- Email: [email protected]
Assessing acute toxicity potential of persulfate ISCO treated water
Chemosphere 93 (2013) 2711–2716
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Assessing acute toxicity potential of persulfate ISCO treated water Chenju Liang ⇑, Chi-Wei Wang Department of Environmental Engineering, National Chung Hsing University, 250 Kuo-Kuang Road, Taichung 402, Taiwan
h i g h l i g h t s Evaluation of the environmental risk of exposure to persulfate is studied. Persulfate is considerably more toxic than its decomposition product sulfate. Very acidic or basic pH conditions may cause toxic effects. Water conductivity reveals minor influence on acute toxicity. The bioassay results serve as reference toxicity values following ISCO application.
a r t i c l e
i n f o
Article history: Received 19 February 2013 Received in revised form 24 August 2013 Accepted 26 August 2013 Available online 4 October 2013 Keywords: Sodium persulfate Remediation In situ chemical oxidation Aquatic toxicity Alkaline activated persulfate
a b s t r a c t Persulfate anion (S2 O2 8 ), a widely used in situ chemical oxidation agent, is increasingly applied for environmental remediation. However, limited information on environmental and toxicological effects is available for the evaluation of the environmental risk of exposure to S2 O2 8 , particularly after its application. In this study, the acute toxic effects on the common carp (Cyprinus carpio) were employed as a 2 2 model to investigate S2 O8 , sulfate ion (decomposition product of S2 O8 ), hydrogen/hydroxide ions and also the mixtures of these ion species. Acute toxicity test results showed 96 h median lethal concentra2 1 for SO2 tions (LC50) of 540 ± 23 mg L1 for S2 O2 8 and 4100 ± 110 mg L 4 . S2 O8 was considerably more . Additionally, solution pH was also an important factor influtoxic than its decomposition product SO2 4 encing toxicity, and S2 O2 8 posed reduced acute toxicity when pH was in the range of 6–10. Water conductivity up to approximately 8000 lS cm1 did not appear to significantly increase fish mortality. In 1 for S2 O2 and 23 ± 2 mg L1 the mixture toxicity test (i.e., S2 O2 8 /OH ), LC50 values of 130 ± 10 mg L 8 for OH were lower than those obtained from the individual toxicity tests and therefore exhibited higher toxicity to fish. However, upon complete decomposition of S2 O2 8 in the mixture, a reduction in acute toxicity may be expected. The results of this study revealed that it may be necessary and/or desirable to con2 addition when potential exposure to an aquatic system is a trol the residual S2 O2 8 and pH after S2 O8 concern. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Persulfate ðS2 O2 8 Þ, which is a strong oxidant with a redox potential of 2.01 V (Eq. (1)), was initially introduced as a potential in situ chemical oxidation (ISCO) agent approximately a decade ago and is now recognized as effective in degrading a variety of organic contaminants present in soil and groundwater. Persulfate reaction chemistry is complicated and involves free radical chain reactions (sulfate radical ðSO 4 Þ and/or hydroxyl radical ðHO Þ formation) upon S2 O2 activations. Chemical oxidation with 8 S2 O2 8 can be combined with the use of heat, transitional metals, hydrogen peroxide, and alkaline pH, to increase the reaction rate (Huling and Pivetz, 2006; Liang and Guo, 2012).
⇑ Corresponding author. Tel.: +886 4 22856610; fax: +886 4 22862587. E-mail address: [email protected] (C. Liang). 0045-6535/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.chemosphere.2013.08.078
2 S2 O2 8 þ 2e ! 2SO4
E ¼ 2:01 V
ð1Þ S2 O2 8
The addition of large quantities of anions to groundwater for contaminant remediation presents a concern about whether there may be toxic effects on the ecosystem. A median lethal concentration at which 50% of organisms die (i.e., LC50) in acute whole effluent toxicity tests, to estimate the lethality of a specific S2 O2 solution, was reported in a Material Safety Data Sheet to 8 be 771 mg L1 (Bluegill sunfish, 96 h exposure); 519 mg L1 (Grass shrimp, 96 h exposure); 133 mg L1 (Daphnia, 48 h exposure); and 163 mg L1 (Rainbow trout, 96 h exposure) (FMC-Corporation, 2009). However, there are still very few ecotoxicology studies that have evaluated S2 O2 ISCO application. It is essential to consider 8 the potential toxicity with respect to residual S2 O2 after ISCO 8 treatment and its associated decomposition end products, including sulfate ion and hydrogen ion. Upon S2 O2 activation, SO 8 4 accepts an electron to form SO2 (Eq. (2)) and hydrolysis of 4
2712
C. Liang, C.-W. Wang / Chemosphere 93 (2013) 2711–2716
S2 O2 8 in water would naturally result in a decrease in pH (Eqs. (3) and (4)). When the solution pH is higher than the pKa value of 1.92, 2 SO2 4 is the major end products of S2 O8 oxidation. Combining Eqs. (3) and (4) yields Eq. (5), which shows that the hydrolysis of 1 mol 2 + of S2 O2 8 would produce 2 mol of SO4 and H . 2 SO 4 þ e ! SO4
ð2Þ
S2 O2 8 þ H2 O ! 2HSO4 þ 1=2 O2
ð3Þ
þ HSO4 ! SO2 4 þH
ð4Þ
pK a ¼ 1:92
2 þ S2 O2 8 þ H2 O ! 2SO4 þ 2H þ 1=2 O2
ð5Þ
Moreover, SO 4 present in the aqueous phase could interact with water and hydroxide ions in accordance with Eqs. (6) and (7), respectively. Addition of OH (e.g., sodium hydroxide) to the 2 S2 O2 8 solution, produces an alkaline activated S2 O8 system, which has been widely used and shown effective for remediating organic contaminants on the basis of the simultaneous presence of active radical species (e.g., SO 4 and HO ) (Liang and Su, 2009). 2 þ SO 4 þ H2 O ! SO4 þ HO þ H
ð6Þ
2 SO 4 þ OH ! SO4 þ HO
ð7Þ
Groundwater quality policy is usually based on a ‘‘chemicalspecific’’ method (Tišler et al., 2004) and only a limited number of chemicals (e.g., excluding S2 O2 8 ) are prescribed by legislation for injection to the subsurface environment in Taiwan. Even though acute toxicity assessment for S2 O2 and other associated 8 ions involved during S2 O2 application might not be related to 8 any legal limits, test organisms exposed to whole S2 O2 treated 8 effluents can be a useful reference for understanding the potential ecological impacts to specific chemicals and combined effects in the S2 O2 treated water. The present study was performed to 8 determine acute toxicity of persulfate, sulfate ion, hydrogen/ hydroxide ions and also mixtures of these ion species. The common carp (Cyprinus carpio, which is one of the biological organisms allowed in Taiwan’s biological acute toxicity testing for the Water Pollution Control Act) was selected for the acute toxicity tests to determine the LC50, and variations of pH and conductivity were applied to assess the causes of toxic effects during the course of experiments. 2. Materials and methods 2.1. Chemicals All reagents were of analytical grade and used without further purification. Magnesium sulfate (MgSO47H2O, 100.4%) and acetone ((CH3)2CO, min 99.5%) were purchased from J.T. Baker (Center Valley, PA, United States of America (USA)); sodium persulfate (Na2S2O8, min 99.0%) and calcium sulfate dehydrate (CaSO42H2O, min 99.0%) were purchased from Merck (Drmstadt, Germany); sulfuric acid (min 95%), potassium chloride (min 99.5%), sodium bicarbonate (min 99.7%), sodium hydroxide (min 99%) and sodium sulfate (min 99%) were purchased from Sigma–Aldrich (St. Louis, MO, USA). The water used was purified using a Millipore reverse osmosis (RO) purification system. 2.2. Fish stock and care C. carpio were obtained from a local aquaculture farm (Changhua, Taiwan), and their body length were 2.0–3.0 cm. C. carpio (approximately 200 fish) were acclimatized for several days in
glass tanks (50 L) to assess whether they were healthy enough to be used in the experiments, i.e., the mortality rate was less than 10% within 7 d prior to the experiment. During the period of acclimatization, C. carpio were fed on commercially available fish feed once a day at least, and the environmental conditions were maintained at a 16 h photo-period and temperature 25 ± 1 °C (controlled by the air conditioner). Water quality parameters including conductivity, dissolved oxygen (DO), and pH were monitored to document the conditions during the test. The fish used in the test was selected from a tank containing healthy acclimatized carp. In addition, before carrying out the experiment, the test fish were quarantined 1 d without feeding. 2.3. Monitoring of water characteristics The pH, temperature, DO and conductivity were measured periodically using a portable pH/ORP meter (Suntex TS-100, New Taipei, Taiwan) equipped with a pH electrode and a temperature probe, a portable DO meter (Hanna HI 9146, Woonsocket, RI, USA) and a pH/conductivity meter (Eutech PC 5000, Singapore) equipped with a Fisher Scientific Accumet four-cell conductivity probe, respectively. The S2 O2 anion concentration was deter8 mined using a spectrophotometer (Hach DR/2400, Loveland, CO, USA) at 400 nm (Liang et al., 2008). 2.4. Experimental design The first phase of the experiments was designed to determine 2 + LC50 for S2 O2 8 , SO4 , H and OH . Note that sulfuric acid and sodium hydroxide were used to adjust the pH of the solutions. The concentration ranges associated with the toxicity of these reagents were unknown prior to the experiment. Therefore, several preliminary range-finding tests were conducted using groups of 5 fish, which were exposed to several widely-spaced sample dilutions to estimate the toxicity range (note that these results exhibit no statistical significance and are not reported.). The determination of the LC50 for S2 O2 8 and OH separately will distinguish the individual toxicities of the ions. Thereafter, in the second phase of the experiments, the toxicity of the simultaneous presence of both S2 O2 8 and OH was evaluated. In the bioassays of these combined ion solutions, the maximum concentration of each ion was based on the LC50 of the more toxic individual ion. For example, in deter mining the anion concentrations in the mixture of S2 O2 8 /OH , if the S2 O2 anion shows a relatively lower LC value that is 1.0 M, 50 8 the maximum OH concentration of 2.0 M would be used because + 1 mol of S2 O2 8 produces 2 mol of H which require 2 mol of OH to maintain its initial pH level upon complete decomposition of S2 O2 8 (see Eq. (5)). In order to determine the acute toxicity of the mixture of S2 O2 8 /OH upon complete decomposition, the mixture solution was heated in a water bath at 70 °C for 4 d to completely decom2 pose S2 O2 8 to SO4 and the resulting solution was cooled to room temperature (25 ± 1 °C) prior to the test. The acute toxicity tests were performed in accordance with the procedure of the Taiwan National Institute of Environmental Analysis (NIEA) method (NIEA, 2005). The toxicity testing procedure includes the preparation of five 1.5 L solutions in 2 L glass beakers, each containing a different concentration of the test ions (i.e., serial dilutions are made with: 20%, 40%, 60%, 80%, and 100% of the test solution). Experiments were carried out using 10 fish in each beaker for a period of 96 h. No food was given during the experiments. The number of fish was counted at least every 24 h and the dead fish were removed from the beaker immediately. The water quality parameters and the mortality rate were determined at the end of 0, 2, 24, 48, 72 and 96 h. Duplicate experiments were conducted and averages were reported. Control tests in RO water only were also conducted side by side.
2713
C. Liang, C.-W. Wang / Chemosphere 93 (2013) 2711–2716
2.5. Statistical analyses
(a)
100
2-
S2O8
80
40 Symbols overlaid
20 0
(b)
100 80
2-
SO4
2 + 3.1. Individual toxicity of S2 O2 8 , SO4 , H and OH 2 + The mortality of C. carpio depending on S2 O2 8 , SO4 , H and OH ions and associated variations of pH and conductivity in solutions during a 96 h period, are shown in Figs. 1 and 2, respectively. The determined LC50 values for individual solutions of persulfate and sulfate anions were 540 ± 23 mg L1 (2.82 mM) and 4100 ± 110 mg L1 (42.72 mM) for S2 O2 and SO2 , respectively (Table 1). 8 4 When comparing pH and conductivities in test solutions, pH for both tests under different ion concentrations were similar at approximately neutral pH conditions (pH 6–7), but the conductiv2 ities were much higher in the SO2 4 solution than the S2 O8 solution (e.g., 8000 lS cm1 for 39.44 mM SO2 solution versus 800 4 lS cm1 for 2.69 mM S2 O2 solution). It appears that water con8 ductivity up to approximately 8000 lS cm1 is not responsible for the fish mortality based on the results of Figs. 1b and 2b showing no dead fish in test solution containing 39.44 mM SO2 4 . Additionally, water characteristics such as pH, which remained in the range of 6 to 8 in these two experiments, can be excluded as a cause of fish mortality based on experimental results reported by Chen et al. (2011) that the variation of pH between 6 to 8 would not significantly affect the physiology of fish. Note that the initial 2 S2 O2 8 concentration used in the S2 O8 toxicity test was 3.36 mM, 2 which would produce 6.72 mM of SO4 upon complete decomposition in accordance with Eq. (5). If 3.36 mM of S2 O2 8 can be completely decomposed in solution, its decomposition end product SO2 4 (6.72 mM) would not cause lethal effects in these short-term acute toxicity experiments. However, S2 O2 8 concentration in these toxicity tests remained nearly the same (less than 2% variation) after 96 h exposure to fish. Therefore, it can be concluded that a possible direct mechanism of S2 O2 toxicity can be related to its 8 oxidative chemical property. Water characteristics such as neutral pH, conductivity (800 lS cm1) and its decomposition end product SO2 4 (39 mM) may insignificantly cause fish mortality.
49.30 mM 39.44 mM 29.58 mM 19.72 mM 9.86 mM Blank
60 40 Symbols overlaid
20 0
Symbols overlaid
(c)
100
H
80
+
0.20 mM 0.16 mM 0.12 mM 0.08 mM 0.04 mM Blank
60 40 20
Symbols overlaid
0
(d)
100
Symbols overlaid
80 3. Results and discussion
3.36 mM 2.69 mM 2.02 mM 1.34 mM 0.67 mM Blank
60
Mortality (%)
Data derived from tests were analyzed by four methods to produce LC50 values and associated 95% confidence intervals, i.e., the Spearman–Karber method, the Trimmed Spearman–Karber method, the Probit method and the Graphical method (only recommended when there are no partial mortalities). The analysis scheme is in accordance with the procedure of acute toxicity data analysis reported in Taiwan NIEA (2005), in which the flowchart for determination of the LC50 for multi-effluent concentration acute toxicity tests was adapted from the U.S. Environmental Protection Agency (USEPA, 2002) (note that the flowchart can be seen in Fig. SM-1 in Supplementary Material (SM)). In brief, when two or more of the observed proportion mortalities are between zero and one and the calculated Chi-square test for data heterogeneity is less than a tabular value at 0.05, the Probit method (a parametric statistical procedure) can be used to obtain an estimate of the LC50 and associated 95% confidence interval. However, if one or more partial mortalities are observed and the zero and 100% of mortalities happen in the lowest and highest aqueous concentrations, respectively, the Spearman–Karber method (a non-parametric statistical procedure) is used for estimating the LC50 and associated 95% confidence interval. Use of the Trimmed Spearman–Karber method (a modification of the Spearman–Karber non-parametric statistical procedure) is only appropriate when the requirements for the Probit method and the Spearman–Karber method are not met. In addition, if there are no partial mortalities, the Graphical method (a mathematical procedure) can be used, but it does not provide a confidence interval for the LC50 estimate.
OH
-
2.5 mM 2.0 mM 1.5 mM 1.0 mM 0.5 mM Blank
60 40 20
Symbols overlaid
0 0
24
48 Time (h)
72
96
2 + Fig. 1. Mortality of C. carpio depending on (a) S2 O2 8 , (b) SO4 , (c) H , and (d) OH concentrations with 96 h aqueous exposures.
Furthermore, the effects of H+ and OH on the fish death were also evaluated to determine their LC50 concentrations (see Fig. 1c and d, respectively). The LC50 for H+ and OH were determined to be 0.18 mg L1 (0.18 mM) and 24 ± 2 mg L1 (1.45 mM), respectively, and also presented in Table 1. The H+ concentration of 0.16 mM (and lower) exhibited no dead fish (corresponding to the measured pH range of 6–8). It should be noted that the calculated initial pH ranged from 3.7 ([H+] = 0.2 mM) to 4.4 ([H+] = 0.04 mM), which is different from the measured pH range and it was found that the solution pH initially increased and then gradually approached neutral during the test in the presence of fish (see Fig. 2c). However, in the initial solution where [H+] = 0.2 mM, all fish died soon after they were placed into the solution (i.e., within 2 h) and the measured pH gradually increased to around 6. This final pH of 6 is lower than the pH of around 7 which resulted from other test concentrations in the presence of live fish. Suomalainen et al. (2005) studied the buffering capacity of fish skin mucus in water and reported that fish skin mucus is an efficient buffer against decreased water pH and the pH of the skin could be higher than that of the mucus. Hence, this may explain the fish were modifying the pH and the solution pH approached neutral during the test. However, when the H+ slightly increased to 0.20 mM and pH decreased to 4, all fish died (just after 2 h exposure, Fig. 1c). As for OH, its concentration of 1.0 mM and a
2714
C. Liang, C.-W. Wang / Chemosphere 93 (2013) 2711–2716
11
11
(a)
10
pH
8
(b)
49.30 mM 39.44 mM 29.58 mM 19.72 mM 9.86 mM Blank
10
3.36 mM 2.69 mM 2.02 mM 1.34 mM 0.67 mM Blank
9
9 8 7
7
Cond. (uS/cm)
2-
SO4
2-
S2O8
6
6
1200
12000
1000
10000
800
8000
600
6000
400
4000
200
2000
0 11
0 11
(c)
10
(d)
10
9 9
pH
8 7
8
6
7
5 6
4 1200
1200 -
+
OH
H
Cond. (uS/cm)
1000 800 600
2.5 mM 2.0 mM 1.5 mM 1.0 mM 0.5 mM Blank
1000
0.20 mM 0.16 mM 0.12 mM 0.08 mM 0.04 mM Blank
800 600
400
400
200
200
0
0 0
24
48
72
96
Time (h)
0
24
48
72
96
Time (h)
2 + Fig. 2. Variations of pH and conductivity in (a) S2 O2 8 , (b) SO4 , (c) H , and (d) OH ions solutions of acute toxicity tests.
measured pH 10 resulted in no dead fish and solution pH would also gradually approach neutral in the presence of fish (Suomalainen et al., 2005). However, when OH increased to above 1.5 mM and pH was greater than 10, fish death occurred. It should be mentioned that conductivities in these tests (100–600 lS cm1) were lower than 1 those in the S2 O2 for 2.02 mM of S2 O2 8 tests (e.g., 700 lS cm 8 solution with no dead fish (see Fig. 1a). Hence the conductivity can be excluded as a major factor causing fish death, and mortality was directly related to H+ or OH species. These results indicated that in addition to the acute toxicity of specific chemical species, pH is also an important factor and solution pH beyond the range of 6–10 may cause acute toxicity effects. Furthermore, Fig. 3 illustrates some relationships between Na2S2O8 and Na2SO4. The LC50 for Na2SO4 is 42.72 mM, which is equivalent to the SO2 concentration resulting from complete 4 decomposition of 21.36 mM Na2S2O8 (Eq. (1)). In addition, as shown in Fig. 3, if the LC50 concentration of 2.82 mM S2 O2 8 is con2 verted to SO2 4 , it generates 5.64 mM SO4 , a concentration that exhibits no toxicity to fish. In the S2 O2 8 ISCO application, if the high concentration of residual S2 O2 can be completely transformed 8
into SO2 4 , a reduction of oxidant impact (e.g., acute toxicity) on the environment may be expected. Therefore, further experiments were performed to verify this assumption. The solution containing 21.36 mM of S2 O2 was thermally 8 decomposed (70 °C, Eq. (8)) to produce a 42.72 mM solution of SO2 (cooled to 25 ± 1 °C) (combining Eqs. (2) and (8)) (Liang 4 et al., 2007; Liang and Su, 2009) and the mortality of fish exposed to this decomposed solution during the 96 h acute toxicity test can be seen in Fig. 4. The LC50 for this decomposed solution of SO2 4 was then determined to be 24.46 mM, which is lower than 42.72 mM determined from the individual SO2 4 test. Moreover, the pH upon 2 decomposition of S2 O2 8 was approximately 4 for two SO4 levels of 42.72 and 34.18 mM, which was toxic and no fish survived (see Fig. 4). It should be noted that these two SO2 4 concentrations (42.72 and 34.18 mM), if tested individually, would exhibit no or partial fish death. Therefore, cause of the acute toxic effects appears to be partly due to the acidic pH (i.e., 4, see Fig. 5). Furthermore, the pH of the decomposed 25.64 mM SO2 solution was 4 initially 4.5 and gradually increased to above 6. Therefore, partial fish death was observed in the early stage and no fish death
2715
C. Liang, C.-W. Wang / Chemosphere 93 (2013) 2711–2716 Table 1 96 h LC50 values for seven acute toxicity bioassays, and analysis modes. Objective
Species
Individual toxicity
Initial conc.
a
1
mM
mM
mg L
S2 O2 8
3.36
2.82 ± 0.12
540 ± 23
SO2 4 H+ OH
49.30
42.72 ± 1.15
4100 ± 110
Probit
0.20 2.50 42.72
0.18 1.43 ± 0.11 24.46 ± 1.96
0.18 24 ± 2 2300 ± 188
Graphical Spearman–Karber Spearman–Karber
2 SO2 4 (S2 O8 decomposed)
Mixture toxicity
Analysis modea
LC50
Spearman–Karber
(S2 O2 8 /OH
1.25/2.50
0.67 ± 0.05/1.34 ± 0.09
130 ± 10/23 ± 2
Spearman–Karber
(S2 O2 8 /OH (decomposed)
3.36/6.72
n.a.
n.a.
n.a.
In accordance with the procedure established by USEPA (2002); n.a.: not available.
21.5
10 21.36 mM
21.0
2-
9
decomposed to 2SO4
8
3.5
pH
100 LC50 = 42.72 mM
80
40
60
30
40
20
2.5 2.0 1.5 1.0
7 6
Na2SO4 (mM)
3.0 LC50 = 2.82 mM
LC50 (%)
Na2S2O8 (mM)
20.5
Tolerable S2O8
5 4 18000
2-
0.5
16000
42.72 mM 34.18 mM 25.64 mM 17.08 mM 8.54 mM Blank
5.64 mM
0
0
Na2S2O8
14000
Na2SO4
Fig. 3. Relationship of Na2S2O8 and Na2SO4 concentrations and LC50 values to C. carpio bioassays.
Cond. (uS/cm)
0.0
2-
SO4 (S2O8 decomposed)
10
20
12000 10000 8000 6000 4000
100 2-
SO4 (S2O8 decomposed)
Mortality (%)
Symbols overlaid
42.72 mM 34.18 mM 25.64 mM 17.08 mM 8.54 mM Blank
80
60
2000
2-
0 0
24
48
72
96
Time (h) Fig. 5. Variations of pH and conductivity in SO2 solution of acute toxicity tests, h i 4 resulting from complete 70 °C thermal S2 O2 decomposition, initial S2 O2 = 8 8 21.36 mM.
40 Symbols overlaid
20
S2 O2 8
ambiant ð1030 CÞ or elevated ð3090 CÞ temperature
!
2SO 4
ð8Þ
0 0
24
48
72
96
Time (h) Fig. 4. Mortality of depending on SO2 h C. carpio i 4 concentrations with 96 h aqueous exposures. Initial SO2 = 42.72 mM (resulting from 21.36 mM Na2S2O8 decompo4 sition at 70 °C).
occurred thereafter when pH increased to near neutral. The conductivity range for this 25.64 mM test was approximately 1 6000 lS cm1 for the SO2 , as discussed 4 (lower than 8000 lS cm earlier) and revealed minor influence on acute toxicity. It should also be noted that the results of control tests (in the presence of RO water only) show no dead fish. DO values were within the range of 4–8 mg L1 during the course of all acute toxicity tests.
2 + 3.2. Mixture toxicity of S2 O2 8 , SO4 , H and OH
Alkaline activation is one of widely used S2 O2 8 activations for ISCO remediation of organic contaminants. In order to understand its toxicity in aquatic systems, two mixed solutions were tested. The first solution is an alkaline activated S2 O2 8 system which con tained S2 O2 8 /OH (at a ratio of 1/2 in accordance with Eq. (5)). The second mixture was a S2 O2 solution decomposed at 70 °C. The 8 S2 O2 toxicity test mixture contained 1.25 mM S2 O2 and 8 /OH 8 2.5 mM OH. These initial concentrations were determined by adopting the initial OH concentration used ([OH] = 2.5 mM) in the individual toxicity test, because it had a lower LC50 than the S2 O2 8 , and therefore controlled the toxicity. The decomposed
2716
C. Liang, C.-W. Wang / Chemosphere 93 (2013) 2711–2716
fore, the partial fish death which occurred in the S2 O2 8 / OH = 3.36 mM/6.72 mM decomposed solution would be mainly due to the acidic pH condition.
100 Symbols overlaid
Mortality (%)
80
4. Conclusion -
2-
S2O8 /OH
60
1.25 mM/2.5 mM 1.00 mM/2.0 mM 0.75 mM/1.5 mM 0.50 mM/1.0 mM 0.25 mM/0.5 mM Blank
40
20
Symbols overlaid
0 0
24
48
72
96
Time (h) Fig. 6. Mortality of C. carpio depending on SO2 concentrations with 96 h 4 /OH h i = 1.25 mM and initial [OH] = 2.5 mM in the aqueous exposures. Initial S2 O2 8
alkaline activated persulfate test.
2 S2 O2 8 mixture contained 3.36 mM S2 O8 and 6.72 mM OH upon complete 70 °C thermal decomposition. The initial concentrations were determined based on the S2 O2 8 concentration level used in its individual toxicity test (i.e., 3.36 mM) and a OH concentration (6.72 mM) that can be neutralized by the H+ (3.36 mM 2 + S2 O2 8 2 = 6.72 mM H ) released from S2 O8 decomposition. Note 2 that in the second test both initial S2 O8 and OH concentration levels would cause 100% fish mortality if the mixed solution was not thermally decomposed. The mortality during the 96 h acute toxicity test in the alkaline activated S2 O2 system (i.e., S2 O2 8 8 /OH = 1.25 mM/2.5 mM) is shown in Fig. 6 and it can be seen that percent mortality is similar to the result of individual OH toxicity test. The LC50 values for OH at 1.43 mM or 1.34 mM in the absence or presence of S2 O2 8 , respectively, are similar, but the LC50 for S2 O2 8 appears to decrease (2.83 mM decreased to 0.67 mM) due to the presence of OH. The toxicity of the mixture of S2 O2 8 and OH seems to be sensitive to and dependent on pH. For example, in the S2 O2 mixture, 8 /OH with initial concentrations = 0.75 mM/1.5 mM, initial pH was approximately 10.5 (see Fig. SM-2a), and partial fish mortality occurred. This can be compared to their individual toxicities where h i S2 O2 at an initial concentration = 2.02 mM, and there was no 8
fish mortality, with a pH of approximately 6.5. In the individual OH test, with an initial concentration = 1.5 mM, there was partial fish toxicity, with a pH initially = 10 which decreased to 7.5 (see both Figs. 1 and 2). Moreover, when the alkaline activated S2 O2 8 system (i.e., S2 O2 8 /OH = 3.36 mM/6.72 mM) was completely decomposed, there were 3 dead fish in the none diluted solution, but no dead fish were observed in all other serial dilution concentration levels (data not shown). Hence the LC50 cannot be calculated because of the insignificant number of fish deaths in these solutions. As shown in Fig. SM-2b, the pH was within the range of 6–8 (except for the initial pH of 5.5 in the S2 O2 8 /OH = 3.36 mM/6.72 mM solution), which is within the non-toxic pH range of 6–10 (as discussed earlier). Also, the observed conductivities in this S2 O2 test (300–1600 lS cm1) were within the 8 /OH acceptable non-toxic conductivity range (i.e., less than 8000 lS cm1). Note that DO values were within the range of 3–8 mg L1 during the course of the mixture toxicity tests. There-
These acute toxicity test results represent an initial investigation into the possible effects of ISCO S2 O2 8 oxidant on aquatic ecosystems. The study is concerned with the acute toxic effects of 2 2 + S2 O2 8 , SO4 , H and OH and also the mixture of S2 O8 /OH on C. carpio. LC50 values were estimated for these ions. The bioassay results indicate that the S2 O2 is considerably more toxic than its 8 decomposition product SO2 4 due to persulfate’s oxidative chemical property. The conductivity of the aqueous solution (mainly due to the presence of Na2SO4) is unlikely to cause significant fish mortality. Additionally, very acidic or basic pH conditions may cause toxic effects. The study shows that mortality in C. carpio by exposure to 2 + S2 O2 8 can be due to the combined effect of S2 O8 and H or OH . On the basis of data generated from these comparative experiments, and the calculated LC50 values, this study recommends control of residual S2 O2 to less than 2.82 mM, control of SO2 less 8 4 than 42.72 mM, pH within the range of 6–10, and conductivity less than 8000 lS cm1, as reference toxicity values following intentional ISCO S2 O2 8 application and exposures in aquatic systems. Acknowledgements The authors acknowledge Prof. Kai-Sung Wang from Department of Public Health, Chung-Shan Medical University, Taichung, Taiwan for valuable discussion about conducting acute toxicity tests. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.chemosphere. 2013.08.078. References Chen, P.-J., Su, C.-H., Tseng, C.-Y., Tan, S.-W., Cheng, C.-H., 2011. Toxicity assessments of nanoscale zerovalent iron and its oxidation products in medaka (Oryzias latipes) fish. Mar. Pollut. Bull. 63, 339–346. FMC-Corporation, 2009. FMC Material Safety Data Sheet of Sodium Persulfate, MSDS ref. # 7775-27-1, revision #12,
Contents lists available at ScienceDirect
Chemosphere journal homepage: www.elsevier.com/locate/chemosphere
Assessing acute toxicity potential of persulfate ISCO treated water Chenju Liang ⇑, Chi-Wei Wang Department of Environmental Engineering, National Chung Hsing University, 250 Kuo-Kuang Road, Taichung 402, Taiwan
h i g h l i g h t s Evaluation of the environmental risk of exposure to persulfate is studied. Persulfate is considerably more toxic than its decomposition product sulfate. Very acidic or basic pH conditions may cause toxic effects. Water conductivity reveals minor influence on acute toxicity. The bioassay results serve as reference toxicity values following ISCO application.
a r t i c l e
i n f o
Article history: Received 19 February 2013 Received in revised form 24 August 2013 Accepted 26 August 2013 Available online 4 October 2013 Keywords: Sodium persulfate Remediation In situ chemical oxidation Aquatic toxicity Alkaline activated persulfate
a b s t r a c t Persulfate anion (S2 O2 8 ), a widely used in situ chemical oxidation agent, is increasingly applied for environmental remediation. However, limited information on environmental and toxicological effects is available for the evaluation of the environmental risk of exposure to S2 O2 8 , particularly after its application. In this study, the acute toxic effects on the common carp (Cyprinus carpio) were employed as a 2 2 model to investigate S2 O8 , sulfate ion (decomposition product of S2 O8 ), hydrogen/hydroxide ions and also the mixtures of these ion species. Acute toxicity test results showed 96 h median lethal concentra2 1 for SO2 tions (LC50) of 540 ± 23 mg L1 for S2 O2 8 and 4100 ± 110 mg L 4 . S2 O8 was considerably more . Additionally, solution pH was also an important factor influtoxic than its decomposition product SO2 4 encing toxicity, and S2 O2 8 posed reduced acute toxicity when pH was in the range of 6–10. Water conductivity up to approximately 8000 lS cm1 did not appear to significantly increase fish mortality. In 1 for S2 O2 and 23 ± 2 mg L1 the mixture toxicity test (i.e., S2 O2 8 /OH ), LC50 values of 130 ± 10 mg L 8 for OH were lower than those obtained from the individual toxicity tests and therefore exhibited higher toxicity to fish. However, upon complete decomposition of S2 O2 8 in the mixture, a reduction in acute toxicity may be expected. The results of this study revealed that it may be necessary and/or desirable to con2 addition when potential exposure to an aquatic system is a trol the residual S2 O2 8 and pH after S2 O8 concern. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Persulfate ðS2 O2 8 Þ, which is a strong oxidant with a redox potential of 2.01 V (Eq. (1)), was initially introduced as a potential in situ chemical oxidation (ISCO) agent approximately a decade ago and is now recognized as effective in degrading a variety of organic contaminants present in soil and groundwater. Persulfate reaction chemistry is complicated and involves free radical chain reactions (sulfate radical ðSO 4 Þ and/or hydroxyl radical ðHO Þ formation) upon S2 O2 activations. Chemical oxidation with 8 S2 O2 8 can be combined with the use of heat, transitional metals, hydrogen peroxide, and alkaline pH, to increase the reaction rate (Huling and Pivetz, 2006; Liang and Guo, 2012).
⇑ Corresponding author. Tel.: +886 4 22856610; fax: +886 4 22862587. E-mail address: [email protected] (C. Liang). 0045-6535/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.chemosphere.2013.08.078
2 S2 O2 8 þ 2e ! 2SO4
E ¼ 2:01 V
ð1Þ S2 O2 8
The addition of large quantities of anions to groundwater for contaminant remediation presents a concern about whether there may be toxic effects on the ecosystem. A median lethal concentration at which 50% of organisms die (i.e., LC50) in acute whole effluent toxicity tests, to estimate the lethality of a specific S2 O2 solution, was reported in a Material Safety Data Sheet to 8 be 771 mg L1 (Bluegill sunfish, 96 h exposure); 519 mg L1 (Grass shrimp, 96 h exposure); 133 mg L1 (Daphnia, 48 h exposure); and 163 mg L1 (Rainbow trout, 96 h exposure) (FMC-Corporation, 2009). However, there are still very few ecotoxicology studies that have evaluated S2 O2 ISCO application. It is essential to consider 8 the potential toxicity with respect to residual S2 O2 after ISCO 8 treatment and its associated decomposition end products, including sulfate ion and hydrogen ion. Upon S2 O2 activation, SO 8 4 accepts an electron to form SO2 (Eq. (2)) and hydrolysis of 4
2712
C. Liang, C.-W. Wang / Chemosphere 93 (2013) 2711–2716
S2 O2 8 in water would naturally result in a decrease in pH (Eqs. (3) and (4)). When the solution pH is higher than the pKa value of 1.92, 2 SO2 4 is the major end products of S2 O8 oxidation. Combining Eqs. (3) and (4) yields Eq. (5), which shows that the hydrolysis of 1 mol 2 + of S2 O2 8 would produce 2 mol of SO4 and H . 2 SO 4 þ e ! SO4
ð2Þ
S2 O2 8 þ H2 O ! 2HSO4 þ 1=2 O2
ð3Þ
þ HSO4 ! SO2 4 þH
ð4Þ
pK a ¼ 1:92
2 þ S2 O2 8 þ H2 O ! 2SO4 þ 2H þ 1=2 O2
ð5Þ
Moreover, SO 4 present in the aqueous phase could interact with water and hydroxide ions in accordance with Eqs. (6) and (7), respectively. Addition of OH (e.g., sodium hydroxide) to the 2 S2 O2 8 solution, produces an alkaline activated S2 O8 system, which has been widely used and shown effective for remediating organic contaminants on the basis of the simultaneous presence of active radical species (e.g., SO 4 and HO ) (Liang and Su, 2009). 2 þ SO 4 þ H2 O ! SO4 þ HO þ H
ð6Þ
2 SO 4 þ OH ! SO4 þ HO
ð7Þ
Groundwater quality policy is usually based on a ‘‘chemicalspecific’’ method (Tišler et al., 2004) and only a limited number of chemicals (e.g., excluding S2 O2 8 ) are prescribed by legislation for injection to the subsurface environment in Taiwan. Even though acute toxicity assessment for S2 O2 and other associated 8 ions involved during S2 O2 application might not be related to 8 any legal limits, test organisms exposed to whole S2 O2 treated 8 effluents can be a useful reference for understanding the potential ecological impacts to specific chemicals and combined effects in the S2 O2 treated water. The present study was performed to 8 determine acute toxicity of persulfate, sulfate ion, hydrogen/ hydroxide ions and also mixtures of these ion species. The common carp (Cyprinus carpio, which is one of the biological organisms allowed in Taiwan’s biological acute toxicity testing for the Water Pollution Control Act) was selected for the acute toxicity tests to determine the LC50, and variations of pH and conductivity were applied to assess the causes of toxic effects during the course of experiments. 2. Materials and methods 2.1. Chemicals All reagents were of analytical grade and used without further purification. Magnesium sulfate (MgSO47H2O, 100.4%) and acetone ((CH3)2CO, min 99.5%) were purchased from J.T. Baker (Center Valley, PA, United States of America (USA)); sodium persulfate (Na2S2O8, min 99.0%) and calcium sulfate dehydrate (CaSO42H2O, min 99.0%) were purchased from Merck (Drmstadt, Germany); sulfuric acid (min 95%), potassium chloride (min 99.5%), sodium bicarbonate (min 99.7%), sodium hydroxide (min 99%) and sodium sulfate (min 99%) were purchased from Sigma–Aldrich (St. Louis, MO, USA). The water used was purified using a Millipore reverse osmosis (RO) purification system. 2.2. Fish stock and care C. carpio were obtained from a local aquaculture farm (Changhua, Taiwan), and their body length were 2.0–3.0 cm. C. carpio (approximately 200 fish) were acclimatized for several days in
glass tanks (50 L) to assess whether they were healthy enough to be used in the experiments, i.e., the mortality rate was less than 10% within 7 d prior to the experiment. During the period of acclimatization, C. carpio were fed on commercially available fish feed once a day at least, and the environmental conditions were maintained at a 16 h photo-period and temperature 25 ± 1 °C (controlled by the air conditioner). Water quality parameters including conductivity, dissolved oxygen (DO), and pH were monitored to document the conditions during the test. The fish used in the test was selected from a tank containing healthy acclimatized carp. In addition, before carrying out the experiment, the test fish were quarantined 1 d without feeding. 2.3. Monitoring of water characteristics The pH, temperature, DO and conductivity were measured periodically using a portable pH/ORP meter (Suntex TS-100, New Taipei, Taiwan) equipped with a pH electrode and a temperature probe, a portable DO meter (Hanna HI 9146, Woonsocket, RI, USA) and a pH/conductivity meter (Eutech PC 5000, Singapore) equipped with a Fisher Scientific Accumet four-cell conductivity probe, respectively. The S2 O2 anion concentration was deter8 mined using a spectrophotometer (Hach DR/2400, Loveland, CO, USA) at 400 nm (Liang et al., 2008). 2.4. Experimental design The first phase of the experiments was designed to determine 2 + LC50 for S2 O2 8 , SO4 , H and OH . Note that sulfuric acid and sodium hydroxide were used to adjust the pH of the solutions. The concentration ranges associated with the toxicity of these reagents were unknown prior to the experiment. Therefore, several preliminary range-finding tests were conducted using groups of 5 fish, which were exposed to several widely-spaced sample dilutions to estimate the toxicity range (note that these results exhibit no statistical significance and are not reported.). The determination of the LC50 for S2 O2 8 and OH separately will distinguish the individual toxicities of the ions. Thereafter, in the second phase of the experiments, the toxicity of the simultaneous presence of both S2 O2 8 and OH was evaluated. In the bioassays of these combined ion solutions, the maximum concentration of each ion was based on the LC50 of the more toxic individual ion. For example, in deter mining the anion concentrations in the mixture of S2 O2 8 /OH , if the S2 O2 anion shows a relatively lower LC value that is 1.0 M, 50 8 the maximum OH concentration of 2.0 M would be used because + 1 mol of S2 O2 8 produces 2 mol of H which require 2 mol of OH to maintain its initial pH level upon complete decomposition of S2 O2 8 (see Eq. (5)). In order to determine the acute toxicity of the mixture of S2 O2 8 /OH upon complete decomposition, the mixture solution was heated in a water bath at 70 °C for 4 d to completely decom2 pose S2 O2 8 to SO4 and the resulting solution was cooled to room temperature (25 ± 1 °C) prior to the test. The acute toxicity tests were performed in accordance with the procedure of the Taiwan National Institute of Environmental Analysis (NIEA) method (NIEA, 2005). The toxicity testing procedure includes the preparation of five 1.5 L solutions in 2 L glass beakers, each containing a different concentration of the test ions (i.e., serial dilutions are made with: 20%, 40%, 60%, 80%, and 100% of the test solution). Experiments were carried out using 10 fish in each beaker for a period of 96 h. No food was given during the experiments. The number of fish was counted at least every 24 h and the dead fish were removed from the beaker immediately. The water quality parameters and the mortality rate were determined at the end of 0, 2, 24, 48, 72 and 96 h. Duplicate experiments were conducted and averages were reported. Control tests in RO water only were also conducted side by side.
2713
C. Liang, C.-W. Wang / Chemosphere 93 (2013) 2711–2716
2.5. Statistical analyses
(a)
100
2-
S2O8
80
40 Symbols overlaid
20 0
(b)
100 80
2-
SO4
2 + 3.1. Individual toxicity of S2 O2 8 , SO4 , H and OH 2 + The mortality of C. carpio depending on S2 O2 8 , SO4 , H and OH ions and associated variations of pH and conductivity in solutions during a 96 h period, are shown in Figs. 1 and 2, respectively. The determined LC50 values for individual solutions of persulfate and sulfate anions were 540 ± 23 mg L1 (2.82 mM) and 4100 ± 110 mg L1 (42.72 mM) for S2 O2 and SO2 , respectively (Table 1). 8 4 When comparing pH and conductivities in test solutions, pH for both tests under different ion concentrations were similar at approximately neutral pH conditions (pH 6–7), but the conductiv2 ities were much higher in the SO2 4 solution than the S2 O8 solution (e.g., 8000 lS cm1 for 39.44 mM SO2 solution versus 800 4 lS cm1 for 2.69 mM S2 O2 solution). It appears that water con8 ductivity up to approximately 8000 lS cm1 is not responsible for the fish mortality based on the results of Figs. 1b and 2b showing no dead fish in test solution containing 39.44 mM SO2 4 . Additionally, water characteristics such as pH, which remained in the range of 6 to 8 in these two experiments, can be excluded as a cause of fish mortality based on experimental results reported by Chen et al. (2011) that the variation of pH between 6 to 8 would not significantly affect the physiology of fish. Note that the initial 2 S2 O2 8 concentration used in the S2 O8 toxicity test was 3.36 mM, 2 which would produce 6.72 mM of SO4 upon complete decomposition in accordance with Eq. (5). If 3.36 mM of S2 O2 8 can be completely decomposed in solution, its decomposition end product SO2 4 (6.72 mM) would not cause lethal effects in these short-term acute toxicity experiments. However, S2 O2 8 concentration in these toxicity tests remained nearly the same (less than 2% variation) after 96 h exposure to fish. Therefore, it can be concluded that a possible direct mechanism of S2 O2 toxicity can be related to its 8 oxidative chemical property. Water characteristics such as neutral pH, conductivity (800 lS cm1) and its decomposition end product SO2 4 (39 mM) may insignificantly cause fish mortality.
49.30 mM 39.44 mM 29.58 mM 19.72 mM 9.86 mM Blank
60 40 Symbols overlaid
20 0
Symbols overlaid
(c)
100
H
80
+
0.20 mM 0.16 mM 0.12 mM 0.08 mM 0.04 mM Blank
60 40 20
Symbols overlaid
0
(d)
100
Symbols overlaid
80 3. Results and discussion
3.36 mM 2.69 mM 2.02 mM 1.34 mM 0.67 mM Blank
60
Mortality (%)
Data derived from tests were analyzed by four methods to produce LC50 values and associated 95% confidence intervals, i.e., the Spearman–Karber method, the Trimmed Spearman–Karber method, the Probit method and the Graphical method (only recommended when there are no partial mortalities). The analysis scheme is in accordance with the procedure of acute toxicity data analysis reported in Taiwan NIEA (2005), in which the flowchart for determination of the LC50 for multi-effluent concentration acute toxicity tests was adapted from the U.S. Environmental Protection Agency (USEPA, 2002) (note that the flowchart can be seen in Fig. SM-1 in Supplementary Material (SM)). In brief, when two or more of the observed proportion mortalities are between zero and one and the calculated Chi-square test for data heterogeneity is less than a tabular value at 0.05, the Probit method (a parametric statistical procedure) can be used to obtain an estimate of the LC50 and associated 95% confidence interval. However, if one or more partial mortalities are observed and the zero and 100% of mortalities happen in the lowest and highest aqueous concentrations, respectively, the Spearman–Karber method (a non-parametric statistical procedure) is used for estimating the LC50 and associated 95% confidence interval. Use of the Trimmed Spearman–Karber method (a modification of the Spearman–Karber non-parametric statistical procedure) is only appropriate when the requirements for the Probit method and the Spearman–Karber method are not met. In addition, if there are no partial mortalities, the Graphical method (a mathematical procedure) can be used, but it does not provide a confidence interval for the LC50 estimate.
OH
-
2.5 mM 2.0 mM 1.5 mM 1.0 mM 0.5 mM Blank
60 40 20
Symbols overlaid
0 0
24
48 Time (h)
72
96
2 + Fig. 1. Mortality of C. carpio depending on (a) S2 O2 8 , (b) SO4 , (c) H , and (d) OH concentrations with 96 h aqueous exposures.
Furthermore, the effects of H+ and OH on the fish death were also evaluated to determine their LC50 concentrations (see Fig. 1c and d, respectively). The LC50 for H+ and OH were determined to be 0.18 mg L1 (0.18 mM) and 24 ± 2 mg L1 (1.45 mM), respectively, and also presented in Table 1. The H+ concentration of 0.16 mM (and lower) exhibited no dead fish (corresponding to the measured pH range of 6–8). It should be noted that the calculated initial pH ranged from 3.7 ([H+] = 0.2 mM) to 4.4 ([H+] = 0.04 mM), which is different from the measured pH range and it was found that the solution pH initially increased and then gradually approached neutral during the test in the presence of fish (see Fig. 2c). However, in the initial solution where [H+] = 0.2 mM, all fish died soon after they were placed into the solution (i.e., within 2 h) and the measured pH gradually increased to around 6. This final pH of 6 is lower than the pH of around 7 which resulted from other test concentrations in the presence of live fish. Suomalainen et al. (2005) studied the buffering capacity of fish skin mucus in water and reported that fish skin mucus is an efficient buffer against decreased water pH and the pH of the skin could be higher than that of the mucus. Hence, this may explain the fish were modifying the pH and the solution pH approached neutral during the test. However, when the H+ slightly increased to 0.20 mM and pH decreased to 4, all fish died (just after 2 h exposure, Fig. 1c). As for OH, its concentration of 1.0 mM and a
2714
C. Liang, C.-W. Wang / Chemosphere 93 (2013) 2711–2716
11
11
(a)
10
pH
8
(b)
49.30 mM 39.44 mM 29.58 mM 19.72 mM 9.86 mM Blank
10
3.36 mM 2.69 mM 2.02 mM 1.34 mM 0.67 mM Blank
9
9 8 7
7
Cond. (uS/cm)
2-
SO4
2-
S2O8
6
6
1200
12000
1000
10000
800
8000
600
6000
400
4000
200
2000
0 11
0 11
(c)
10
(d)
10
9 9
pH
8 7
8
6
7
5 6
4 1200
1200 -
+
OH
H
Cond. (uS/cm)
1000 800 600
2.5 mM 2.0 mM 1.5 mM 1.0 mM 0.5 mM Blank
1000
0.20 mM 0.16 mM 0.12 mM 0.08 mM 0.04 mM Blank
800 600
400
400
200
200
0
0 0
24
48
72
96
Time (h)
0
24
48
72
96
Time (h)
2 + Fig. 2. Variations of pH and conductivity in (a) S2 O2 8 , (b) SO4 , (c) H , and (d) OH ions solutions of acute toxicity tests.
measured pH 10 resulted in no dead fish and solution pH would also gradually approach neutral in the presence of fish (Suomalainen et al., 2005). However, when OH increased to above 1.5 mM and pH was greater than 10, fish death occurred. It should be mentioned that conductivities in these tests (100–600 lS cm1) were lower than 1 those in the S2 O2 for 2.02 mM of S2 O2 8 tests (e.g., 700 lS cm 8 solution with no dead fish (see Fig. 1a). Hence the conductivity can be excluded as a major factor causing fish death, and mortality was directly related to H+ or OH species. These results indicated that in addition to the acute toxicity of specific chemical species, pH is also an important factor and solution pH beyond the range of 6–10 may cause acute toxicity effects. Furthermore, Fig. 3 illustrates some relationships between Na2S2O8 and Na2SO4. The LC50 for Na2SO4 is 42.72 mM, which is equivalent to the SO2 concentration resulting from complete 4 decomposition of 21.36 mM Na2S2O8 (Eq. (1)). In addition, as shown in Fig. 3, if the LC50 concentration of 2.82 mM S2 O2 8 is con2 verted to SO2 4 , it generates 5.64 mM SO4 , a concentration that exhibits no toxicity to fish. In the S2 O2 8 ISCO application, if the high concentration of residual S2 O2 can be completely transformed 8
into SO2 4 , a reduction of oxidant impact (e.g., acute toxicity) on the environment may be expected. Therefore, further experiments were performed to verify this assumption. The solution containing 21.36 mM of S2 O2 was thermally 8 decomposed (70 °C, Eq. (8)) to produce a 42.72 mM solution of SO2 (cooled to 25 ± 1 °C) (combining Eqs. (2) and (8)) (Liang 4 et al., 2007; Liang and Su, 2009) and the mortality of fish exposed to this decomposed solution during the 96 h acute toxicity test can be seen in Fig. 4. The LC50 for this decomposed solution of SO2 4 was then determined to be 24.46 mM, which is lower than 42.72 mM determined from the individual SO2 4 test. Moreover, the pH upon 2 decomposition of S2 O2 8 was approximately 4 for two SO4 levels of 42.72 and 34.18 mM, which was toxic and no fish survived (see Fig. 4). It should be noted that these two SO2 4 concentrations (42.72 and 34.18 mM), if tested individually, would exhibit no or partial fish death. Therefore, cause of the acute toxic effects appears to be partly due to the acidic pH (i.e., 4, see Fig. 5). Furthermore, the pH of the decomposed 25.64 mM SO2 solution was 4 initially 4.5 and gradually increased to above 6. Therefore, partial fish death was observed in the early stage and no fish death
2715
C. Liang, C.-W. Wang / Chemosphere 93 (2013) 2711–2716 Table 1 96 h LC50 values for seven acute toxicity bioassays, and analysis modes. Objective
Species
Individual toxicity
Initial conc.
a
1
mM
mM
mg L
S2 O2 8
3.36
2.82 ± 0.12
540 ± 23
SO2 4 H+ OH
49.30
42.72 ± 1.15
4100 ± 110
Probit
0.20 2.50 42.72
0.18 1.43 ± 0.11 24.46 ± 1.96
0.18 24 ± 2 2300 ± 188
Graphical Spearman–Karber Spearman–Karber
2 SO2 4 (S2 O8 decomposed)
Mixture toxicity
Analysis modea
LC50
Spearman–Karber
(S2 O2 8 /OH
1.25/2.50
0.67 ± 0.05/1.34 ± 0.09
130 ± 10/23 ± 2
Spearman–Karber
(S2 O2 8 /OH (decomposed)
3.36/6.72
n.a.
n.a.
n.a.
In accordance with the procedure established by USEPA (2002); n.a.: not available.
21.5
10 21.36 mM
21.0
2-
9
decomposed to 2SO4
8
3.5
pH
100 LC50 = 42.72 mM
80
40
60
30
40
20
2.5 2.0 1.5 1.0
7 6
Na2SO4 (mM)
3.0 LC50 = 2.82 mM
LC50 (%)
Na2S2O8 (mM)
20.5
Tolerable S2O8
5 4 18000
2-
0.5
16000
42.72 mM 34.18 mM 25.64 mM 17.08 mM 8.54 mM Blank
5.64 mM
0
0
Na2S2O8
14000
Na2SO4
Fig. 3. Relationship of Na2S2O8 and Na2SO4 concentrations and LC50 values to C. carpio bioassays.
Cond. (uS/cm)
0.0
2-
SO4 (S2O8 decomposed)
10
20
12000 10000 8000 6000 4000
100 2-
SO4 (S2O8 decomposed)
Mortality (%)
Symbols overlaid
42.72 mM 34.18 mM 25.64 mM 17.08 mM 8.54 mM Blank
80
60
2000
2-
0 0
24
48
72
96
Time (h) Fig. 5. Variations of pH and conductivity in SO2 solution of acute toxicity tests, h i 4 resulting from complete 70 °C thermal S2 O2 decomposition, initial S2 O2 = 8 8 21.36 mM.
40 Symbols overlaid
20
S2 O2 8
ambiant ð1030 CÞ or elevated ð3090 CÞ temperature
!
2SO 4
ð8Þ
0 0
24
48
72
96
Time (h) Fig. 4. Mortality of depending on SO2 h C. carpio i 4 concentrations with 96 h aqueous exposures. Initial SO2 = 42.72 mM (resulting from 21.36 mM Na2S2O8 decompo4 sition at 70 °C).
occurred thereafter when pH increased to near neutral. The conductivity range for this 25.64 mM test was approximately 1 6000 lS cm1 for the SO2 , as discussed 4 (lower than 8000 lS cm earlier) and revealed minor influence on acute toxicity. It should also be noted that the results of control tests (in the presence of RO water only) show no dead fish. DO values were within the range of 4–8 mg L1 during the course of all acute toxicity tests.
2 + 3.2. Mixture toxicity of S2 O2 8 , SO4 , H and OH
Alkaline activation is one of widely used S2 O2 8 activations for ISCO remediation of organic contaminants. In order to understand its toxicity in aquatic systems, two mixed solutions were tested. The first solution is an alkaline activated S2 O2 8 system which con tained S2 O2 8 /OH (at a ratio of 1/2 in accordance with Eq. (5)). The second mixture was a S2 O2 solution decomposed at 70 °C. The 8 S2 O2 toxicity test mixture contained 1.25 mM S2 O2 and 8 /OH 8 2.5 mM OH. These initial concentrations were determined by adopting the initial OH concentration used ([OH] = 2.5 mM) in the individual toxicity test, because it had a lower LC50 than the S2 O2 8 , and therefore controlled the toxicity. The decomposed
2716
C. Liang, C.-W. Wang / Chemosphere 93 (2013) 2711–2716
fore, the partial fish death which occurred in the S2 O2 8 / OH = 3.36 mM/6.72 mM decomposed solution would be mainly due to the acidic pH condition.
100 Symbols overlaid
Mortality (%)
80
4. Conclusion -
2-
S2O8 /OH
60
1.25 mM/2.5 mM 1.00 mM/2.0 mM 0.75 mM/1.5 mM 0.50 mM/1.0 mM 0.25 mM/0.5 mM Blank
40
20
Symbols overlaid
0 0
24
48
72
96
Time (h) Fig. 6. Mortality of C. carpio depending on SO2 concentrations with 96 h 4 /OH h i = 1.25 mM and initial [OH] = 2.5 mM in the aqueous exposures. Initial S2 O2 8
alkaline activated persulfate test.
2 S2 O2 8 mixture contained 3.36 mM S2 O8 and 6.72 mM OH upon complete 70 °C thermal decomposition. The initial concentrations were determined based on the S2 O2 8 concentration level used in its individual toxicity test (i.e., 3.36 mM) and a OH concentration (6.72 mM) that can be neutralized by the H+ (3.36 mM 2 + S2 O2 8 2 = 6.72 mM H ) released from S2 O8 decomposition. Note 2 that in the second test both initial S2 O8 and OH concentration levels would cause 100% fish mortality if the mixed solution was not thermally decomposed. The mortality during the 96 h acute toxicity test in the alkaline activated S2 O2 system (i.e., S2 O2 8 8 /OH = 1.25 mM/2.5 mM) is shown in Fig. 6 and it can be seen that percent mortality is similar to the result of individual OH toxicity test. The LC50 values for OH at 1.43 mM or 1.34 mM in the absence or presence of S2 O2 8 , respectively, are similar, but the LC50 for S2 O2 8 appears to decrease (2.83 mM decreased to 0.67 mM) due to the presence of OH. The toxicity of the mixture of S2 O2 8 and OH seems to be sensitive to and dependent on pH. For example, in the S2 O2 mixture, 8 /OH with initial concentrations = 0.75 mM/1.5 mM, initial pH was approximately 10.5 (see Fig. SM-2a), and partial fish mortality occurred. This can be compared to their individual toxicities where h i S2 O2 at an initial concentration = 2.02 mM, and there was no 8
fish mortality, with a pH of approximately 6.5. In the individual OH test, with an initial concentration = 1.5 mM, there was partial fish toxicity, with a pH initially = 10 which decreased to 7.5 (see both Figs. 1 and 2). Moreover, when the alkaline activated S2 O2 8 system (i.e., S2 O2 8 /OH = 3.36 mM/6.72 mM) was completely decomposed, there were 3 dead fish in the none diluted solution, but no dead fish were observed in all other serial dilution concentration levels (data not shown). Hence the LC50 cannot be calculated because of the insignificant number of fish deaths in these solutions. As shown in Fig. SM-2b, the pH was within the range of 6–8 (except for the initial pH of 5.5 in the S2 O2 8 /OH = 3.36 mM/6.72 mM solution), which is within the non-toxic pH range of 6–10 (as discussed earlier). Also, the observed conductivities in this S2 O2 test (300–1600 lS cm1) were within the 8 /OH acceptable non-toxic conductivity range (i.e., less than 8000 lS cm1). Note that DO values were within the range of 3–8 mg L1 during the course of the mixture toxicity tests. There-
These acute toxicity test results represent an initial investigation into the possible effects of ISCO S2 O2 8 oxidant on aquatic ecosystems. The study is concerned with the acute toxic effects of 2 2 + S2 O2 8 , SO4 , H and OH and also the mixture of S2 O8 /OH on C. carpio. LC50 values were estimated for these ions. The bioassay results indicate that the S2 O2 is considerably more toxic than its 8 decomposition product SO2 4 due to persulfate’s oxidative chemical property. The conductivity of the aqueous solution (mainly due to the presence of Na2SO4) is unlikely to cause significant fish mortality. Additionally, very acidic or basic pH conditions may cause toxic effects. The study shows that mortality in C. carpio by exposure to 2 + S2 O2 8 can be due to the combined effect of S2 O8 and H or OH . On the basis of data generated from these comparative experiments, and the calculated LC50 values, this study recommends control of residual S2 O2 to less than 2.82 mM, control of SO2 less 8 4 than 42.72 mM, pH within the range of 6–10, and conductivity less than 8000 lS cm1, as reference toxicity values following intentional ISCO S2 O2 8 application and exposures in aquatic systems. Acknowledgements The authors acknowledge Prof. Kai-Sung Wang from Department of Public Health, Chung-Shan Medical University, Taichung, Taiwan for valuable discussion about conducting acute toxicity tests. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.chemosphere. 2013.08.078. References Chen, P.-J., Su, C.-H., Tseng, C.-Y., Tan, S.-W., Cheng, C.-H., 2011. Toxicity assessments of nanoscale zerovalent iron and its oxidation products in medaka (Oryzias latipes) fish. Mar. Pollut. Bull. 63, 339–346. FMC-Corporation, 2009. FMC Material Safety Data Sheet of Sodium Persulfate, MSDS ref. # 7775-27-1, revision #12,