., DOI: 10.1016/j.scitotenv.2016.02.003; Das et al., Environ. Sci. Pollut. Res., 2015, 22, 18333, DOI: 10.1007/s11356-015-5491-6). More specifically, the transport and toxicity to organisms that are exposed to nZVI analogues need to be better understood, which should include model systems that can systematically evaluate the risks associated with their use in in situ applications. Basnet et al. examined the contribution of model biofilms to the transport of Pd-doped nZVI materials with different coatings, where the toxicity of the nanomaterials was also examined on growing and sessile state bacteria (Basnet et al., Environ. Sci.: Nano, 2016, 3, 127, DOI: 10.1039/c5en00109a). In work by Yang et al., the first multigenerational investigation of C. elegans after exposure and removal of nZVI materials at concentration ranges measured at remediation sites was conducted to examine the long term effects of exposure (Yang et al., Chemosphere, 2016, 150, 615, DOI: 10.1016/j.chemosphere.2016.01.068). The last study examined the reactivity and subsequent products resulting from implementing nZVI-based remediation in contaminated aquifers. The work by Tsarev et al. examined the reactions of U(VI) with nZVI, including specific product formation and effectiveness for uranium immobilization at a range of molar ratios that have been observed in the environment, using a suite of characterization methods (Tsarev et al., Environ. Sci. Technol., 2016, 50, 2595, DOI: 10.1021/acs.est.5b06160).

Nanoparticle transport and toxicity in biofilm-coated sand

The work by Basnet et al. (Basnet et al., Environ. Sci.: Nano, 2016, 3, 127, DOI: 10.1039/c5en00109a) examined the transport of nanomaterials in model aquifer systems to systematically evaluate the roles that composition, engineered coating, and subsurface granular media characteristics impart on their effectiveness as a remediation method for contaminants. The authors examined Pd-doped nZVI (Pd-nZVI) coated with 700 kDa carboxymethyl cellulose (CMC) or 577 Da rhamnolipid (RL) biosurfactant through packed columns of quartz sand with and without homogeneous biofilms comprising P. aeruginosa. Importantly for the development of remediation methods, the transport behavior of the Pd-nZVI was examined in the presence of the biofilm, but the impact of the material on the biofilm was also investigated over 24 h in diluted cell culture media.

To more completely describe the transport behavior through the column and observed effects of nanoparticle aggregation on toxicity, the colloidal stability and electrophoretic mobility (EPM) of the Pd-nZVI particles were examined in monovalent and divalent salt solutions, where the authors reported increased aggregation and decreased magnitude of the EPM with increasing ionic strength. The extent of aggregation was higher in the presence of the divalent CaCl₂ solutions, where the authors attributed the lower EPM to charge screening and Ca²⁺-induced bridging between the engineered coatings' carboxylic functional groups.

The authors reported similar trends in transport behavior for the Pd-nZVI in solutions containing the monovalent and divalent salt solutions and in the presence of the biofilm, but distinctions based on the surface characteristics were reported. A semi-quantitative comparison of the transport potential for the two Pd-nZVI species was reported using their calculated attachment efficiency (α_pc), which was compared to the authors' previous work on transport through the same uncoated sand (Basnet et al., Water Res., 2015, 68, 354, DOI: 10.1016/j.watres.2014.09.039). In NaCl solutions (Fig. 1a and b), α_pc increased in the presence of biofilm and also increasing ionic strength, which was consistent with the authors' previous work in the absence of the biofilms. However, the authors reported that the relative increase in attachment efficiency for experiments containing the biofilm was higher for RL-coated materials than their CMC-coated counterparts. Furthermore, the increase in α_pc was reported to be larger for higher concentrations of the nanoparticles (0.15 g L⁻¹ to 1 g L⁻¹), which the authors attributed to significant reduction in the nanoparticle transport when the biofilms are present and suggest a possible overestimation of transport potential when clean sand experiments are used for natural environments. In the CaCl₂ solutions (Fig. 1c and d), enhanced deposition resulting in larger α_pc values was reported, but significant change from the bare sand was only reported at 3 mmol L⁻¹ CaCl₂ for the RL-coated nanoparticles. The authors reported no significant changes in α_pc for CMC-coated nanoparticles with increasing ionic strength, but α_pc values in the presence of the biofilm were up to 26-fold higher than the transport experiments in uncoated sand. The authors attributed these observed differences for both monovalent and divalent studies to preferential retention of larger aggregates, which were examined in the influent and effluent using dynamic light scattering (DLS) and showed a decrease in hydrodynamic size from the influent to the effluent; thus, physical straining was proposed as an important mechanism for nanomaterial retention in the presence of the biofilm.

Fig. 1 Calculated particle-collector attachment efficiency (α_pc) in monovalent salt (NaCl) for (a) RL-coated and (b) CMC-coated Pd-nZVI in clean and biofilm-coated sand columns. The α_pc values in divalent salt (CaCl₂) for (c) RL-coated and (d) CMC-coated Pd-nZVI in clean and biofilm-coated sand columns. Error bars represent standard deviations. The dashed lines are included to guide the eye. The circular red and triangular blue symbols in Fig. 1a and b correspond to the transport experiment conducted at higher Pd-nZVI concentration (1 g L⁻¹) in biofilm coated-sand and clean sand, respectively. For all other conditions, the particle concentration was 0.15 g L⁻¹. Reprinted with permission from Basnet et al., Environ. Sci.: Nano, 2016, 3, 127. Copyright 2016 Royal Society of Chemistry.

Because the responsible use of these materials for remediation applications requires limited impact to the ecosystem, the interaction of the materials with the biofilms was investigated. Based on optical density measurements of the films, quantitative estimates of the biofilms formed were investigated at two different concentrations of the coated Pd-nZVI and compared to the bare nanoparticles. In the presence of RL-coated species, a decrease in biofilm formation was observed at both concentrations. In contrast, a slight decrease in biofilm formation was observed for the CMC-coated nanoparticles at the lowest concentration (15 mg L⁻¹) and increased formation at the higher concentration (150 mg L⁻¹), which was attributed to the nanoparticles because the supernatant containing CMC was reported to also inhibit growth. Furthermore, no statistical difference was observed for the viability of the biofilms exposed to nanoparticles that were grown for 24 h even at 1 g L⁻¹, which suggests the exposure does not affect membrane integrity. The authors attributed the observed results to an extracellular aggregation process that limits the impact on cell viability, as was previously reported for silver nanoparticles.

Overall, the current work provides a systematic study of the transport properties of candidate nanomaterials for targeted remediation, which provided further insight into the role biofilms likely play in transport through granulated media and the role of the nanomaterial properties that should be evaluated to implement sustainable nanomaterial-based solutions. The authors also suggested similar additional studies that investigate the role of nanomaterial aging for more accurate predictions.

Investigating long-term toxicity of nZVI-based remediation

Studies examining the effects of growth and development of microorganisms and viability in the sessile state as observed in the previous section (vide supra) are an important contribution to the overall development of sustainable approaches to nanotechnology implementation. Another essential consideration is the long term impact of the introduction and exposure of the nanomaterials to the ecosystem.

The work by Yang et al. (Yang et al., Chemosphere, 2016, 150, 615, DOI: 10.1016/j.chemosphere.2016.01.068) examined the multigenerational (reproductive) toxicity of C. elegans when exposed to three iron species, including CMC-coated nZVI at concentration levels found at test sites after ground remediation with similar nZVI materials (Wei et al., Water Res., 2010, 44, 131, DOI: 10.1016/j.watres.2009.09.012). The authors first tested the mortality of the invertebrates in embryo-rearing media, which did not result in significant increase of C. elegans mortality and was consistent with previous studies. These results were consistent with other Fe based species, where Fe₃O₄ nanoparticles and Fe(II)_aq also did not show increased mortality compared to the control. For the reproductive toxicity measurements, the organisms were exposed to increasing concentrations of the different iron species from (5 to 100) mg L⁻¹ and denoted as the parental generation, F0. The subsequent filial generations were evaluated based on number of offspring per worm and brood size. A dose dependent response in the number of F0 offspring was reported, where the highest adverse effects were observed for Fe₃O₄ nanoparticles and Fe(II)_aq but effects were observed for all Fe species at 5 mg L⁻¹ (Fig. 2A). The authors also examined the brood size of the offspring, which is reported to be a predictor for a toxicant's ability to affect reproduction, and observed similar decreased offspring numbers at the highest concentration of Fe containing species (Fig. 2B).

Fig. 2 Reproductive toxicity of CMC-nZVI, nFe₃O₄, and Fe(II)_aq exposure in F0 and F1 generations of C. elegans. (A) Synchronized wild-type L4 larvae of F0 generation were exposed to CMC-nZVI, nFe₃O₄, and Fe(II)_aq for 48 h with E. coli OP50 added as a food source. Subsequently, the offspring of each worm were scored. (B) The total brood size/offspring of each worm in F1 generation was scored. Differences compared to the control (non-exposed) were considered significant at p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) by one-way ANOVA and Games-Howell post-hoc test. N.D., not detectable. Reprinted with permission from Yang et al., Chemosphere, 2016, 150, 615. Copyright 2016 Elsevier Ltd.

Because the two highest concentrations of Fe species resulted in the most toxic responses, additional experiments were reported that examined the multigenerational effects of the CMC-coated nZVI particles after the nZVI was removed. A significant decrease in the brood size was observed in the F1 and F2 generations. However, a recovery of brood size to the control levels was observed in F3 and F4 generations. The authors reported the cause of the observed decrease was not due to oxidative stress induced by the nanoparticles or dissolved ions, which is a commonly reported mechanism for toxicity for nZVI, based on the measured, unchanging reactive oxygen species accumulation. The authors posit the increased Fe accumulation in the F1 generation as the cause for the observed increase in toxicity. However, the authors suggest further work is necessary to investigate the role of Fe accumulation on reproductive toxicity. Overall, the work provides further information on the long term impacts of Fe-based nanomaterial remediation practices on exposed ecosystems.

Elucidating byproducts resulting from in situ remediation

Another important factor in the development of viable remediation methods is determination of the reaction products with the contaminant and their associated risks. The work by Tsarev et al. (Tsarev et al., Environ. Sci. Technol., 2016, 50, 2595, DOI: 10.1021/acs.est.5b06160) examined the interaction of nZVI with highly mobile U(VI) to identify the resulting distribution of solid phase products over a range of U : Fe molar ratios and over long timescales, 1 year. The composition of the synthesized nZVI was first examined by dissolution experiments and X-ray absorption near edge structure (XANES) measurements. The authors reported ≈75% of the particle was Fe⁰ with good agreement between both measurements. The initial reaction of the U with the nanoparticles in the anaerobic chamber resulted in a decrease in the soluble U(VI) concentration to below detection limits in less than 1 h. The resulting products were examined with a suite of instrumentation.

The most salient peaks observed in diffraction data were attributed to Fe₂(OH)₂CO₃ (a known product of anaerobic Fe⁰ corrosion), magnetite, and UO₂. The rates of formation of discernable UO₂ diffraction peaks were distinct for each U : Fe molar ratio, where the 1 : 4 ratio exhibited observable peaks after 1 day that were similar in intensity to the 1 : 21 sample after 8 days. To examine the morphology and distribution of the U and Fe species, the authors used transmission electron microscopy (TEM) and energy dispersive X-ray spectroscopy (Fig. 3). The TEM images exhibited fibrous structures consistent with Fe₂(OH)₂CO₃ and found the Fe and U distributed in two distinct phases, Fe-rich fibers and U-rich aggregates. Evidence was present for both Fe-substituted mixed phases and sorbed U onto the surface, and additional measurements were conducted to examine the formation and evolution of each phase.

Fig. 3 (Left) Phase-uncorrected Fourier-transformed EXAFS spectra and (center) corresponding raw EXAFS data (k-weight = 3) of selected samples. (Right) High-angle annular dark-field image with Fe and U elemental contents colored with yellow and cyan, respectively. Reprinted with permission from Tsarev et al., Environ. Sci. Technol., 2016, 2595. Copyright 2016 American Chemical Society.

Examination of the edge energy positions in extended X-ray absorption fine structure (EXAFS) spectra provided further information about the mixed phase products. For the 1 : 4 molar ratio sample, a model containing U–O and U–U single scattering pathways was sufficient to match the observed data, indicating the sample was UO₂. In the 1 : 21 sample, additional scattering pathways were necessary and provided further evidence for either Fe-substitution or U sorption. The authors also posited the evolution of uranyl species from 1 day that are transformed from a U distribution that contains U(V) species and U complexed with carbonate. At 8 days, the assigned uranyl contribution is removed, and the predominant signatures are from UO₂ and U complexed with CO₃, where the authors attributed the latter to sorption. In lower molar ratios of U (1 : 110), the U–U pathway was not detected and the authors surmised that sorption was the primary reaction pathway. They also examined the Fe–U atomic distances, which provided further evidence for sorption onto the Fe₂(OH)₂CO₃.

Overall, the work provides a comprehensive evaluation of the products that form under representative field-scale applications of nZVI. The authors reported U immobilization after reaction with nZVI remained complete for at least 1 year in anaerobic environments, and they suggested the application may also be effective in O₂-limited diffusion environments. Understanding the interaction, structure and incorporation of U in the product distribution at environmentally relevant conditions provides necessary information to more efficiently develop and implement engineered nanomaterial-based remediation strategies.

---