- Top of page
- MATERIALS AND METHODS
- SUPPLEMENTAL DATA
- Supporting Information
Metal oxide nanoparticles (NPs) (e.g., ZnO, TiO2) as well as fullerenes (e.g., C60) and carbon nanotubes (CNT) are among the most studied NPs in ecotoxicology as a consequence of their wide application in novel technologies [1-3], and their production rate is expected to continue to rise in the coming years . Therefore, it is necessary to evaluate the hazard associated with engineered nanomaterials by evaluating their effects on organisms. For that, fate and distribution of nanomaterials in the environment should be assessed and studied to establish an accurate prediction on exposure and derive risk . The development of nano-related detection techniques is of major importance to allow exposure estimations of nanomaterials that are being released to the environment.
As a consequence of the extensive production and use of engineered nanomaterials, they are likely to be released into the environment. These occurrences inevitably will lead to some degree of environmental and human exposure, instigating the need to catch up the fast growth of this multidisciplinary field .
As metal oxide NPs, zinc oxide NPs (ZnO-NPs) have received much attention as a result of their wide range of applications in nanotechnology. Because of their high ultraviolet radiation absorption , ZnO-NPs have been applied in a variety of personal care products such as sunscreens, toothpastes, and cosmetics [8, 9]. They can also be used in coatings , antifouling paints  environmental remediation processes , wastewater treatment , textiles , ceramics, rubber processing, food additives, and biosensors , or even used as catalysts in batteries  and antibacterial agents .
A considerable amount of literature already exists for ZnO-NPs, reporting the toxicity effects to several organisms, such as bacteria [16-18], algae [19-21], and fish [22, 23], as well as human cells . A recent review published on this topic  lists the toxicity data available for ZnO NPs to aquatic organisms. But most of these studies are based on short-term or acute exposure tests, sometimes lacking ecological relevance, such as long-term effects or important ecological traits or endpoints. Additionally, NPs' size and solubility effects are often not fully considered and toxicity related.
It has been reported that when bulk counterparts are used to produce smaller particles, their physicochemical features change, increasing their surface reactivity , thus enabling them to interact or penetrate more efficiently with or into organisms, possibly triggering adverse responses [8, 13]. Therefore, size is an important NP characteristic to be considered when toxicity or fate studies are conducted.
Daphnia magna is a freshwater invertebrate that has been used extensively for the past 20 yr in regulatory testing and ecotoxicological research. Several features of this invertebrate make it suitable for laboratory testing . In addition to their small size, high fecundity, short life cycle, reproduction by parthenogenesis, ubiquitous occurrence, and ease in laboratory handling , they can be used to evaluate functional traits (e.g., filtration, by feeding inhibition tests) or derive effects from the individual to the population scale (e.g., reproduction tests). Cladocerans have been considered a good toxicity model organism to predict the toxicity of pollutants to ecosystems due to their high sensitivity to environmental pollutants , including to the metal Zn , and representativeness in food-web chains as food and energy link between primary producers and secondary consumers .
To our knowledge, to date only acute or short-term effects of ZnO-NPs to D. magna have been studied, and the effect of particle size or solubility has not been taken into account [30-32]. Therefore, the present study aimed to evaluate the toxicity-related effects of size of ZnO particles to D. magna considering 2 different nanoparticulate sizes (30 nm and 80–100 nm) and a microsized (> 200 nm) particle, which were also related and compared with ionic counterparts. This was carried out by studying the effects on daphnids' immobilization, feeding activity, and reproduction on exposure to ZnO particles and ZnCl2 and by characterizing the exposure for all ZnO particles in American Standards for Testing and Materials (ASTM) hard water (transmission electron microscopy [TEM] and energy-dispersive X-ray spectroscopy), as well as their dissolution behavior in the same medium.
- Top of page
- MATERIALS AND METHODS
- SUPPLEMENTAL DATA
- Supporting Information
It has been reported that at a water pH of 7.8, a 1 g L−1 dispersion of ZnO NPs will lead to a dissolution of 10 mg Zn L−1 (i.e., < 2%) . In the study of Franklin et al. , the dissolution of 100 mg L−1 of 30-nm particles was carried out for 40 h, resulting in a dissolution of 15 mg L−1. This is not in accordance with the results of the present study, in which we observed a lower dissolution of 0.51 mg Zn L−1 for the 30-nm NPs. Although this was not expected, the discrepancies on the values found can be related to the quality of the NPs, such as the content of hydroxides (content of the amorphous and crystalline phases), but also on the media used for their dispersion.
The rate of dissolution of a particle is usually considered proportional to its surface area, considering also the same shape, which will lead to a faster dissolution for lower-scale particles when compared with larger-sized particles or bulk materials, for the same mass. In the present study, the >200-nm ZnO particles showed a higher dissolution in 48 h than the 2 NPs studied, which goes against this theory and also against results obtained by Franklin et al. , where similar dissolutions were observed between ZnO-NPs of 30 nm and bulk ZnO particles. Therefore, dissolution may be considered not just dependent on NP size but also on details of their nanostructure and most likely on the agglomerates' size and structure, which are also dependent on the methodology used for production and functionalization and the media in which they were dispersed.
The LC50 values for the acute toxicity tests to D. magna ranged between 0.76 mg Zn L−1 and 1.32 mg Zn L−1. For the same species and clone, Heijerick et al.  reported a 48-h EC50 value for the acute toxicity of Zn of 0.39 mg Zn L−1, which is lower than the LC50 values obtained in our results. The differences in toxicity may be due to the different concentrations of cations used to prepare both media (ASTM and M4 medium), because according to their results, the toxicity of Zn can be reduced by different concentrations of cations (e.g., Ca2+, Mg2+, Na+, K+) used in test media.
The acute toxicity of both NPs (30 nm and 80–100 nm) and the microsized ZnO particles were found to be similar and not dependent on the initial particle size. Differences in agglomerate sizes were observed after 48 h in ASTM hard water and were not directly related to their initial size. Previous studies have already reported similar toxicity between nanoscale ZnO particles and the respective bulk counterparts [22, 31, 42]. Although it is known that particle size plays an important role in the NPs' toxicity [14, 43], this was not observed in the present study, where there was no relationship between the toxicity and initial particle size, which can be justified by the fact that daphnids were not exposed to the individual NPs but instead to the agglomerates that may have modified its toxicity. This lack of relationship observed between the acute toxicity of nanomaterials and their size is in agreement with a previous study in which all Zn forms were very toxic, with similar LC50 values between them all .
From the EC50 values obtained for the feeding behavior of daphnids, based on zinc concentrations, it can be observed that test organisms were more sensitive to ZnO-NPs than to the ionic zinc, with lower values for the ZnO-NPs of smaller particle size (Table 2). Higher concentrations of ionic zinc (> 3.2 mg Zn L−1) will be needed to inhibit their feeding activity in 50% of the population. Therefore, considering that the EC50 values based on zinc for NPs were < 2 mg Zn L−1 and lower than the ionic zinc concentration inducing feeding inhibition, we assume that the concentrations of ZnO particles affecting D. magna were mainly constituted by the particulate form. In addition, ZnO agglomerates changed their size throughout the 48 h, with 30-nm ZnO-NPs showing larger agglomerates after this period.
Several studies have suggested that the toxicity of ZnO-NPs is mainly the result of the dissolution of zinc ions present in suspensions [19, 29]. This is not yet a consensual assumption, however, because other studies believe that their dissolution does not account for the total observed toxicity of ZnO-NPs , as the present study has also shown.
As mentioned above, particle size is one of the features that bring special attention to NPs studies. Their smaller size provides a larger surface area and reactivity, thus allowing them to penetrate into cells and organisms more efficiently. This will possibly induce higher toxicity effects where the same material in the bulk form may be inert [13, 14, 43]. This was partly observed in our results in the feeding behavior tests for the 2 NPs studied (30 nm and 80–100 nm), possibly because toxicity was mainly driven by the particulate form. For the acute toxicity, however, microsized ZnO particles presented a lower LC50 value when compared with those of ZnO-NPs (Table 2). In this last case, the acute toxicity was possibly driven by both the particulate and the ionic forms. When looking at the dissolution rates, the >200-nm particles released more ions during 48 h when compared with both NPs, which may then explain the similarity between LC50 values found between these and the ionic form.
It is also known that NPs are susceptible for aggregation in aqueous suspensions. This process can change their physicochemical properties, making them less available to cause toxicity [11, 44]. Therefore, regarding the zeta potential values obtained, we can consider that the particles of >200 nm with a highest negative value (–19.8) would be more stable than the NPs whose zeta potential values were closer to 0 due to electrostatic stabilization. Their dissolution was higher, however, which contradicts these findings to a certain extent.
For the postexposure period, it was observed that daphnids were not able to recover from the ionic zinc previous effects, as shown in Figure 2, even at low concentrations (0.4 mg Zn L−1 and 0.8 mg Zn L−1), where effects were not detected during the exposure periods. For the 2 NPs studied and the microsized particles, all concentrations tested in the exposure period were still inducing lower feeding rates than the control during the 4-h period. This highlights that even after short exposure periods (e.g., 24 h) there will be long-lasting effects in daphnids exposed to Zn forms.
The number of offspring produced was the most sensitive endpoint used in the present study. No published data about chronic toxicity of ZnO-NPs to the cladoceran D. magna was found in the literature for comparison. Taking into account the key water characteristics used in the ecotoxicological tests, however, Heijerick et al.  reported EC50 and no-observed-effect concentration values for Zn toxicity to D. magna of 0.0083 mg L−1 and 0.0055 mg L−1, respectively. These values are lower than those obtained in the present study, which again may be related to the cations composition of the test medium.Given the scarcity of studies concerning long-term effects of NPs, it becomes difficult to assess, compare, and understand their potential risk to aquatic environments.
Contrary to results of the immobilization tests and similarly on the feeding inhibition tests, the effect of ionic zinc on the reproduction endpoint did not allow us to observe a complete dose–response curve (EC50 > 0.25 mg Zn L−1). Concentrations inducing both sublethal and lethal effects on reproduction along the 21-d of exposure were very close, with a small interval between them (> 0.25 mg Zn L−1 and < 0.4 mg Zn L−1, respectively; Table 2). This also shows that acute effects on time are important to consider. In this case, results obtained in the reproduction test for the ionic zinc derived a lower LC50 value for 21 d than the value reported for the 48 h (0.76 mg Zn L−1and 0.21 mg Zn L−1for 48 h [without food] and 21 d [within the chronic assay], respectively).
Similar to our results, Sánchez-Ortíz et al.  also showed a decrease in population growth in 2 cladocerans species (Ceriodaphnia dubia and Daphnia pulex) with increasing concentrations of zinc in the medium. These authors observed no reproduction at 1 mg Zn L−1 and mortality after 1 wk at concentrations ≥ 0.25 mg Zn L−1. A similar pattern was observed in our results. From approximately the second week of exposure to ionic zinc, daphnids showed low reproduction rates and started to die in all the replicates of the highest 3 concentrations used (0.30 mg Zn L−1, 0.35 mg Zn L−1, and 0.40 mg Zn L−1), reaching adult mortality rates > 50%. Therefore, according to the standardized protocol followed, these data were not included in the statistical approach. Contrary to what was found in the present study, Sánchez-Ortíz et al.  showed that the population growth of D. pulex was higher at 0.125 mg Zn L−1 than in the control, resulting in an hormesis (stimulatory effect of sublethal concentrations).
When looking at the EC50 values obtained for reproduction performance (Table 2) and the dissolution results (Table 1), it can be concluded that the effects are driven mainly by the ionic form of Zn. At the end of the 48-h period of the dissolution experiments, the amount of zinc dissolved was similar to or higher than the EC50 causing reduction in the number of offspring.
During the reproduction tests in the present study, daphnids visibly showed difficulty changing their moults after 2 wk of exposure (except the control) and consequently died (data not shown). A recent study of D. magna suggested that exposures of waterborne zinc resulted in whole-body zinc burden and consequently leads to a decrease of Ca2+ body contents as a result of the competition between these 2 ions for ion regulatory surfaces . The depletion of Ca2+ levels (hypocalcemia) inhibits the filtration rates of daphnids, leading inevitably to a decrease of food uptake . As a consequence, less energy will be available to allow a normal growth rate and reproduction [28, 47]. These findings could then explain the results obtained in the present study for the low feeding rates and the reproduction output of daphnids at the highest concentrations tested. Therefore, we can hypothesize that surplus zinc levels can indirectly affect the moulting process of D. magna.