Knowledge of the toxicity of nanomaterials and nano-enabled products is very important for directing their uses and protect the environment from possible adverse effects. The understanding of their ecotoxicity is fundamental to environmental health and safety as well as for public acceptance 1. Nanomaterials have been extensively produced and used in a wide spectrum of applications, including the enhancement of existing products, the advancement micrometer-sized technology, and the development of new medical and imaging techniques. Therefore, it can be assumed that so-called nanowastes are being generated and are reaching the aquatic environment 2. Although many types of nanomaterials exist, metallic nanoparticles are useful because they are easily visualized, easily modified to have different surface chemistries, and easily made into a wide variety of shapes and sizes 3.
Silver nanomaterials are the most common materials incorporated into commercial products. More than 30% of the nano-enabled products currently on the market contain silver nanoparticles because of their antimicrobial properties. This means that silver nanoparticles have great potential to reach the environment 4. The oxidation of silver nanoparticles and the subsequent release of silver ions were thought to be the major mechanism by which they pose a hazard to the environment 5, 6. The high toxicity of silver ions and their bioaccumulation potential are already known 7, 8 to interfere in the regulation of sodium uptake in aquatic organisms 8, 9.
Recently, a new type of silver nanomaterial with antimicrobial activity has been described. It is a silver vanadate nanowire with micrometer length and diameter of less than 100 nm, decorated with silver nanoparticles 10. These nanowires have shown antibacterial activity to three different strains of Staphylococcus aureus, with a minimum growth inhibitory concentration value, 10-fold lower than oxacillin 10. The proposed antibacterial mechanism was that chemisorbed Ag+ is released from Ag nanoparticles through a direct dissociation between the nanoparticle and the bacterial cell wall causing toxic effects 11. They are also being considered as a potential antifouling agent; therefore, the release of this material in the environment is possible. The ecotoxicity of silver nanomaterials has been extensively evaluated 9, 12, 13. The predominant studies are of silver nanoparticles with a spherical shape 14–16; however, there are few studies that address nanowire ecotoxicity.
George et al. 4 tested different silver nanoparticles, including silver nanowires, and found that silver nanoparticles with different shapes cause toxicity in fish cell lines and zebrafish embryos. They showed that the surface defects in the crystal structure play a role in the toxicity along with silver release. Silica nanowires have been tested in zebrafish embryos and cell cultures 17, 18. Silicon carbide nanowires (SiCNW) were tested with different organisms such as Hyalella azteca, Chironomus dilutus, Lumbriculus variegatus, and Lampsilis siliquoidea19, but no literature was found using Daphnia to test the toxicity of nanowires. Daphnids are considered a keystone species in aquatic toxicology because they are filter-feeders and are able to ingest nanoparticles. They have also been proposed as a model organism for the ecotoxicological testing of nanomaterials 20, 21.
The present study sought to assess the toxicity of silver vanadate nanowires decorated with silver nanoparticles using Daphnia as a test organism and to evaluate the role of silver and vanadium release in the toxicity of this material.
MATERIALS AND METHODS
Synthesis of the silver vanadate nanowires
Briefly, a solution of silver nitrate (2 mM in 35 ml; Merck, 99.8%) was added dropwise in 35 ml of an ammonium vanadate solution (1 mM; Merck, 99%) under mechanical agitation and heated for 10 min at 70°C. The mixture (100 ml) was then placed in a hydrothermal reactor with pressure gauges and a thermocouple inside the reaction medium. This system was heated at 180°C for 16 h. The resulting product was removed and centrifuged at 775 g for 10 min to separate the nanowires, which were then washed thoroughly with absolute ethanol and dried under vacuum for 10 h. This sample was called SVSN-LQES1.
Fourier transform infrared (FTIR) spectra were recorded on an ABB Bomem MB-series spectrometer, model FTLA2000-102, using the transmittance mode in the 4000–200 cm−1 wavenumber region. The sample was diluted and pelletized in analytical grade KBr. The X-ray powder diffraction (XRD) analysis was performed at room temperature using a Shimadzu XRD-7000 diffractometer operating with CuKα radiation (λ = 1.5406 Å, 40 kV, 30 mA). Data were collected in the 10° < 2 θ < 70° range with a speed of 2° min−1. The morphology of the obtained silver vanadate nanowires was confirmed using scanning electron microscopy (Jeol, JMS 6360LV) and transmission electron microscopy (Zeiss CEM-902 analyzer with EELS). The zeta potential of SVSN-LQES1 was determined in Zeta Sizer (Malvern) in a suspension of 1 mM/L KCl and the surface area determined (Quantachrome, NOVA 4200 E).
Silver vanadate nanowires (SVSN-LQES1) at 100 mg/L were sonicated in an ultrasonic bath (Colle-Palmer 8891) for 40 s in MilliQ water before testing. The material was tested using appropriate dilutions. To investigate if the presence of the nanowires in the suspension was promoting toxicity, we tested this solution before and after filtration using 0.45-µm cellulose acetate membranes.
Acute toxicity tests were performed with Daphnia similis according to the Organisation for Economic Co-operation and Development test guideline 23. For each of four replicates, five organisms 6- to 24-h-old were exposed for 48 h in 10-ml test solutions or suspensions under static conditions at 20°C ± 2°C in the dark. After exposure, immobilized organisms were counted.
For comparison, solutions of Ag and V ions were prepared and tested in parallel with the nanowires. Silver nitrate (AgNO3; Merck, 99.8%) and vanadium pentoxide (V2O5; Sigma Aldrich, 98%) were used as the source of Ag and V ions. The V solution was sonicated in an ultrasonic bath for 4 min to enhance solubility. To verify the possible synergistic effect of V and Ag we prepared a 1:1 (w/w) solution of both ions.
Tests were considered valid when mortality in the negative controls did not exceed 10%. The results were statistically analyzed using the Trimmed Spearman–Karber method 24 for estimating the median effective (immobilization) concentration (EC50), which was expressed in micrograms per liter. The observed mortality of Daphnia (binary data) was modeled by logistic regression with the following generalized logistic model: logit (p) = µ + b (concentration), where logit (p) is the log odds of a Daphnia dying, µ is an intercept value, and b is the fixed effect of concentration. Model estimation was performed by iterative estimation of the likelihood function (Systat Software). The estimate of logit (p) was used to obtain the predicted probability of mortality (p): P = elogit(p)/(1+ elogit(p)).
The sensitivity of the D. similis culture was monitored monthly with sodium chloride as a reference substance, and the culture was not used if the results were not within two standard deviations of the average EC50 obtained for the reference substance.
Total silver determination
Total Ag release was determined in a sonicated (40 s) and filtered (0.45 µm) stock suspension of the SVSN-LQES1 at 100 mg/L prepared in Daphnia testing medium. We did not expect that the nanowire would pass through the 0.45-µm filter because of its length. However, we could not rule out that some Ag nanoparticles detached from the nanowires could pass through the filter. Total Ag concentration was measured by inductively coupled plasma-atomic emission spectrometry (ICP-AES, OPTIMA 3000DV; Perkin Elmer). A calibration curve was performed with standard solutions of Ag (silver AAS standard; Tec Lab) and a correlation factor (r2) of 0.998 was obtained. A sample dilution in 2% (v/v) of HNO3 was introduced into the spectrometer to be atomized.
RESULTS AND DISCUSSION
The silver vanadate nanowire decorated with Ag nanoparticle samples, SVSN-LQES1 showed a crystalline structure that can be assigned to β-AgVO3 (JCPDS 861154) according to XRD data (Fig. 1A). The diffraction peaks at 2θ = 38.11°, 44.30°, and 64.44° were assigned to metallic silver (Ag0) (JCPDS 870597). Figure 1B shows the FTIR spectra of the silver vanadate nanostructures. The bands observed in the range 400 to 1000 cm−1 can be assigned to the V-O vibrations 25. The band observed at 848 cm−1 can be attributed both to the V-O stretching vibration and the Ag-O-V vibration 26. We also observed a band at approximately 1632 cm−1 assigned to the vibrations of water molecules. The band observed at 964 cm−1 was assigned to the asymmetric stretching vibrations of VO3 terminal groups 27. Some analyses were performed with transmission electron microscopy (TEM) in dark-field showing difference in contrast between the structures. This contrast difference suggests that the nanoparticles on the surface of the silver vanadate nanowires have metallic character 10.
The SVSN-LQES1 showed a high morphological yield, which displayed as a mesh of nanowires decorated with silver nanoparticles presenting a width of a 100 nm (Fig. 2). The diameter of the silver nanoparticles that decorated the surface of these nanowires varied in size and its distribution is presented in Figure 3. The formation of Ag nanoparticles on the surface of the nanowires can be attributed to reduction of AgNO3. In the studied conditions, NH4VO3 reacts with AgNO3 forming silver vanadate at the same time that the reduction generates Ag ions on the nanowires, which act as nucleation and growing sites of the Ag nanoparticles 28. The nanowires presented zeta potential of −20.1 mV and surface area of 23 m2/g−1.
Sonication (40 s) of nanowires was necessary to promote the dispersion of the material in the toxicity test solution. The integrity of the nanowires was evaluated before and after sonication (40 s) and the sonication resulted in nanowires of shorter length (Fig. 2A,B).
To preliminarily test the nanowires and investigate if the presence of the nanowires in the suspension was promoting toxicity to Daphnia, we tested them before and after filtration. The 48h-EC50 obtained for the nanowire suspension was 1 µg/L expressed in mass of nanowires. After filtration, the toxicity decreased to an equivalent EC50 of 34 µg/L (expressed in equivalent mass of nanowires) indicating that the presence of the nanowires was influencing the observed toxicity.
It is expected that SVSN-LQES1 in suspension would release Ag and V to testing media. We tested ionic Ag and V solutions. Three independent experiments were performed and V showed to be much less toxic than Ag (Fig. 4). The mean EC50 for Ag was 1.1 µg/L (SD = 0.67). The mean EC50 of V was 1400 µg/L (SD = 120). Both values are similar to the EC50 reported in the literature 5, 29. We can conclude that V, if released from the nanowires, does not seem to contribute to the toxicity (Fig. 4). Because V and Ag have the same mode of action, inhibiting the (Na, K)–ATPase 30, 31, an additional experiment was conducted to verify any synergistic or additive effect. We tested Ag and V ions combined in one solution 1:1 and the toxicity of the mixture corresponded to the Ag toxicity (Fig. 4).
Therefore, the released Ag could be responsible for the toxicity of the SVSN-LQES1, but the percentage of Ag released from the nanowires was 0.9% (0.85 mg/L of Ag in a100 mg/L). Three independent experiments were performed for the SVSN-LQES1 suspension. The toxicity results were expressed in the relative amount of Ag expected to be released from nanowires; when compared to the toxicity of Ag solution, it was evident that the toxicity of the nanowires could not be explained by the release of Ag in the solution (Fig. 5). Some studies also observed that the release of Ag was not the only factor responsible for the toxicity of the tested nanomaterials 4, 32.
Therefore, in our preliminary tests when SVSN-LQES1 was filtered, the remaining toxicity could be explained by Ag ions released to the test media along with some small amount of nanoparticles that eventually passed through the filter.
Nanowires can be trapped in aquatic organisms, staying in their guts for a time before being eliminated 33–37. This would increase the exposure to the Ag ions released by the SVSN-LQES1. In this scenario, the concentration of Ag ions inside the gut would be greater than the concentration of Ag ions in the media, or the ingestion of nanomaterial by testing organisms could occlude the digestive tract limiting the respiratory function and consequently decreasing mobility 19.
During the experiments, we also observed an increase in the size of droplets in the Daphnia as a response to the nanowire exposure (Fig. 6). Asghari et al. 38 also described small bubbles under the carapace of the Daphnia exposed to Ag nanoparticles that were similar to the droplets observed in our study. These droplets are consistent with lipid deposits, which in Daphnia are frequently observed when the animals are living in good conditions 39. However, Sosak-Świderska et al. 40 observed an increase in the size of lipid droplets in Daphnia magna when exposed to parathion. In our laboratory, we also observed droplets in D. similis when the organisms were exposed to multiwall carbon nanotubes (J. G. Honório, Faculty of Technology–Unicamp, Limeira, São Paulo, Brazil, unpublished data). Further studies will be necessary to investigate the importance of this response when Daphnia is exposed to nanomaterials or other toxicants.
Silver vanadate nanowires decorated with silver nanoparticles (SVSN-LQES1) are acutely toxic to D. similis. The release of silver from the nanomaterial trapped in the gut along with the silver released to the test media seems to be responsible for the observed toxicity. Although toxic to Daphnia, V does not contribute to the toxicity of SVSN-LQES1. The increase in lipid droplets appears to be related to the exposure of the organisms to the nanomaterial, but the significance of this response needs further investigation.
The authors acknowledge financial support from the São Paulo Research Foundation (FAPESP), National Council for Scientific and Technological Development (CNPq), and Coordination for the Improvement of Higher Education Personnel (CAPES). R. Holtz acknowledges FAPESP for the PhD scholarship Grant 2008/57974-9. This is a contribution of the Millennium Institute of Complex Materials and National Institute for Science, Technology and Innovation of Functional Complex Materials. The authors also thank M. B. Bohrer for the identification of the lipid droplets, E. Zeiger for helpful comments, and T. Henry for the statistical analysis and helpful comments.