Environmental Toxicology and Chemistry, Vol. 33, No. 5, pp. 1122–1128, 2014 # 2014 SETAC Printed in the USA
INTERACTION OF FUNCTIONALIZED FULLERENES AND METAL ACCUMULATION IN DAPHNIA MAGNA ZHI-GUO YU and WEN-XIONG WANG* Division of Life Science, The Hong Kong University of Science and Technology, Kowloon, Hong Kong, China (Submitted 8 October 2013; Returned for Revision 12 December 2013; Accepted 10 January 2014) Abstract: In aquatic environments, transformation of pollutants by association with functionalized carbon-based nanoparticles can dramatically change their cycling pathways. The present study quantified the uptake and depuration behavior of cadmium and zinc bound with functionalized fullerene nanoparticles (f-nC60) in a freshwater cladoceran, Daphnia magna, in a well-dispersed medium. Metal uptake proceeded with a linear pattern during the 8-h exposure period, and the uptake rate constants (ku) were 1.3-fold to 1.4-fold higher for Cd or comparable for Zn bound with f-nC60 than those of the free ones. The assimilation efficiencies of Cd and Zn bound with f-nC60 were significantly enhanced when compared with those metals bound with algal food. Furthermore, the depuration of metals bound with f-nC60 was relatively slower compared to the depuration of metals bound with carbon nanotubes. A longer exposure to f-nC60 resulted in an even slower depuration of metals. The authors conclude that metal binding with f-nC60 as modified nanoparticles could serve as a new pathway for the elevated metal accumulation in Daphnia. Environ Toxicol Chem 2014;33:1122–1128. # 2014 SETAC Keywords: Daphnia
Cadmium
Zinc
Functionalized fullerene
Biokinetics
relatively slow depuration of fullerenes in a freshwater cladoceran, Daphnia magna, likely make these animals carriers of fullerenes to higher trophic levels. Daphnia magna is a filterfeeder and has been a model organism in ecotoxicological studies. Cadmium and Zn are the common toxic metals found in freshwater ecosystems [16,17]. Therefore, the objective of the present study was to examine how functionalized fullerenes (particle size 100 nm) affect the uptake and depuration of Cd and Zn in a freshwater cladoceran, D. magna. Previously, we have characterized the influences of nonfunctionalized carbon nanotubes on the uptake of metals from the dissolved and dietary phases and demonstrated that physical blocking played a role in metal accumulation in D. magna [18]. However, no direct interaction between toxic metals and nonfunctionalized carbon nanotubes was found in this earlier study. In the present study, we investigated the functionalized fullerenes that can bind with metals directly by acid adsorption. Compared with carbon nanotubes, fullerenes are much smaller in size and may have a greater potential to bind with metals because of their high specific surface areas. We conducted the dissolved uptake and depuration experiments of Cd and Zn in interacting with functionalized fullerene nanoparticles (f-nC60). The study will provide important information for our current understanding of the interaction between carbon-based nanoparticles and the biological fate and transport of toxic metals in aquatic organisms.
INTRODUCTION
Among the manufactured carbon-based nanomaterials, fullerenes were among the first detected and bulk-produced in nanotechnology [1]. Subsequent studies reported their interactions with biological targets [2] and raised a wide interest in the behavior of fullerenes in water. Functionalized fullerenes are now rapidly applied in clinical and material usage, such as personal care products, coatings, and drug delivery vehicles [1]. As a result of application of nanotechnological products, the risk of their release into the environment (e.g., waste disposal or accidental release) during or after their use has increased. These nanomaterials may finally enter into ecologically sensitive environments such as aquatic and terrestrial ecosystems [3]. In aquatic environments, agglomeration of fullerene nanoparticles (nC60) can significantly influence their environmental fate, transport, and toxicity [4]. They will interact with and affect the behavior of other pollutants such as organics and metals [5]. Nevertheless, the bioaccumulation of toxic metals interacting with fullerene (and derivative C60) has been rarely investigated. Because fullerenes are entirely insoluble in water, there is a major technical difficulty in the preparation of medium with C60 in aquatic systems [2]. Adding functional groups to fullerene C60 molecules (e.g., –COOH) could increase their hydrophilicity [1]. These functionalized C60 molecules can form negatively charged agglomerates of several C60 molecules (namely, nC60), which are surrounded by water molecules [6,7]. Using water mixing and vigorous stirring in the laboratory, agglomeration allows fullerenes to remain suspended for weeks or months [8]; this mobility likely makes them more available for aquatic organisms. Many studies have investigated the bioaccumulation and toxicological effects of fullerenes (C60) on various organisms [9–14]. Tervonen et al. [15] showed that the rapid uptake and
MATERIALS AND METHODS
Preparation of aqueous f-nC60 suspension
Functionalized fullerenes (1,2-methanofullerene C60-61carboxylic acid, C62H2O2, molecular weight ¼ 778.68 g/mol, as f-C60) was purchased from Sigma-Aldrich and used as received. The diameter of the nanomaterial was not given by the manufacturer. According to the published data [19], the van der Waals diameter of a C60 fullerene molecule is approximately 1.1 nm. However, the average particle size for f-nC60 is
* Address correspondence to [email protected]. Published online 16 January 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/etc.2520 1122
Interaction of fullerenes and metals in Daphnia magna
120 10 nm as measured by dynamic light scattering (Brookhaven Instruments). The calculated carboxyl group content was 5.65% of weight. Stock suspension of f-nC60 was made by mixing 2 mg of fnC60 with 1 g/L polyvinylpyrrolidone (PVP) solution in a 40-mL glass tube, followed by sonication using a probe sonicator at an intensity of 60 W for 30 min [18,20] to disperse the nanoparticles. The suspension did not undergo a filtration process and was collected for further use; the total carbon concentration of the stock suspension was analyzed by a total organic carbon analyzer (minimum detection limit of 0.05 mg/L). The concentration of f-nC60 in the test suspension was analyzed by a spectrophotometer. A calibration curve was established by recording 6 different dilutions of 10 mg f-nC60/L (PVP suspension) at 288 nm (r2 ¼ 0.99). The calibration curve was used to estimate the concentration of f-nC60 in suspensions in the following experiments. It was difficult to reach a full extent of exfoliation by sonication; thus, the f-nC60 concentrations calculated from the calibration curves were estimated values. The average surface charge of functionalized fullerenes was characterized by a phase analysis light scattering technique. Organisms, medium, and radioisotopes
The test organism, D. magna, was cultured in filtered (GF/C Whatman) pond water collected from the Hong Kong University of Science and Technology campus at a density of 1 individual/ 10 mL of water. Before the experiments, the green algae Chlamydomonas reinhardtii raised previously in Woods Hole modified CHU 10 (WC) medium [17] were fed to daphnids at 5 104 cells/mL (neonates 3 d old) or 105 cells/mL (adults >3 d old) daily. The growth conditions for both green algae and daphnids were 23.5 8C with a 14 h to 10 h light to dark cycle. Individual culture systems were maintained for healthy neonates, and then 7-d-old organisms were used for the following bioassays. In all treatments, modified freshwater (SM7) with a low calcium concentration (20 mg/L) [17,18,21] was used. The low hardness medium helped to minimize f-nC60 precipitation. The pH of solutions was adjusted to 8.0 0.2 by spiking 0.01 M HCl or NaOH. Radioisotopes 109CdCl2 (Boston, MA, USA) and 65ZnCl2 (Roskilde, Denmark) were used as tracers in the present study. Radioisotopes were diluted in 0.1 N HCl as stock solution. The radioactivity was measured by a Wallac 1480 NaI(T1) gamma detector at 88 keV for 109Cd and 1115 keV for 65Zn (Turku, Finland). Counting propagated accuracy was greater than 95%. Daphnia effects on stability of dispersed f-nC60 and acute toxicity
In a biokinetic study, a uniform suspension with relative stability was required to provide ideal conditions for the experiment. Thus, the first experiment tested the effects of daphnids on the stability of f-nC60 suspension at different concentrations by monitoring the concentration and zetapotential of f-nC60 over time. Specifically, 200-mL breakers were filled with 100 mL of test solution (PVP dispersed f-nC60 in modified SM7) at different concentrations of f-nC60 (0.5 mg/L, 1 mg/L, and 4 mg/L). Six replicates were prepared for each concentration treatment. Ten 7-d-old D. magna were added to each of the 3 replicates of each concentration treatment, while no D. magna was added in the remaining 3 replicates used as the control. The f-nC60 concentrations were analyzed every 2 h over a period of 8 h. The test medium was not renewed and animals were not fed during the experimental period. The 48-h acute toxicity tests were carried out in SM7, with 4 PVP concentrations (0.005 g/L, 0.05 g/L, 1 g/L, and 5 g/L) or 3 f-
Environ Toxicol Chem 33, 2014
1123
nC60 concentrations (0.5 mg/L, 1 mg/L, and 4 mg/L, dispersed in 1 g PVP/L) using 7-d-old D. magna. Only PVP toxicity and PVP dispersed f-nC60 toxicity were conducted in the present study. Each test consisted of a control (only SM7 medium) and different concentration treatments for PVP and PVP-dispersed fnC60; each treatment had 4 replicates. In each beaker, 100 mL of test medium was allocated with 10 daphnids, which had been cultured in M7 medium for 7 d from birth. No food was provided during the exposure period. After 48 h of exposure, the daphnids that did not resume swimming on gentle agitation were considered to be immobilized and were enumerated. The 48-h median effective concentration and 95% confidence intervals were calculated according to the trimmed Spearman-Karber method based on nominal PVP and f-nC60 concentrations. Metal adsorption onto functionalized fullerenes
In the suspension, metals either exist as free form or are bound with f-nC60; thus, quantifying the fraction of Cd and Zn bound with functionalized fullerenes was essential. The adsorption experiment was therefore carried out at 2 initial concentrations of f-nC60 (1 mg/L and 4 mg/L) by adding Cd (0.5 mg/L as CdCl2 and 109Cd as tracer) or Zn (1 mg/L as ZnCl2 and 65Zn as tracer) to the suspension. The suspension was adjusted with 0.01 M HCl or NaOH to achieve a pH of 8.0 0.2. The adsorption kinetics of 109Cd and 65Zn onto f-nC60 was then measured over 24 h at 20 1 8C. At different time intervals, a 10-mL aliquot was collected and ultracentrifuged using a 1-kDa unltrafilter (EMD Millipore) [22], and the radioactivity of the suspension before and after filtration was measured. The amounts of metals adsorbed onto f-nC60 were calculated as the difference between the initial and the final radioactivities in suspension. The reproducibility of the procedure was checked by repeating experiments up to 3 times. Uptake of metals bound to functionalized fullerenes
We quantified the uptake of metals bound to functionalized fullerenes at 2 concentrations (1 mg- and 4 mg f-nC60/L). The fnC60 was suspended in the PVP-SM7 (1 mg/mL) medium with addition of 0.5 mg/L for Cd and 1 mg/L for Zn. Radioisotopes 109 Cd and 65Zn were also spiked as the radiotracers to follow the uptake behavior of Cd and Zn by the animals, as described earlier [18]. Each concentration treatment had 6 beakers. Based on the adsorption kinetic experiment, the final retained free metal ions in the suspension achieved equilibrium after 24 h. Three beakers were used as prepared, and the suspension from the other 3 beakers was poured into a stirred cell with 1-kDa membrane under 75 psi pressure of N2 to remove the f-nC60. The filtrates with undetectable f-nC60 were also used for the 8-h dissolved uptake experiment. Thus, the first 3 replicates were used to quantify the uptake from the suspension (including metal bound to f-nC60 and free metal ions), and the filtrates were used to quantify the uptake from free metal ions in the solution. Before the exposure, selected daphnids with uniform size were depurated in artificial freshwater for 2 h to clear their guts. Thirty daphnids for each treatment were equally divided into 3 replicates and placed into beakers containing 100 mL medium for 8 h. No food was added during the experimental period in order to avoid the uptake of metals associated with food. At 2 h, 4 h, 6 h, and 8 h, daphnids were collected and rinsed with SM7 for 1 min to remove the weakly adsorbed metals. An aliquot of water was also sampled to measure the radioactivity in the aqueous phase. After the radioactivity measurements, the animals were returned to the medium immediately. After 8 h, the animals were dried at 80 8C to measure their dry weights. A
Environ Toxicol Chem 33, 2014
We specifically quantified the depuration of metals bound with f-nC60 with and without the presence of algae following short-term (15 min) and long-term (24 h) exposure. Assimilation efficiencies of Cd and Zn associated with algae were also quantified as the control. To radiolabel the f-nC60, 109Cd and 65 Zn (plus stable Cd and Zn to achieve the intended concentrations) were added to the f-nC60 suspension at different concentrations and equilibrated for 12 h. The algae C. reinhardtii was radiolabeled by 109Cd and 65Zn using the methods described in Tan and Wang [17]. Briefly, C. reinhardtii at the early log phase were harvested and resuspended in a modified WC medium (without EDTA, Zn, and Cu) at an initial cell density of 2 105 cells/mL. Radioisotopes were added at 148 kBq/L of 109Cd and 296 kBq/L of 65Zn. After 3 d of growth, the algae were centrifuged and resuspended in M7 medium immediately before the experiment. The algal cell density was determined using a hemocytometer. In the pulse feeding experiments, there were in total 3 treatments (and 3 replicates for each treatment), including 1) radiolabeled algae, 2) radiolabeled f-nC60 at 1 mg/L (with an initial Cd and Zn concentration of 1 mg/L and 2 mg/L, respectively), and 3) radiolabeled f-nC60 at 4 mg/L (with an initial Cd and Zn concentration of 1 mg/L and 2 mg/L, respectively). The SM7 medium was amended with PVP. After pulse exposure for 15 min, the daphnids were depurated in SM7 medium with or without algae at a density of 105 cells/mL for 24 h. The daphnids were assayed for radioactivity every 3 h during the first 12 h and every 12 h during the subsequent 12 h. The water and food were renewed after each radioactivity measurement. The assimilation efficiency of metals was calculated as the percentage of radioactivity retained in the daphnids after 24 h of depuration. To measure the efflux rate constant of metals, the daphnids were exposed to radiolabeled f-nC60 at 0.5 mg/L and 1 mg/L, with initial Cd and Zn concentrations of 0.5 mg/L, in SM7 medium amended with PVP for 24 h. The daphnids were then removed from the exposed medium, radioassayed, and then depurated in SM7 medium with food. Again, the animals were radioassayed regularly during the 24-h depuration period. One-way analysis of variance was performed to compare metal adsorption by functionalized carbon nanoparticles, dissolved uptake, and depuration rates of Cd and Zn. RESULTS AND DISCUSSION
Stability of functionalized C60 and 48-h acute toxicity of D. magna
Time-dependent zeta-potential and ultraviolet-visible absorption tests were first conducted to confirm the stability of the f-nC60 suspension. The stability of PVP-facilitated f-nC60 suspensions in SM7 with or without daphnids is shown in Figure 1. The calibration curves for the intensity of absorbance as a function of f-nC60 concentration were first established for the suspension. It was clear that f-nC60 remained highly stable when they were dispersed by PVP. The results of zeta-potentials of f-nC60 indicated that f-nC60 was rich in carboxyl and resulted
Zeta-potential (mV)
Depuration of metals bound with functionalized fullerenes
A
-20 -25
No daphnids With daphnids
-30 -35 -40 -45
B f-nC60 (mg/L)
linear regression was conducted between the accumulated Cd and Zn in daphnids and the exposure time to calculate the influx rate (I, mg/g/h). Since the exposed concentrations of metals in the medium were different among the different treatments, we also calculated the dissolved uptake rate constant (ku) as the influx rate divided by the ambient concentration of metals, which can be used as the first-order parameter for comparison among the different treatments.
Z.-G. Yu and W.-X. Wang
C % f-nC60
1124
0
10
20
30
40
50
4 3
0.5 mg/L 1 mg/L 4 mg/L
2 1 0 100 80 60
0.5 mg/L 1 mg/L 4 mg/L
40 20 0
5
10 15 Time (h)
20
25
Figure 1. (A) Zeta-potentials of dispersed f-nC60 in suspension with and without the presence of daphnids. (B) Concentrations of f-nC60 remaining in suspension without daphnids. (C) Percentage f-nC60 remaining in suspension with the presence of daphnids.
in higher electrostatic repulsion. The stability of f-nC60 thus ensured their dispersability in the medium for a relatively long period (24 h). Functionalized nC60 suspension was partially destabilized by the presence of daphnids during exposure. After 24 h, the concentrations of f-nC60 dispersed in the test medium were 74% to 86% of the initial concentrations, and the zeta-potential also changed over time accordingly (Figure 1). With prolonged exposure, f-nC60 agglomeration was present at the bottom of the test solution. Therefore, D. magna had an effect on the stability of polyvinylpyrrolidone- f-nC60, similar to previous findings by Robert et al. [23] and Yu and Wang [18]. In our earlier study [18], only 2% to 8% of the initial concentration of dispersed functionalized carbon nanotubes was lost during the 8h exposure to daphnids. One possibility for such destabilization was that the animals ingested the dispersant nanoparticles as a result of their filter-feeding. However, such a decrease of the f-
Interaction of fullerenes and metals in Daphnia magna
Environ Toxicol Chem 33, 2014
Adsorption of metals on functionalized nC60 and metal accumulation
The oxygen-containing carboxyl functionalized groups had great capacity to sorb metals from water because of their high specific surface areas and functionalized group densities. The percentage of metals absorbed by the f-nC60 at 2 different concentrations (1 mg/L and 4 mg/L) is shown in Figure 2. Sorption equilibrium was rapidly reached (within 4 h), and 87% to 92% of Cd and 82% to 84% of Zn were adsorbed onto fullerenes at the 2 f-nC60 concentrations. Thus, most metals were bound to f-nC60, with only a small fraction remaining in the dissolved phase in suspension. The concentrations of f-nC60 had little influence on the proportion of metals sorbed onto f-nC60. The partitioning coefficient (kd) can therefore be calculated as the amount of metals sorbed onto the nanomaterials divided by the amount of metals remaining in solution. Accordingly, the kd
A
100 75
Adsorbed onto f-nC60 (%)
50 Cd Zn
25 0 100
values of Cd and Zn were 2.88 106 and 1.31 106 L/kg, respectively. These calculated kd values were much higher than those measured for functionalized carbon nanotubes [18]. In our earlier study, the calculated kd values of Cd and Zn for functionalized single-walled nanotubes were 3.75 105 L/kg and 2.50 105 L/kg, while for functionalized multiwalled nanotubes values were 8.30 104 and 6.25 104, respectively. Generally, the adsorption capacity of different carbon-based nanomaterials was higher for Cd than for Zn, and smaller particle size could lead to higher adsorption efficiency for metals. Despite only a small fraction of metals (
INTERACTION OF FUNCTIONALIZED FULLERENES AND METAL ACCUMULATION IN DAPHNIA MAGNA ZHI-GUO YU and WEN-XIONG WANG* Division of Life Science, The Hong Kong University of Science and Technology, Kowloon, Hong Kong, China (Submitted 8 October 2013; Returned for Revision 12 December 2013; Accepted 10 January 2014) Abstract: In aquatic environments, transformation of pollutants by association with functionalized carbon-based nanoparticles can dramatically change their cycling pathways. The present study quantified the uptake and depuration behavior of cadmium and zinc bound with functionalized fullerene nanoparticles (f-nC60) in a freshwater cladoceran, Daphnia magna, in a well-dispersed medium. Metal uptake proceeded with a linear pattern during the 8-h exposure period, and the uptake rate constants (ku) were 1.3-fold to 1.4-fold higher for Cd or comparable for Zn bound with f-nC60 than those of the free ones. The assimilation efficiencies of Cd and Zn bound with f-nC60 were significantly enhanced when compared with those metals bound with algal food. Furthermore, the depuration of metals bound with f-nC60 was relatively slower compared to the depuration of metals bound with carbon nanotubes. A longer exposure to f-nC60 resulted in an even slower depuration of metals. The authors conclude that metal binding with f-nC60 as modified nanoparticles could serve as a new pathway for the elevated metal accumulation in Daphnia. Environ Toxicol Chem 2014;33:1122–1128. # 2014 SETAC Keywords: Daphnia
Cadmium
Zinc
Functionalized fullerene
Biokinetics
relatively slow depuration of fullerenes in a freshwater cladoceran, Daphnia magna, likely make these animals carriers of fullerenes to higher trophic levels. Daphnia magna is a filterfeeder and has been a model organism in ecotoxicological studies. Cadmium and Zn are the common toxic metals found in freshwater ecosystems [16,17]. Therefore, the objective of the present study was to examine how functionalized fullerenes (particle size 100 nm) affect the uptake and depuration of Cd and Zn in a freshwater cladoceran, D. magna. Previously, we have characterized the influences of nonfunctionalized carbon nanotubes on the uptake of metals from the dissolved and dietary phases and demonstrated that physical blocking played a role in metal accumulation in D. magna [18]. However, no direct interaction between toxic metals and nonfunctionalized carbon nanotubes was found in this earlier study. In the present study, we investigated the functionalized fullerenes that can bind with metals directly by acid adsorption. Compared with carbon nanotubes, fullerenes are much smaller in size and may have a greater potential to bind with metals because of their high specific surface areas. We conducted the dissolved uptake and depuration experiments of Cd and Zn in interacting with functionalized fullerene nanoparticles (f-nC60). The study will provide important information for our current understanding of the interaction between carbon-based nanoparticles and the biological fate and transport of toxic metals in aquatic organisms.
INTRODUCTION
Among the manufactured carbon-based nanomaterials, fullerenes were among the first detected and bulk-produced in nanotechnology [1]. Subsequent studies reported their interactions with biological targets [2] and raised a wide interest in the behavior of fullerenes in water. Functionalized fullerenes are now rapidly applied in clinical and material usage, such as personal care products, coatings, and drug delivery vehicles [1]. As a result of application of nanotechnological products, the risk of their release into the environment (e.g., waste disposal or accidental release) during or after their use has increased. These nanomaterials may finally enter into ecologically sensitive environments such as aquatic and terrestrial ecosystems [3]. In aquatic environments, agglomeration of fullerene nanoparticles (nC60) can significantly influence their environmental fate, transport, and toxicity [4]. They will interact with and affect the behavior of other pollutants such as organics and metals [5]. Nevertheless, the bioaccumulation of toxic metals interacting with fullerene (and derivative C60) has been rarely investigated. Because fullerenes are entirely insoluble in water, there is a major technical difficulty in the preparation of medium with C60 in aquatic systems [2]. Adding functional groups to fullerene C60 molecules (e.g., –COOH) could increase their hydrophilicity [1]. These functionalized C60 molecules can form negatively charged agglomerates of several C60 molecules (namely, nC60), which are surrounded by water molecules [6,7]. Using water mixing and vigorous stirring in the laboratory, agglomeration allows fullerenes to remain suspended for weeks or months [8]; this mobility likely makes them more available for aquatic organisms. Many studies have investigated the bioaccumulation and toxicological effects of fullerenes (C60) on various organisms [9–14]. Tervonen et al. [15] showed that the rapid uptake and
MATERIALS AND METHODS
Preparation of aqueous f-nC60 suspension
Functionalized fullerenes (1,2-methanofullerene C60-61carboxylic acid, C62H2O2, molecular weight ¼ 778.68 g/mol, as f-C60) was purchased from Sigma-Aldrich and used as received. The diameter of the nanomaterial was not given by the manufacturer. According to the published data [19], the van der Waals diameter of a C60 fullerene molecule is approximately 1.1 nm. However, the average particle size for f-nC60 is
* Address correspondence to [email protected]. Published online 16 January 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/etc.2520 1122
Interaction of fullerenes and metals in Daphnia magna
120 10 nm as measured by dynamic light scattering (Brookhaven Instruments). The calculated carboxyl group content was 5.65% of weight. Stock suspension of f-nC60 was made by mixing 2 mg of fnC60 with 1 g/L polyvinylpyrrolidone (PVP) solution in a 40-mL glass tube, followed by sonication using a probe sonicator at an intensity of 60 W for 30 min [18,20] to disperse the nanoparticles. The suspension did not undergo a filtration process and was collected for further use; the total carbon concentration of the stock suspension was analyzed by a total organic carbon analyzer (minimum detection limit of 0.05 mg/L). The concentration of f-nC60 in the test suspension was analyzed by a spectrophotometer. A calibration curve was established by recording 6 different dilutions of 10 mg f-nC60/L (PVP suspension) at 288 nm (r2 ¼ 0.99). The calibration curve was used to estimate the concentration of f-nC60 in suspensions in the following experiments. It was difficult to reach a full extent of exfoliation by sonication; thus, the f-nC60 concentrations calculated from the calibration curves were estimated values. The average surface charge of functionalized fullerenes was characterized by a phase analysis light scattering technique. Organisms, medium, and radioisotopes
The test organism, D. magna, was cultured in filtered (GF/C Whatman) pond water collected from the Hong Kong University of Science and Technology campus at a density of 1 individual/ 10 mL of water. Before the experiments, the green algae Chlamydomonas reinhardtii raised previously in Woods Hole modified CHU 10 (WC) medium [17] were fed to daphnids at 5 104 cells/mL (neonates 3 d old) or 105 cells/mL (adults >3 d old) daily. The growth conditions for both green algae and daphnids were 23.5 8C with a 14 h to 10 h light to dark cycle. Individual culture systems were maintained for healthy neonates, and then 7-d-old organisms were used for the following bioassays. In all treatments, modified freshwater (SM7) with a low calcium concentration (20 mg/L) [17,18,21] was used. The low hardness medium helped to minimize f-nC60 precipitation. The pH of solutions was adjusted to 8.0 0.2 by spiking 0.01 M HCl or NaOH. Radioisotopes 109CdCl2 (Boston, MA, USA) and 65ZnCl2 (Roskilde, Denmark) were used as tracers in the present study. Radioisotopes were diluted in 0.1 N HCl as stock solution. The radioactivity was measured by a Wallac 1480 NaI(T1) gamma detector at 88 keV for 109Cd and 1115 keV for 65Zn (Turku, Finland). Counting propagated accuracy was greater than 95%. Daphnia effects on stability of dispersed f-nC60 and acute toxicity
In a biokinetic study, a uniform suspension with relative stability was required to provide ideal conditions for the experiment. Thus, the first experiment tested the effects of daphnids on the stability of f-nC60 suspension at different concentrations by monitoring the concentration and zetapotential of f-nC60 over time. Specifically, 200-mL breakers were filled with 100 mL of test solution (PVP dispersed f-nC60 in modified SM7) at different concentrations of f-nC60 (0.5 mg/L, 1 mg/L, and 4 mg/L). Six replicates were prepared for each concentration treatment. Ten 7-d-old D. magna were added to each of the 3 replicates of each concentration treatment, while no D. magna was added in the remaining 3 replicates used as the control. The f-nC60 concentrations were analyzed every 2 h over a period of 8 h. The test medium was not renewed and animals were not fed during the experimental period. The 48-h acute toxicity tests were carried out in SM7, with 4 PVP concentrations (0.005 g/L, 0.05 g/L, 1 g/L, and 5 g/L) or 3 f-
Environ Toxicol Chem 33, 2014
1123
nC60 concentrations (0.5 mg/L, 1 mg/L, and 4 mg/L, dispersed in 1 g PVP/L) using 7-d-old D. magna. Only PVP toxicity and PVP dispersed f-nC60 toxicity were conducted in the present study. Each test consisted of a control (only SM7 medium) and different concentration treatments for PVP and PVP-dispersed fnC60; each treatment had 4 replicates. In each beaker, 100 mL of test medium was allocated with 10 daphnids, which had been cultured in M7 medium for 7 d from birth. No food was provided during the exposure period. After 48 h of exposure, the daphnids that did not resume swimming on gentle agitation were considered to be immobilized and were enumerated. The 48-h median effective concentration and 95% confidence intervals were calculated according to the trimmed Spearman-Karber method based on nominal PVP and f-nC60 concentrations. Metal adsorption onto functionalized fullerenes
In the suspension, metals either exist as free form or are bound with f-nC60; thus, quantifying the fraction of Cd and Zn bound with functionalized fullerenes was essential. The adsorption experiment was therefore carried out at 2 initial concentrations of f-nC60 (1 mg/L and 4 mg/L) by adding Cd (0.5 mg/L as CdCl2 and 109Cd as tracer) or Zn (1 mg/L as ZnCl2 and 65Zn as tracer) to the suspension. The suspension was adjusted with 0.01 M HCl or NaOH to achieve a pH of 8.0 0.2. The adsorption kinetics of 109Cd and 65Zn onto f-nC60 was then measured over 24 h at 20 1 8C. At different time intervals, a 10-mL aliquot was collected and ultracentrifuged using a 1-kDa unltrafilter (EMD Millipore) [22], and the radioactivity of the suspension before and after filtration was measured. The amounts of metals adsorbed onto f-nC60 were calculated as the difference between the initial and the final radioactivities in suspension. The reproducibility of the procedure was checked by repeating experiments up to 3 times. Uptake of metals bound to functionalized fullerenes
We quantified the uptake of metals bound to functionalized fullerenes at 2 concentrations (1 mg- and 4 mg f-nC60/L). The fnC60 was suspended in the PVP-SM7 (1 mg/mL) medium with addition of 0.5 mg/L for Cd and 1 mg/L for Zn. Radioisotopes 109 Cd and 65Zn were also spiked as the radiotracers to follow the uptake behavior of Cd and Zn by the animals, as described earlier [18]. Each concentration treatment had 6 beakers. Based on the adsorption kinetic experiment, the final retained free metal ions in the suspension achieved equilibrium after 24 h. Three beakers were used as prepared, and the suspension from the other 3 beakers was poured into a stirred cell with 1-kDa membrane under 75 psi pressure of N2 to remove the f-nC60. The filtrates with undetectable f-nC60 were also used for the 8-h dissolved uptake experiment. Thus, the first 3 replicates were used to quantify the uptake from the suspension (including metal bound to f-nC60 and free metal ions), and the filtrates were used to quantify the uptake from free metal ions in the solution. Before the exposure, selected daphnids with uniform size were depurated in artificial freshwater for 2 h to clear their guts. Thirty daphnids for each treatment were equally divided into 3 replicates and placed into beakers containing 100 mL medium for 8 h. No food was added during the experimental period in order to avoid the uptake of metals associated with food. At 2 h, 4 h, 6 h, and 8 h, daphnids were collected and rinsed with SM7 for 1 min to remove the weakly adsorbed metals. An aliquot of water was also sampled to measure the radioactivity in the aqueous phase. After the radioactivity measurements, the animals were returned to the medium immediately. After 8 h, the animals were dried at 80 8C to measure their dry weights. A
Environ Toxicol Chem 33, 2014
We specifically quantified the depuration of metals bound with f-nC60 with and without the presence of algae following short-term (15 min) and long-term (24 h) exposure. Assimilation efficiencies of Cd and Zn associated with algae were also quantified as the control. To radiolabel the f-nC60, 109Cd and 65 Zn (plus stable Cd and Zn to achieve the intended concentrations) were added to the f-nC60 suspension at different concentrations and equilibrated for 12 h. The algae C. reinhardtii was radiolabeled by 109Cd and 65Zn using the methods described in Tan and Wang [17]. Briefly, C. reinhardtii at the early log phase were harvested and resuspended in a modified WC medium (without EDTA, Zn, and Cu) at an initial cell density of 2 105 cells/mL. Radioisotopes were added at 148 kBq/L of 109Cd and 296 kBq/L of 65Zn. After 3 d of growth, the algae were centrifuged and resuspended in M7 medium immediately before the experiment. The algal cell density was determined using a hemocytometer. In the pulse feeding experiments, there were in total 3 treatments (and 3 replicates for each treatment), including 1) radiolabeled algae, 2) radiolabeled f-nC60 at 1 mg/L (with an initial Cd and Zn concentration of 1 mg/L and 2 mg/L, respectively), and 3) radiolabeled f-nC60 at 4 mg/L (with an initial Cd and Zn concentration of 1 mg/L and 2 mg/L, respectively). The SM7 medium was amended with PVP. After pulse exposure for 15 min, the daphnids were depurated in SM7 medium with or without algae at a density of 105 cells/mL for 24 h. The daphnids were assayed for radioactivity every 3 h during the first 12 h and every 12 h during the subsequent 12 h. The water and food were renewed after each radioactivity measurement. The assimilation efficiency of metals was calculated as the percentage of radioactivity retained in the daphnids after 24 h of depuration. To measure the efflux rate constant of metals, the daphnids were exposed to radiolabeled f-nC60 at 0.5 mg/L and 1 mg/L, with initial Cd and Zn concentrations of 0.5 mg/L, in SM7 medium amended with PVP for 24 h. The daphnids were then removed from the exposed medium, radioassayed, and then depurated in SM7 medium with food. Again, the animals were radioassayed regularly during the 24-h depuration period. One-way analysis of variance was performed to compare metal adsorption by functionalized carbon nanoparticles, dissolved uptake, and depuration rates of Cd and Zn. RESULTS AND DISCUSSION
Stability of functionalized C60 and 48-h acute toxicity of D. magna
Time-dependent zeta-potential and ultraviolet-visible absorption tests were first conducted to confirm the stability of the f-nC60 suspension. The stability of PVP-facilitated f-nC60 suspensions in SM7 with or without daphnids is shown in Figure 1. The calibration curves for the intensity of absorbance as a function of f-nC60 concentration were first established for the suspension. It was clear that f-nC60 remained highly stable when they were dispersed by PVP. The results of zeta-potentials of f-nC60 indicated that f-nC60 was rich in carboxyl and resulted
Zeta-potential (mV)
Depuration of metals bound with functionalized fullerenes
A
-20 -25
No daphnids With daphnids
-30 -35 -40 -45
B f-nC60 (mg/L)
linear regression was conducted between the accumulated Cd and Zn in daphnids and the exposure time to calculate the influx rate (I, mg/g/h). Since the exposed concentrations of metals in the medium were different among the different treatments, we also calculated the dissolved uptake rate constant (ku) as the influx rate divided by the ambient concentration of metals, which can be used as the first-order parameter for comparison among the different treatments.
Z.-G. Yu and W.-X. Wang
C % f-nC60
1124
0
10
20
30
40
50
4 3
0.5 mg/L 1 mg/L 4 mg/L
2 1 0 100 80 60
0.5 mg/L 1 mg/L 4 mg/L
40 20 0
5
10 15 Time (h)
20
25
Figure 1. (A) Zeta-potentials of dispersed f-nC60 in suspension with and without the presence of daphnids. (B) Concentrations of f-nC60 remaining in suspension without daphnids. (C) Percentage f-nC60 remaining in suspension with the presence of daphnids.
in higher electrostatic repulsion. The stability of f-nC60 thus ensured their dispersability in the medium for a relatively long period (24 h). Functionalized nC60 suspension was partially destabilized by the presence of daphnids during exposure. After 24 h, the concentrations of f-nC60 dispersed in the test medium were 74% to 86% of the initial concentrations, and the zeta-potential also changed over time accordingly (Figure 1). With prolonged exposure, f-nC60 agglomeration was present at the bottom of the test solution. Therefore, D. magna had an effect on the stability of polyvinylpyrrolidone- f-nC60, similar to previous findings by Robert et al. [23] and Yu and Wang [18]. In our earlier study [18], only 2% to 8% of the initial concentration of dispersed functionalized carbon nanotubes was lost during the 8h exposure to daphnids. One possibility for such destabilization was that the animals ingested the dispersant nanoparticles as a result of their filter-feeding. However, such a decrease of the f-
Interaction of fullerenes and metals in Daphnia magna
Environ Toxicol Chem 33, 2014
Adsorption of metals on functionalized nC60 and metal accumulation
The oxygen-containing carboxyl functionalized groups had great capacity to sorb metals from water because of their high specific surface areas and functionalized group densities. The percentage of metals absorbed by the f-nC60 at 2 different concentrations (1 mg/L and 4 mg/L) is shown in Figure 2. Sorption equilibrium was rapidly reached (within 4 h), and 87% to 92% of Cd and 82% to 84% of Zn were adsorbed onto fullerenes at the 2 f-nC60 concentrations. Thus, most metals were bound to f-nC60, with only a small fraction remaining in the dissolved phase in suspension. The concentrations of f-nC60 had little influence on the proportion of metals sorbed onto f-nC60. The partitioning coefficient (kd) can therefore be calculated as the amount of metals sorbed onto the nanomaterials divided by the amount of metals remaining in solution. Accordingly, the kd
A
100 75
Adsorbed onto f-nC60 (%)
50 Cd Zn
25 0 100
values of Cd and Zn were 2.88 106 and 1.31 106 L/kg, respectively. These calculated kd values were much higher than those measured for functionalized carbon nanotubes [18]. In our earlier study, the calculated kd values of Cd and Zn for functionalized single-walled nanotubes were 3.75 105 L/kg and 2.50 105 L/kg, while for functionalized multiwalled nanotubes values were 8.30 104 and 6.25 104, respectively. Generally, the adsorption capacity of different carbon-based nanomaterials was higher for Cd than for Zn, and smaller particle size could lead to higher adsorption efficiency for metals. Despite only a small fraction of metals (
