Environ. Sci. Technol. 2002, 36, 1477-1484
Chemical Speciation and Toxicity of Nickel Species in Natural Waters from the Sudbury Area (Canada) RUPASRI MANDAL, NOURI M. HASSAN, JOHN MURIMBOH, CHUNI L. CHAKRABARTI,* AND MARGARET H. BACK Ottawa-Carleton Chemistry Institute, Department of Chemistry, Carleton University, 1125 Colonel By Drive, Ottawa, Ontario K1S 5B6, Canada UCU RAHAYU AND DAVID R. S. LEAN Department of Biology, University of Ottawa, Ottawa, Ontario K1N 6N5, Canada
Metal complexation properties of dissolved organic carbon (DOC) in freshwaters are recognized but poorly understood. Here, we investigated the release of free nickel from Ni-DOC complexes using nickel-polluted freshwaters from Sudbury (Canada). We used the Competing Ligand Exchange Method with Chelex-100 as the competing ligand to measure the rate of free Ni2+ ion released by the dissociation of Ni-DOC complexes. The kinetic studies showed that the fastest kinetically distinguishable component representing ∼30-95% of the total nickel had a dissociation rate coefficient similar to that reported for [Ni(H2O)6]2+. High concentrations of Ca2+ and Mg2+ caused a larger amount of the DOC-bound nickel to be released as free Ni2+ ion. Growth inhibition of the freshwater alga Pseudokirchneriella subcapitata was highly correlated with the Ni/DOC ratio, the free plus labile nickel concentration, and the dissociation rate coefficient. While the levels of metals were not sufficient to kill Daphnia magna, these test organisms were immobilized in the same samples that showed algal growth inhibition. Only one sample caused 22% death of Hydra attenuata. The algal toxicity tests were consistent with the kinetic speciation results and are consistent with the hypothesis that dissolved [Ni(H2O)6]2+ plus other labile nickel species are toxic forms of Ni present.
Introduction Trace metal complexation by dissolved organic carbon (DOC) plays an important role in solubilizing metals and in regulating transport, bioavailability, and toxicity of metals in freshwaters. Since most of these metals in freshwaters are DOC-bound, their study is of primary importance in understanding the behavior and fate of these metals in freshwaters (1, 2). Freshwater systems are often far removed from equilibrium and, hence, require kinetic approaches to determine the chemical speciation of metals (2). Kinetic studies of metal speciation not only can differentiate chemical species according to their kinetic parameters but also can give * Corresponding author telephone: (613)520-2600, ext 3839; fax: (613)520-3749/3830; e-mail:
[email protected]. 10.1021/es015622e CCC: $22.00 Published on Web 02/22/2002
2002 American Chemical Society
information on the distribution of the chemical species in the system at any time during the kinetic process. The results obtained can therefore be used to estimate the bioavailability of metal if a kinetic model can be constructed to represent the process of biological uptake. For nonequilibrium systems, kinetic speciation would give a more realistic description of bioavailability of trace metals than metal speciation based on local equilibrium approximations. The kinetics of metal complex dissociation in freshwaters can be modeled employing the Competing Ligand Exchange Method (CLEM) using Chelex-100 as the competing ligand. Chelex was used by Figura and McDuffie (3, 4) to develop a method for differentiating trace metal species on the basis of relative lability, classifying them by operationally defined terms: “very labile”, “moderately labile”, “ slowly labile”, or “inert” based on the characteristic time scale of the measuring technique (anodic stripping voltammetry). However, the method of Figura and McDuffie has the limitations of its operationally defined classification. Nickel is an essential micronutrient for some marine algae, but micromolar Ni2+ free ion, i.e., [Ni(H2O)6]2+, has been reported to be toxic to a variety of algae, invertebrates, and fish (5). However, interactions of Ni(II) with dissolved organic matter have not been much studied, perhaps because of the absence of convenient, sensitive methods for measuring free Ni2+ ions in the presence of organic complexants. A recent study (6) of nickel speciation and complexation kinetics in freshwaters by ligand exchange and differential pulse cathodic stripping voltammetry has found that more than 99.9% of dissolved nickel is bound by organic ligands with strong affinity (log K ) 12.1-14.9). Langford and Gutzman (7) have used CLEM for kinetic speciation of nickel with the chromophore 4-(2-pyridylazo)resorcinol (PAR) as a competing ligand and spectrophotometric detection of the Ni-PAR complex. However, most of the reported investigations of nickel were done on unpolluted freshwaters, and very little is known about the chemical speciation of nickel in metalpolluted freshwaters. Polluted freshwaters from Ni mining and refining areas typically contain Ni along with other 3d transition metals and strong ligands (anthropogenic and natural) along with naturally occurring humic substances and large concentrations of Ca and Mg. The above multimetal and multiligand mixtures set the stage for double-exchange reactions (metalmetal exchange and ligand-ligand exchange) to occur, involving reestablishment of coordination equilibrium (8, 9), which is sometimes slow. In a ground-breaking paper published recently, Xue et al. (6) highlighted the significance of slow kinetics of Ni complexation in freshwaters and its consequencessthe unpredictable concentration levels of free nickel ions, which have been reported to be toxic to freshwater biota (5). Information on the competition between nickel and major cations for naturally occurring organic complexants in freshwaters is scarce. Recently, we have studied the competition of Ca(II) and Mg(II) with Ni(II) for binding sites of a well-characterized fulvic acid in model systems at constant pH and ionic strength (10). We have reported that, in model solutions containing humic substances and trace quantities of nickel and a large excess of major cations (Ca2+ and Mg2+), the competition of Ca2+ and Mg2+ with Ni2+ for binding sites on humic substances results in weak Ni(II)humate complexes that are labile, releasing free Ni2+ ions. The effect of major cations on the binding of trace metals by humic substances has been studied mainly for model systems (11-16). There are, however, no published reports on the linkage between nickel speciation and toxicity of the nickel VOL. 36, NO. 7, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1477
TABLE 1. pH, Conductivity, and DOC of the Sudbury Area Water Samples conductivity (µS/cm)
pH
DOC (mg/L)
sample
sampling sites
1998
1999
1998
1999
1998
1999
1 2 3 4 5 6 7 8 9 10
Copper Cliff, wastewater treatment outflow Junction Creek upstream of Copper Cliff Kelly Lake Alice Lake conservation area Coniston Creek downstream Moose Lake treated discharge Moose Creek upstream Onaping River downstream Onaping River downstream of sewage treatment plant
6.4 a 6.6 a 6.9 6.8 6.9 6.6 6.8 6.9
6.2 7.8 7.6 7.2 7.6 8.1 7.9 7.6 a 7.2
1980 a 1250 a 898 528 1520 1060 93.2 64.1
2920 800 1345 200 831 420 1671 1640 a 221
7.09 a 6.45 a 6.44 6.03 6.47 6.40 6.45 6.51
7.09 6.91 6.84 6.03 6.47 6.98 6.39 3.51 a 6.39
a
Sample not collected.
species in Ni-polluted freshwater environments. Here, we present the first published report relating Ni speciation in nickel-polluted freshwaters to its toxicity. The objectives of this research are (i) to determine the release of free nickel from the Ni-DOC complexes in nickelpolluted freshwaters and (ii) to establish the linkage between nickel speciation and toxicity.
Kinetic Model CLEM. The kinetic model proposed by Olson and Shuman (17) was modified (18) to provide the dissociation kinetics of complex ML, where M is a metal ion and L is a ligand such as DOC. Consider an aqueous mixture of n components in which each component, designated MLi, undergoes a firstorder or pseudo-first-order reaction independently and simultaneously with all other components: k1
MLi {\ } M + Li k -1
(slow)
(1)
where k1 and k-1 are the rate coefficients for the forward and the reverse reactions, respectively. In CLEM, with Chelex as the competing ligand, the Chelex reacts with M as follows: k2
M + Chelex {\ } M-Chelex k -1
(fast)
(2)
The charges have been omitted for simplicity. In the presence of a large excess of Chelex, the overall reaction 3 is irreversible:
ML + Chelex f M-Chelex + L
(3)
The model assumes that (a) reaction 2 is much faster than reaction 1, (b) M-Chelex complex is thermodynamically much stronger than ML complex, (c) CChelex . CM, and (d) Chelex does not bind with the ML. The total concentration of all dissociating components, CML, decreases exponentially with time: n
CML(t) )
∑C i)1
0 MLi
exp(-kit)
(4)
0 where CML is the initial concentration of MLi, the ith i component. The concentration of the nickel complexes remaining in the test solution, CML(t), is measured as a function of time. The time taken for completion of reaction 2, i.e., its rate, will establish the theoretical upper limit of the dissociation reaction that can be observed with this technique.
Experimental Section Materials. Chelex-100 resin (Bio-Rad, 100-200 mesh, sodium form) was soaked in ultrapure water, stored, and then filtered 1478
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 7, 2002
just before use. Ultrapure water of resistivity 18.2 MΩ‚cm was obtained direct from a Milli-Q-Plus water purification system (Millipore Corporation). Dissolved Organic Carbon. The DOC of the filtered water samples was measured using O I Analytical model 1010 TOC analyzer. Apparatus. Screw-capped Teflon cylindrical bottles (500 mL capacity) were used to fabricate the reactors for the kinetic studies of the samples using graphite furnace atomic absorption spectrometry (GFAAS) (18), for which a PerkinElmer Zeeman atomic absorption spectrophotometer, model 4100ZL, equipped with an AS-70 autosampler and transversely heated graphite tubes (Perkin-Elmer), were used. The instrumental parameters used in the GFAAS determination of nickel were as follows: analysis line, 232.0 nm; ashing temperature, 1000 °C; and atomization temperature, 2300 °C. An Accumet 20 pH/mV/conductivity meter (Fisher Scientific) was used for the pH and conductivity determinations. Cleaning Procedure. All containers were made of Teflon and were precleaned by soaking in 10% HNO3 (AR grade) for 1 week at room temperature, followed by a 5× rinse with ultrapure water. Finally, they were kept filled with ultrapure water until used. Sampling Sites and Sampling Protocol. Water samples (6.6 L) were collected in October 1998 and in June 1999 from several sampling sites in the Sudbury (Canada) area using precleaned Teflon bottles. The sampling sites were chosen applying the only criterion that the sites should provide nickel-polluted freshwaters, which were required for studying the applicability of the nickel speciation techniques developed in our laboratories using model solutions (10). The sites were from upstream and downstream of the nickel tailing ponds and outflows from the treatment tanks receiving effluents from nickel ore processing and smelting operations. A list of the sampling sites, pH values, conductivities, and DOC values of all the samples is presented in Table 1. The sample temperatures were 24 ( 1 °C in 1998 and 24 ( 0.5 °C in 1999. The samples were filtered through 0.45-µm membrane filters (CN-6 Metricel, 47 mm diameter, Gelman, Ann Arbor, MI) in our laboratory within 24 h of the sample collection to filter out the particulate matter. These filters, made of mixed cellulose esters, had the advantage of providing filtrates having low blanks for the metals. Each 0.45-µm filter was able to filter ∼500 mL of water samples without clogging. A subsample (250 mL) of the filtered samples was acidified with nitric acid (Ultrex II) to pH 1.5. The acidified samples were used to determine the total metal concentrations of trace metals (Ni, Cu, Co, Al, Fe) using GFAAS, of major metals (Ca, Mg) using flame atomic absorption spectrometry and flame atomic emission spectrometry (Na, K), and of total anions (NO3-, F-, Cl-, Br-,
SO42-) using ion chromatography. The remaining filtered samples were stored in the dark at 4 °C and were used for the kinetic measurement within 72 h of the filtration. Subsamples were used for toxicity tests (see below). Method for Kinetic Measurement. The kinetics of uptake by Chelex-100 of free nickel arising from the dissociation of Ni-DOC complexes as a function of time was measured by GFAAS as follows. One percent (w/v) of pretreated Chelex100 resin was added to a 300 mL of the above samples placed in the reactor, and the mixture was stirred with a Tefloncoated magnetic stirring bar. Test solutions for the kinetic run were withdrawn from the reactor through a nylon membrane filter placed at the end of the reactor tube, which filtered out the Chelex-100 resin and injected a clear filtrate into the graphite furnace. The zero time for the kinetic measurement was taken as the time at which the Chelex-100 resin first came in contact with the test solution in the reactor. The concentrations of the nickel complexes remaining in the test solution, CML(t), (eq 4) was measured by GFAAS as a function of time. For the determination of nickel concentration, 20 µL of the test solution was injected into the graphite furnace where it was dried, ashed, and atomized. The signal was measured in the peak area mode. Each completed determination was followed by a 4-s cleanup cycle of the graphite furnace at 2400 °C. During the drying, ashing, and cleanup cycles, the internal argon gas was passed through the graphite furnace at 250 mL/min. The internal argon gas flow was interrupted during the atomization cycle but was restored for the cleanup cycle. The changes in the pH of the test solutions between the beginning and the end of the experiment were found to be 20%. The relative standard deviation among replicate determinations was typically e5%. Kinetic Data Analysis. The experimental data were analyzed for discrete values of the dissociation rate coefficients by a nonlinear regression method based on the Marquardt-Levenberg algorithm (19). All the experimental data were used in the fitting. Toxicity Tests. Toxicity tests were conducted for the samples collected in 1999 using the standard method recommended by Environment Canada (20). The inhibition of growth yield of exponentially growing cultures of the alga Pseudokirchneriella subcapitata (formerly called Selenastrum capricornutum in this method) incubated at 24 °C under continuous, cool, white fluorescent lights provided a measure of metal toxicity (21). Each water sample was tested at 6 dilutions and a control in 4 replicates was run concurrently. Cell counts in the culture used for the bioassays were made each day so that the growth rate conformed to that required in the standard method. In general, fresh media was inoculated on Wednesday for use on the following Monday. Algae were grown in a 200-µL volume of 96-well microplates at sample dilutions of 50, 25, 12.5, 6.3, and 3.1% as well as distilled water (4 replicates). Starting concentrations were 10 000 cells/mL. In addition, 10 µL of culture medium was added to each sample to ensure that no nutrients or trace elements were missing from the test samples. Cells were counted after 72 h using a Neubaueur (hemacytometer) counting chamber (22) at 100× magnification. The cladoceran Daphnia magna has been used as an international standard organism for toxicity tests (23). Ten neonates, less than 24 h old, from a culture of rapidly reproducing adults were exposed in 3 replicates of 25 mL at concentrations of 100, 50, 25, 12.5, and 6.3% of the sample for 24 h. A zero sample control consisting of culture media
FIGURE 1. Percentage of nickel-DOC complexes remaining undissociated as a function of time. Symbols for samples are arranged in increasing order of free plus labile nickel concentrations. The error bars represent 95.5% confidence intervals. (a) Water samples collected in October 1998: 3, sample 10 (Ni/DOC 9.8 × 10-5 mol/g); O, sample 7 (Ni/DOC 1.4 × 10-4 mol/g); 0, sample 6 (Ni/DOC 5.9 × 10-4 mol/g); 4, sample 3 (Ni/DOC 7.2 × 10-4 mol/g); ], sample 8 (Ni/DOC 2.0 × 10-3 mol/g); b, sample 1 (Ni/DOC 7.3 × 10-3 mol/g). (b) Water samples collected in June 1999: O, sample 10 (Ni/DOC 9.3 × 10-5 mol/g); 0, sample 6 (Ni/DOC 1.8 × 10-4 mol/L); b, sample 3 (Ni/DOC 6.0 × 10-4 mol/g); ], sample 4 (Ni/DOC 3.5 × 10-4 mol/g); 4, sample 8 (Ni/DOC 6.8 × 10-4 mol/g); 3, sample 1 (Ni/DOC 8.2 × 10-4 mol/g). and a positive control of 0.3 µg/mL potassium chromate was also used. The end point was death, but the sublethal end point of immobility was also used, in which case the organism was unable to swim or move but the heart was still beating. The freshwater cnidarian, Hydra attenuata, provides another widely used toxicity estimate for industrial effluents and toxic chemicals (24). It is easy to culture, it is ubiquitous in freshwaters, and recognizable morphological changes occur because of exposure to toxic chemicals. Clubbed tentacles, shortened body, and tulip phase are sublethal end points used after 96-h exposures (25). Death and disintegration of the body occurs under more severe exposure conditions. Three hydras were exposed in triplicate in 4-mL samples at sample dilutions as above. The positive control was 0.2 µg/mL potassium chromate.
Results and Discussion Nickel Released from DOC-Bound Nickel Complexes. Percentage of the Ni-DOC complexes remaining undissociated in the solution was plotted as a function of time for the water samples collected in October 1998 (Figure 1a) and June 1999 (Figure 1b). Dissociation rate coefficients (see Kinetic Data Analysis) and concentrations of the kinetically distinguishable components of the nickel complexes are VOL. 36, NO. 7, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1479
TABLE 2. Kinetically Distinguishable Components of Ni Complexes in Water Samples Collected in October 1998 kinetically distinguishable components
sample
Ni/DOC (mol/g)
C1 (%)
C2 (%)
1 3 5 6 7 8 9 10
7.2 × 10-4 2.0 × 10-3 5.3 × 10-4 5.9 × 10-4 1.4 × 10-4 7.3 × 10-3 2.3 × 10-4 9.8 × 10-5
96.6 ( 5.0 87.9 ( 4.8 69.5 ( 4.0 71.5 ( 5.0 62.9 ( 1.5 95.3 ( 1.5 65.9 ( 2.5 50.0 ( 1.3
12.2 ( 0.5 30.4 ( 1.0 26.1 ( 1.7 37.1 ( 1.3 4.9 ( 0.1 33.8 ( 0.8 49.5 ( 1.8
dissociation rate coefficients C1, k1 × 103 C2, k2 × 104 (s-1) (s-1) 5.9 ( 0.1 5.5 ( 0.1 4.5 ( 0.2 5.1 ( 0.2 4.0 ( 0.2 6.3 ( 0.2 4.1 ( 0.1 3.0 ( 0.0
1.6 ( 0.1 1.4 ( 0.5 1.5 ( 0.5 0.7 ( 0.2 5.7 ( 0.2 0.9 ( 0.1 0.5 ( 0.1
TABLE 3. Kinetically Distinguishable Components of Ni Complexes in Water Samples Collected in June 1999
FIGURE 2. (a) Concentrations of free plus labile nickel vs Ni/DOC (mol/g) ratio and (b) dissociation rate coefficients vs Ni/DOC (mol/ g) ratio for the water samples collected in June 1999. presented in Tables 2 (1998) and 3 (1999), which show that the concentration of the fastest kinetically distinguishable component (C1) representing ∼30-95% of the total nickelDOC complexes has a dissociation rate coefficient similar to that reported for [Ni(H2O)6]2+ in the literature (7). Figure 2 shows that, for most of the samples, the concentration of free plus labile nickel (a) and the corresponding dissociation rate coefficients (b) increases with increasing Ni/DOC (mol/ g) ratio. Morel and Hering (26) have reported that ca. 40% of DOC is fulvic acid and ca. 10% is humic acid. Tipping (27) has suggested that ca. 50% of DOC is humic substances. If we assume that ca. 50% of the DOC is humic substances with increasing concentrations of nickel at the fixed, low concentration of the DOC (the DOC values are almost the same for all the samples, as shown in Table 1), the Ni/DOC ratios become large enough to saturate all the strong binding sites (∼1-10% of the total binding sites) of the DOC; the remaining nickel then binds to the weak binding sites (∼99-90% of the total binding sites) of the DOC. These weak nickel complexes are labile, releasing free Ni2+ ion. According to the Free Ion Activity Model (28, 29), free metal ion activity is a good predicator of metal bioavailability and, hence, of toxicity. In Figure 2, in the absence of knowledge of the concentrations of binding sites of the DOC and hence of the Ni/DOC mole ratio, which would have been more desirable, we have plotted Ni/DOC (mol/g) ratio. Since the DOC concentrations of the water samples are similar (Table 1), Ni/DOC (mol/g) ratio provides a reasonable surrogate for Ni/DOC mole ratio. Table 2 for the 1998 samples shows that the Ni/DOC (mol/ g) ratio ranged from 9.8 × 10-5 to 7.3 × 10-3 and that the fraction of labile nickel species varied from 50 to 97%. For 1480
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 7, 2002
kinetically distinguishable components
sample
Ni/DOC (mol/g)
C1 (%)
C2 (%)
1 2 3 4 5 6 7 8 10
3.5 × 10-4 1.0 × 10-3 6.8 × 10-4 8.2 × 10-4 2.8 × 10-4 1.8 × 10-4 1.2 × 10-4 6.0 × 10-4 9.3 × 10-5
88.0 ( 5.5 88.2 ( 4.8 75.6 ( 4.5 82.7 ( 5.0 71.5 ( 5.0 48.1 ( 2.5 40.0 ( 1.5 87.8 ( 4.8 30.7 ( 1.3
11.5 ( 0.6 12.2 ( 0.5 24.4 ( 1.0 16.6 ( 0.6 28.4 ( 1.0 50.6 ( 1.8 59.3 ( 3.1 11.9 ( 0.5 68.4 ( 3.7
dissociation rate coefficients C1, k1 × 103 C2, k2 × 104 (s-1) (s-1) 7.1 ( 0.1 6.9 ( 0.2 5.4 ( 0.1 6.1 ( 0.2 5.1 ( 0.2 5.0 ( 0.1 4.5 ( 0.2 9.5 ( 0.1 4.0 ( 0.0
2.7 ( 0.1 1.8 ( 0.1 1.3 ( 0.1 1.7 ( 0.1 1.4 ( 0.5 1.2 ( 0.5 1.0 ( 0.2 3.3 ( 0.1 0.7 ( 0.1
sample 8, the Ni/DOC ratio (7.3 × 10-3 mol/g) was much higher than that (7.2 × 10-4 mol/g) of sample 1. Yet, both sample 1 and sample 8 have almost the same fraction of labile species (C1 ) 95.0%), which has almost the same dissociation rate coefficients (k1 ) 6.0 × 10-3 s-1). It is also noteworthy that sample 1 has higher conductivity (1.98 mS/ cm), which is probably due to Ca2+ and Mg2+, than that of sample 8 (1.06 mS/cm). A conductivity of 1.98 mS/cm is roughly 1000 times higher than the conductivity of lakes and streams on the Canadian Shield. Table 3 for the 1999 samples shows that the Ni/DOC (mol/ g) ratio ranged from 9.3 × 10-5 to 1.0 × 10-3 and that the fraction of labile nickel ranged from 31 to 88%. In general, the fraction of labile nickel and the corresponding dissociation rate coefficients are lower in samples with low Ni/DOC ratio. However, there are a few exceptions. For sample 1 (Ni/DOC ) 3.5 × 10-4 mol/g) and sample 8 (Ni/DOC ) 6.0 × 10-4 mol/g), the Ni/DOC ratios are lower than those of the following samples: sample 2 (Ni/DOC ) 1.0 × 10-3 mol/g), sample 4 (Ni/DOC ) 8.2 × 10-4 mol/g), and sample 3 (Ni/ DOC ) 6.8 × 10-4 mol/g). Yet, both samples 1 and 8 have higher dissociation rate coefficients (Figure 2b) than those of samples 2-4. These results can be explained as follows. Metal Competition. As demonstrated in our publication on model systems (10), the presence of a massive excess of major cations (Ca2+ and Mg2+) results in the release of larger fractions of bound nickel as free Ni2+ ion, as seen from Table 4: the higher concentrations of major cations (Ca2+ and Mg2+) in sample 1 (8.8 × 10-3 M Ca2+ and 1.6 × 10-3 M Mg2+) and sample 8 (1.2 × 10-2 M Ca2+) result in the release of a larger fraction of free Ni2+ ion as compared to sample 2 (1.4 × 10-3 M Ca2+ and 1.2 × 10-3 M Mg2+), sample 4 (6.4 × 10-4 M Ca2+), and sample 3 (1.8 × 10-3 M Ca2+). Such high concentrations of Ca2+ and Mg2+are not normally found in the poorly buffered lakes and rivers of this region and may have resulted from effluent treatment activities. Our previous work (30) has
TABLE 4. Concentrations of Total Cations and Anions in Water Samples Collected in June 1999 sample
[Ni] (µM)
[Fe] (µM)
[Al] (µM)
[Cu] (µM)
[Co] (µM)
[Ca] (mM)
[Na] (mM)
[K] (mM)
[Mg] (mM)
[NO3-] (µM)
[F-] (µM)
[Cl-] (mM)
[Br-] (µM)
[SO42-] (mM)
1 2 3 4 5 6 7 8 10
2.5 7.3 4.7 4.9 1.8 1.3 0.9 2.1 0.6
0.4 2.4 1.7 0.8 1.8 1.6 0.2 0.5 1.9
1.5 4.1 1.3 4.5 1.8 7.6 7.2 3.1 2.4
0.9 0.6 0.04 0.09 nd 0.1 0.1 0.06 nd
0.1 0.6 0.3 nd nd nd 0.1 nd nd
8.8 1.4 1.8 0.6 1.8 1.3 7.5 12 1.0
13 2.1 8.6 1.7 3.9 2.8 7.0 8.0 1.0
1.4 0.06 0.5 0.5 0.04 nd 1.0 1.0 nd
1.6 1.2 1.3 0.07 1.1 0.5 nd nd nd
131.6 36.3 69 3.9 nd 71 137 1474 5.5
nd nd nd 5.5 18 nd nd 23 84
2.7 3.2 2.9 0.1 0.8 0.04 0.04 0.01 1.2
nd nd 4.3 nd nd 13 16 nd nd
20 1.6 5.4 0.6 1.1 7.8 7.3 0.08 3.4
shown that Cu(II) and Co(II) at concentrations (Table 4) that were present in the water samples would have no effect on the DOC-bound nickel, whereas the Fe and Al (Table 4) present in the samples do have an appreciable effect of releasing Ni2+ ions from DOC-bound nickel (but not reported here). However, the major cations (Ca2+ and Mg2+) at the high concentrations that were present in the water samples had greater effect of releasing Ni2+ ions and other labile nickel species from the DOC-bound nickel (10). The result of sample 8 might be also partly due to the fact that the DOC for this sample was lower (3.51 mg/L) than that of the other samples (∼6-7 mg/L, Table 1). Effect of Major Anions and pH. As Table 4 shows, the major anions present in the samples are NO3- and SO42-. It is reported in the literature (6) that, in freshwaters, inorganic species of nickel are NiCO3 and Ni2+. We, however, did not determine NiCO3. Since NiCO3 is labile, it contributes to the free Ni2+ ions and other nickel labile species determined by CLEM/GFAAS; hence, its omission does not affect the main conclusions of this work. The pH of all the water samples was similar; hence, the pH effect would be also similar and was not discussed. At the pH and the nickel concentrations of the water samples, equilibrium calculations have shown that there would be no hydrolysis of nickel. Polyelectrolyte Effects of DOC (26). When a cation reacts with an acidic functional group of a polyanion, the chemical free energy is augmented by the long-range Coulombic attraction emanating from all the neighboring, nonreacting, negative sites (26). At relatively high pH, this polyelectrolyte effect can strengthen greatly the binding of metal ions by polyanions. The change in free energy of reaction (∆G°) between a metal (M) and the functional group (L) (reaction 1) is the sum of the chemical free energy change (∆G°chem) due to the reaction with the functional group itself and the Coulombic free energy change (∆G°coul) due to electrostatic interactions between M and all the charges on the polyanion (26):
∆G° ) ∆G°chem + ∆G°coul
(5)
According to eq 5, the binding energy of the Ni-DOC complex has two components: (i) chemical binding energy and (ii) Coulombic binding energy. Ca2+ and Mg2+ could compete with Ni2+ for specific binding sites and, because of the competitive advantage of Ca2+ and Mg2+ (especially Ca2+) over Ni2+ in terms of their rates of exchange of coordinated water (31) and their presence in overwhelming excess over the Ni2+, could outcompete Ni2+ for specific binding sites and also preferentially bind to other sites for which they may have greater affinity for binding than Ni2+. Moreover, the overwhelming excess of Ca2+ and Mg2+ would result in a large amount of Ca2+ and Mg2+ in the diffuse electrical double layer of the DOC polyanions, thereby decreasing significantly the total binding energy of the Ni-DOC by screening the charge on the DOC polyanions. Na+ and K+, although not
TABLE 5. WHAM Predictions for Speciation of Nickel in Water Samples Collected in June 1999 WHAM predictions
exptl
Ni2+
sample
Ni/DOC (mol/g)
1 2 3 4 5 6 7 8 10
3.5 × 10-4 1.0 × 10-3 6.8 × 10-4 8.2 × 10-4 2.8 × 10-4 1.8 × 10-4 1.2 × 10-4 6.0 × 10-4 9.3 × 10-5
free free Ni2+ ions plus nonspecific intrinsic ions plus other small binding binding other small Ni species of Ni of Ni Ni species (%) (%) (%) (%) 98.5 81.9 79.7 82.8 75.8 55.9 45.5 90.2 36.5
0.2 0.1 0.2 0.3 0.1 0.2 0.2 0.1 0.3
15.9 15.6 21.1 16.5 31.1 46.4 56.3 11.3 65.1
88.0 88.2 75.6 82.7 71.5 48.1 40.0 87.8 30.7
competing with nickel for the specific binding sites of DOC, would nevertheless provide strong screening of the DOC polyanionic charges by their presence in large excess (Table 4). This would result in further weakening of the Ni-DOC bond and release of more free Ni2+ ions and other labile nickel species, as observed. WHAM Predictions and Experimental Results. The Windermere Humic Aqueous Model (WHAM), model V (32), was used to predict the nickel binding by DOC in the water samples and to compare its prediction with the experimental results. As suggested by Tipping (27), the originator of WHAM, in the WHAM calculation only 50% of the DOC was assumed to be “active” fulvic acid. The WHAM predictions (Table 5) agree well with the experimental results (R2 ) 0.99), even though the model V, unlike the model VI (33), is not optimized for competition of major ions (Ca2+ and/or Mg2+) and ionic strength effects. Bioavailability of and Ecotoxicological Tests on the 1999 Samples. Control samples containing no added sample typically had an algal yield of 200 000-400 000 cells/mL after 72 h of growth under continuous light (Figure 3). One control yield (sample 5) was above 700 000 cells/mL. While this differences may seem high, they represent only about 5, 6, and 7 doublings of the starting level of 10 000 cells/mL. Whenever samples are transferred from one culture to another, a lag phase exists, and this can vary in length by several hours. Furthermore, although the algae were taken from the standard culture media at exponential growth phases, it was at different times. Although all conformed to the standard method used, doubling time (which ranged from 14, 12, and 10 for these three levels) hours would not be surprising and could account for the observed differences in the control biomass. All samples contained added growth media to ensure that none of the samples were nutrientlimited. The results (Figure 3) provide some valuable insights not normally reported in algal toxicological assays. The VOL. 36, NO. 7, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1481
FIGURE 3. Algal bioassays as a function of percent water sample for the water samples collected in June 1999. The error bars represent 95.5% confidence intervals. expected response with increasing sample from 3.6 to 100% was observed for sample 1. Here, algal yield decreased abruptly with only 3.6% of the water sample, followed by very little change even to 100%. The values ((SD) can be used to provide a clear percent inhibition of 75%. This is the number normally reported using this bioassay. Samples 4, 7, and 10, however, showed stimulation with added sample at the lower concentrations, followed by inhibition with further enrichment. Despite having all elements thought to be required for growth present in the control samples, enrichment with the water samples caused an increase in algal yield. Without intermediate points, a researcher would conclude that there was no toxicity, just stimulation! 1482
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 7, 2002
Consequently, here we express the inhibition as a percent of the maximum algal yield. Without question, both stimulation and inhibition were going on at the same time. Figure 4 shows an excellent correlation between (a) percent inhibition and Ni/DOC (mol/g) ratio (R2 ) 0.73), (b) percent inhibition and free plus labile nickel concentration (R2 ) 0.69), and (c) percent inhibition and dissociation rate coefficients (R2 ) 0.75). As explained earlier (Figure 2), the higher Ni/DOC (mol/g) ratio resulted in more nickel being bound to the weaker sites, making the Ni-DOC complex labile. In Figure 4a,b, sample 1 appeared to be an outlier, but as shown in Table 4, it is also high in copper concentration (9.0 × 10-7 M). Sample 2 also contained copper but at a
FIGURE 4. Algal yield expressed as a fraction of maximum algal yield as a function of (a) Ni-DOC (mol/g) ratio, (b) free plus labile nickel concentration, and (c) dissociation rate coefficients, k1. lower concentration (6.0 × 10-7 M) and fitted with the other data better. The copper concentration expected to inhibit growth rates by 50% is 1.2 × 10-6 M. All other samples were low in Cu. For the alga P. subcapitata, the IC50 for algal growth yield was 1.3 × 10-6 M for Ni and 1.1 × 10-6 M for copper (34). Our values are clearly in the range where toxic effects should be observed. With sample 1, the additional stress with copper was observed. Typical values expected for the lethal dose sufficient to kill 50% of the D. magna (LD50) are 2.6 × 10-5 M Ni (34) and 7.1 × 10-7 M for Cu (35). In the 48 h D. magna lethality test, no mortality was observed in any samples but the sublethal effect where organisms show immobility was observed in samples 1, 2, 3, 4, 6, 7, and 8 at 10, 20, 7, 3, 3, 7, and 3%, respectively. No effect was seen in samples 5 and 10. The high Cu values in sample 1 might have been a factor, but the 20% inhibition in sample 2 corresponded to the highest Ni level. Consistent with these observations, only sample 1 resulted in 22% death of H. attenuata in the 96-h test. Both the results of kinetic speciation and the bioassay toxicity tests of the nickel species in the nickel-polluted water samples from the Sudbury (Canada) area show that the release of free Ni2+ ion and other labile nickel species from the Ni-DOC complex and the resulting toxicity, in general, depend on the Ni/DOC ratio. The kinetic lability of the nickel species varies in a reasonable way with the Ni/DOC (mol/g) ratio, suggesting that the kinetic speciation model gives chemically significant metal species. The results of both the kinetic speciation and the bioassay toxicity tests agree reasonably well, suggesting that kinetic lability is a good predictor of nickel toxicity.
The environmental significance of these results is that in nickel-polluted freshwaters a large part of the dissolved nickel is probably in the form of a free Ni2+ ion plus other labile nickel species. High concentrations of major cations (Ca2+ and Mg2+) result in a larger release of the DOC-bound nickel as free Ni2+ ions that are toxic (5, 7). The presence of calcium and magnesium carbonate bedrocks and the occurrence of acid rains to dissolve them into the aquatic environment aggravate this problem. This knowledge will help environmental managers to develop better environmental policies and regulations constraining discharge of effluents.
Acknowledgments Financial support was received from Nickel Producers Environmental Research Association (USA), Inco Ltd., and Falconbridge Ltd. Research grants from NSERC, Metals in the Environment Research Network Grant, Ontario Power Generation Inc., and the Mining Association of Canada are acknowledged. We thank Dr. E. Tipping for providing guidance with WHAM calculations and Dr. E. DabekZlotorzynska, Environment Canada, for providing some equipment and chemicals. J.M. received a NSERC postgraduate scholarship, and N.M.H. received a graduate scholarship from the Government of Libya.
Literature Cited (1) Anderson, M. A.; Morel, F. M. M. Limnol. Oceanogr. 1982, 120, 591. (2) Langford, C. H.; Cook, R. L. Analyst 1995, 120, 591. (3) Figura, P.; McDuffie, B. Anal. Chem. 1979, 51, 120. (4) Figura, P.; McDuffie, B. Anal. Chem. 1980, 52, 1433. VOL. 36, NO. 7, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1483
(5) Stokes, P. In Metal Ions in Biological Systems; Sigel, H., Ed.; Marcel Dekker: New York, 1988; Vol. 23, pp 31-46. (6) Xue, H. B.; Jansen, S.; Prasch, A.; Sigg, L. Environ. Sci. Technol. 2001, 35, 539. (7) Langford, C. H.; Gutzman, D. W. Anal. Chim. Acta 1992, 256, 183. (8) Morel, F. M. M.; Hering, J. G. Principles and Applications of Aquatic Chemistry; John Wiley & Sons: New York, 1993; pp 404-405. (9) Hering. J. G.; Morel, F. M. M. In Aquatic Chemical Kinetics; Stumm, W., Ed.; John Wiley & Sons: New York, 1990; pp 145171. (10) Mandal, R.; Salam, M. S. A.; Murimboh, J.; Hassan, N. M.; Chakrabarti, C. L.; Back, M. H.; Gre´goire, D. C. Environ. Sci. Technol. 2000, 34, 2201. (11) Hering, J. G.; Morel, F. M. M. Environ. Sci. Technol. 1988, 22, 1234. (12) Cao, Y.; Conklin, M.; Betterton, E. Environ. Health Perspect. 1995, 103, 29. (13) Cabaniss, S. E.; Shuman, M. S. Geochim. Cosmochim. Acta 1988, 52, 185. (14) Sunda, W. G.; Hanson, P. J. In Chemical Modeling in Aqueous Systems: Speciation, Sorption, Solubility, and Kinetics; Jenne, E. A., Ed.; ACS Symposium Series 93; American Chemical Society: Washington, DC, 1979. (15) Buffle, J.; Deladoey, P.; Greter, F. L.; Haerdi, W. Anal. Chim. Acta 1980, 116, 255. (16) van den Hoop, M. A. G. T.; van Leeuwen, H. P.; Pinheiro, J. P.; Mota, A. M.; Goncalves, M. de L. S. Colloids Surf. 1995, 95, 305. (17) Olson, D. L.; Shuman, M. S. Geochim. Cosmochim. Acta 1985, 49, 1371. (18) Chakrabarti, C. L.; Back, M. H.; Gregoire, D. C.; Schroeder, W. H. Anal. Chim. Acta 1994, 293, 95. (19) Marquardt, D. W. J. Soc. Ind. Appl. Math. 1963, 11, 431. (20) Environment Canada. Biological Test Method: Growth Inhibition Test Using the Freshwater Alga Pseudokirchneriella subcapitata; Environment Protection Series 1/RM/25; Environment Canada: Ottawa, ON, Canada, 1992.
1484
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 7, 2002
(21) Nyholm, N.; Peterson, H. G. Laboratory Bioassays with Micro Algae. In Plants for Environmental Studies; Wang, W., Gorsuch, J. W., Hughes, J. S., Eds.; CRC Press: Boca Raton, FL, 1997. (22) Blaise, C.; Trottier, S. Environ. Toxicol. Water Qual. 2000, 15, 352. (23) Environment Canada. Biological Test Method: Acute Lethality Test Using Daphnia spp.; Environment Protection Series 1/RM/ 11; Environment Canada: Ottawa, ON, Canada, 1990; p 57. (24) Trottier, S.; Blaise, C.; Kusui, T.; Johnson, E. M. Environ. Toxicol. Water Qual. 1997, 12, 265. (25) Campbell, R. D.; Bode, H. R. In Terminology for Morphology and Cell Types in Hydra: Research Methods; Lenhoff, H. M., Ed.; Plenum Press: New York, 1983; pp 5-14. (26) Morel, F. M. M.; Hering, J. G. Principles and Applications of Aquatic Chemistry; John Wiley & Sons: New York, 1993; pp 380-388. (27) Tipping, E. private communication. (28) Morel, F. M. M. Principles of Aquatic Chemistry; John Wiley & Sons: New York, 1983; pp 300-308. (29) Campbell, P. G. C. In Metal Speciation and Bioavailability in Aquatic Systems; Tessier, A., Turner, D. R., Eds.; IUPAC/John Wiley & Sons: New York, 1995; p 45. (30) Mandal, R.; Sekaly, A. L. R.; Murimboh, J.; Hassan. N. M.; Chakrabarti, C. L.; Back, M. H.; Gregoire, D. C.; Schroeder, W. H. Anal. Chim. Acta 1999, 295, 309. (31) Morel, F. M. M.; Hering, J. G. Principles and Applications of Aquatic Chemistry; John Wiley & Sons: New York, 1993; p 400. (32) Tipping, E. Comput. Geosci. 1994, 20, 973. (33) Tipping, E. Aquat. Geochem. 1998, 4, 3. (34) Rahayu, U. M.Sc. Thesis, University of Ottawa, Ottawa, Canada, 2001. (35) Winner, R. W. Water Res. 1986, 19, 449.
Received for review July 24, 2001. Revised manuscript received January 6, 2002. Accepted January 17, 2002. ES015622E