Antimony release from contaminated mine soils and its migration in four typical soils using lysimeter experiments

Antimony release from contaminated mine soils and its migration in four typical soils using lysimeter experiments

Ecotoxicology and Environmental Safety 133 (2016) 1–9 Contents lists available at ScienceDirect Ecotoxicology and Environmental Safety journal homep...

420KB Sizes 0 Downloads 155 Views

Ecotoxicology and Environmental Safety 133 (2016) 1–9

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety journal homepage: www.elsevier.com/locate/ecoenv

Antimony release from contaminated mine soils and its migration in four typical soils using lysimeter experiments Yu-xian Shangguan a,b, Long Zhao a, Yusheng Qin b, Hong Hou a,n, Naiming Zhang c a State Key Laboratory of Environmental Criteria and Risk Assessment, Chinese Research Academy of Environmental Sciences, No. 8 Dayangfang, Beijing 100012, China b Soil and Fertilizer Institute, Sichuan Academy of Agricultural Sciences, Chengdu 610066, China c Yunnan Agricultural University, Kunming 650201, China

art ic l e i nf o

a b s t r a c t

Article history: Received 25 March 2016 Received in revised form 19 June 2016 Accepted 20 June 2016

Antimony (Sb) can pose great risks to the environment in mining and smelting areas. The migration of Sb in contaminated mine soil was studied using lysimeter experiments. The exchangeable concentration of soil Sb decreased with artificial leaching. The concentrations of Sb retained in the subsoil layers (5–25 cm deep) were the highest for Isohumosol and Ferrosol and the lowest for Sandy soil. The Sb concentrations in soil solutions decreased with soil depth, and were adequately simulated using a logarithmic function. The Sb migration pattern in Sandy soil was markedly different from the patterns in the other soils which suggested that Sb may be transported in soil colloids. Environmental factors such as water content, soil temperature, and oxidation–reduction potential of the soil had different effects on Sb migration in Sandy soil and Primosol. The high Fe and Mn contents in Ferrosol and Isohumosol significantly decreased the mobility of Sb in these soils. The Na and Sb concentrations in soils used in the experiments positively correlated with each other (Po 0.01). The Sb concentrations in soil solutions, the Sb chemical fraction patterns, and the Sb/Na ratios decreased in the order Sandy soil 4Primosol 4 Isohumosol 4Ferrosol, and we concluded that the Sb mobility in the soils also decreased in that order. & 2016 Elsevier Inc. All rights reserved.

Keywords: Antimony Mobility Migration Lysimeter experiments Sb mining soil

1. Introduction Antimony (Sb) and its alloys have been widely used in semiconductor devices, batteries, munitions, brakes, and fire prevention materials (Okkenhaug et al., 2016; Wilson et al., 2010). The evidence suggests that human are strongly influencing the environmental geochemistry of Sb (Sharifi et al., 2016). Antimony concentrations in soil in areas where smelting is performed may reach 59.8 g kg  1, which is much higher than the maximum permissible concentration (35 mg kg  1) in soil recommended by the World Health Organization (WHO, 1996). High concentrations of Sb have been found to cause chronic toxic and carcinogenic effects in humans (Hammel et al., 2000). There are more Sb reserves in China than in any other country, and China produced about 80% of all Sb around the world in 2014 (U.S. Geological Survey, 2015). Serious Sb pollution has been found in several Chinese provinces (6946–16,389 mg kg  1), especially around the largest Sb mine in the world, in Xikuangshan, Hunan Province (Li et al., 2014; Fu et al., 2016). n

Corresponding author. E-mail address: [email protected] (H. Hou).

http://dx.doi.org/10.1016/j.ecoenv.2016.06.030 0147-6513/& 2016 Elsevier Inc. All rights reserved.

The mobilization of Sb in a soil depends on the soil characteristics and environmental factors. Synthetic metal hydroxides have been used as adsorbents of Sb in most studies in which the mobility of Sb has been investigated. Metal (hydr)oxides have been shown to adsorb Sb and affect the migration of Sb in soils (Wilson et al., 2010; Huang et al., 2012). Leleyter and Probst (1999) found that 40–75% of the Sb in soil was adsorbed to iron hydroxides. Strong evidence for preferential binding of Sb to iron oxides as inner-sphere surface complexes has been provided using extended X-ray adsorption fine structure measurements (Ilgen and Trainor, 2011; Guo et al., 2014). Vithanage et al. (2013) found direct evidence for strong interactions between Sb and Fe–O in red earth soils. A significant proportion of the Sb in soil has also been found to be retained by organic matter (Sh et al., 2012). The potential for Sb to migrate through soil at small-arms firing ranges has been found to increase as the soil pH increases and the oxidation–reduction potential (Eh) of the soil decreases (Martin et al., 2013). The Eh was recognized as an important factor controlling the Sb mobility in an acidic floodplain soil reported by Frohne et al. (2011). The solubility of Sb has been found to decrease as the clay and organic matter contents in soil increase (Tighe et al., 2013). Antimony has been found to be poorly mobile in soil, and Sb

2

Y.-x. Shangguan et al. / Ecotoxicology and Environmental Safety 133 (2016) 1–9

concentrations in soil leachates of less than 0.45 μg L  1 have been found (Hou et al., 2013). The migration of Sb through soil has been measured qualitatively in most previous studies, and there is a paucity of quantitative data on the distributions of Sb in different media. Lysimeter experiments are able to simulate field conditions more closely than the commonly used column tests. Lysimeter experiments can provide quantitative information on the mass balance of a material of interest and allow environmental factors to be monitored that cannot usually be monitored in field experiments. The arsenic (As), chromium (Cr), molybdenum (Mo), vanadium(V), cadmium (Cd), copper(Cu), nickel(Ni), zinc(Zn), cobalt(Co), and their controlling factors in soils were studied using lysimeters (Shaheen et al., 2014a, 2014b). However, only limited studies had to quantitatively investigate the migration of Sb in contaminated soil using lysimeters. We measured the migration of exogenous Sb through uncontaminated soil in an earlier study (Hou et al., 2013). Too little exogenous Sb was present in that study to allow us to observe Sb migration through the sublayer profile, but different Sb fraction and migration patterns were found in experiments using exogenous Sb and naturally aged Sb-contaminated soil. In the study presented here, we added Sb-contaminated soil to the surfaces of uncontaminated soils to simulate the release of Sb from naturally aged Sb-contaminated soil, to allow us to follow Sb migration through the soil. The Sb-contaminated soil that was used had been contaminated by Sb smelting and the soil had aged for many years. The effects of environmental factors (such as the water content and oxidation–reduction potential of the soil and the ion concentration in the soil) on Sb migration through the soil were also investigated. The main objectives were to determine (1) the solubility of Sb present in four “naturally” contaminated soils and the Sb chemical fractions in those soils, (2) the release of Sb from contaminated soil and the Sb distributions between the soil and soil solution in each soil, and (3) which factors affected Sb migration in each soil. The results of the study will help us describe the Sb migration process and provide value insight into remediation of Sb-contaminated soil.

ultisols for the US soil taxonomy) (Brady and Weil, 1996; Shi et al., 2004); Primosol (L2) from Changping, Beijing, (115°50′17″E, 40°02′ 18″N) (Primosol for Chinese soil classification and Inceptisols for the US soil taxonomy); Isohumosol (L3) from Hailun, Heilongjiang Province (126°38′00″E, 47°26′00″N) (Isohumosol for Chinese soil classification and Mollisols for the US soil taxonomy); and Sandy soil (L4) from Chaoyang, Beijing (116°26′26″E, 40°03′28″N) (Sandy soil for Chinese soil classification and Shifting Sand for the US soil taxonomy). These soils were collected and added to the lysimeters in 2008 following the method described by Hou et al. (2013). The properties of the soils and the methods used to determine them are shown in Table S1. A porous ceramic cup, to collect soil solution, was fixed at each of the depths 10, 15, 25, 35, 55, 85, and 115 cm below the soil surface. A circular opening 5 cm in diameter in the center of the bottom of each container was connected to a sump tank to allow leachate to be collected. The lysimeters were allowed to drain freely, so leachate drained out whenever a soil was saturated.

2. Materials and methods

2.4. Soil sampling and analysis

2.1. Sb-contaminated soil

After the leaching experiment was complete, samples of the contaminated soil (0–5 cm depth) and soil from 5 to 25-cm deep were collected using a soil corer (5 cm diameter). Soil sample from 5 to 25-cm deep was divided into 2-cm slices. Each sample was freeze dried and passed through a 2-mm sieve. The total Sb concentration in each sample was determined after digesting the sample in acid (Hou et al., 2005). Each analysis was conducted in triplicate, and the relative standard deviation for the total Sb concentration in each sample was o10%. The Sb concentrations in the extracts were determined using an inductively couple plasmamass spectrometry instrument (7500c; Agilent Technologies, Santa Clara, CA, USA), the limit of determination was 0.15 μg L  1. A soil reference material (GBW07309 from China) was analyzed to determine the reliability of the method. The Sb concentration that was found (0.94 70.03 mg kg  1) agreed well with the reference concentration (0.93 70.32 mg kg  1). The sequential extraction procedure described by Hou et al. (2005) was used to extract the Sb in each soil sample as eight operationally defined chemical fractions, the exchangeable, carbonate bound, metal–organic complex bound (Me-org), easily reducible metal oxide bound (ReMeOx), H2O2 extractable organic material bound (H2O2-Org), amorphous metal oxide bound (Am-MnOx), crystalline Fe oxide bound (Cr-FeOx), and residual fractions.

The Sb-contaminated soil that was used was a clay loam soil collected from the Xikuangshan mining area (27°45′28″N, 111°29′ 08″E), Hunan Province, China, in April 2013. Surface soil samples (from 0 to 20 cm deep) were collected from near a mineral waste dump. The properties of the soil are shown in Table S1. The Sb concentration in the soil was 1081 mg kg  1. The Sb-contaminated soil contained Na at a high concentration because Na is used in the smelting process. The contaminated soil was air dried, passed through a 2-mm sieve, and homogenized by mixing. 40 kg of this soil was placed on top of each lysimeter, forming a layer 5 cm deep. The contaminated soil on each lysimeter was allowed to equilibrate for approximately one month under natural conditions (outside the room) before the leaching experiments were conducted. 2.2. Lysimeters Four lysimeters (L1–L4) were used. Each lysimeter contained the contaminated soil (0–5 cm depth) and one of four representative Chinese soils (5–140 cm depth). The representative soils were: Ferrosol (L1) from Qiyang, Hunan Province (111°52′32″ E, 26°45′12″N) (Ferrosol for the Chinese soil classification and

2.3. Leaching Artificial rainfall (Table S2) was prepared for each lysimeter to match the typical composition and acidity of precipitation in the area the main soil used in that lysimeter had been collected from (Hou et al., 2013). A leaching experiment was conducted from 10 June to 11 November 2013. L1, L2, L3, and L4 received 942 mm, 448 mm, 335 mm, and 448 mm, respectively, artificial rainfall (as a spray) over the course of the experiment. Natural precipitation (302 mm) fell during the experiment. The lysimeters were left open to the environment to ensure they were at or close to their field capacities. The soil solution and leachate from each lysimeter was sampled every two weeks, and the samples were digested and analyzed to avoid the colloid bind Sb (Hu et al., 2008). The Eh, water potential, volumetric moisture content in the soil, and temperature in each lysimeter were recorded by a data logger every 30 min, and the data were aggregated to give daily means.

Y.-x. Shangguan et al. / Ecotoxicology and Environmental Safety 133 (2016) 1–9

2.5. Definition of the Sb/Na ratio

3

contributed 67.5–69.4% of the total Sb concentration in the contaminated soil before the leaching experiment. The Sb in these fractions was immobile and stable under field conditions, and the contribution of the Sb in these fractions to the total Sb concentration in the contaminated soil increased to 77.3–78.0% after the leaching experiment. The Sb in the contaminated soil was mostly in the immobile fractions, but the exchangeable Sb fraction contributed more than 2% (about 20 mg kg  1) of the total Sb concentration in the contaminated soil. The Sb concentration in the exchangeable fraction in the contaminated soil was much higher than 5 mg kg  1, which is the EU waste acceptance criterion for landfills (European Commission, 2003). These results demonstrate that Sb in the contaminated soil would pose high levels of risk to organisms.

Most of the Sb and Na in the soil solution were released from the contaminated soil. A larger proportion of the Sb than of the Na in the soil solution had been adsorbed by the soil. The ratio between the Sb and Na concentrations in the soil solution could be used to indicate the Sb migration rate using the convective–dispersive equation. The derivation of the equation was complex, and it is provided as Supplementary material. The derived equation was CSb/CNa ¼k  expm(1  R), in which k and m are constant. The CSb/CNa term will decrease as R increases and increase as the Sb migration rate increases.

3. Results and discussion 3.2. Antimony distribution in the sublayers and soil solutions 3.1. Antimony concentrations and chemical fractions in the contaminated soil

The Sb released from the contaminated soil migrated into the sublayers and became distributed downwards. The Sb concentrations in the sublayers (5–25 cm) were much higher after the leaching experiment (Fig. 2a). The Sb concentrations were highest near the contaminated soil and decreased rapidly and approached background concentrations in deeper sublayers in L1, L2, and L3. Antimony was distributed evenly through the soil between 5 and 25 cm deep in L4. The highest Sb concentrations were found at different depths in the different lysimeters after the leaching experiment was performed. More artificial rainfall was applied to L1 (942 mm) than to the other lysimeters, and the highest Sb concentrations were found at 5–9 cm and 15–19 cm deep in L1. Lysimeter of L2 received less artificial rainfall (448 mm), and the highest Sb concentrations were at 5–9 cm and 15–17 cm deep. Lysimeter of L3 received the least artificial rainfall (335 mm), and had a single maximum Sb concentration at 5–13 cm deep. Unlike in the other lysimeters, the Sb concentration in L4 increased uniformly between 5 and 25 cm deep. This was because Sb adsorbs only weakly onto sand, so it is more mobile in Sandy soils than in other soils (Griggs et al., 2011). Between 6.10% and 160% of the Sb released by the contaminated soil was adsorbed by the sublayer soil 5–25 cm deep (Table 1). These results demonstrate that Sb was adsorbed effectively by the Isohumosol and Ferrosol but less effectively by the Primosol and Sandy soil. The Sb chemical fraction in the sublayer soil profiles was very different from the Sb chemical fraction in the contaminated soil at the surface in each lysimeter (Fig. 2b). The Am-MnOx and Cr-FeOx

The Sb concentrations in the contaminated soil in the lysimeters decreased slightly after the leaching experiment was performed. The Sb concentration was 1081 mg kg  1 in the soil that was added to each lysimeter, and the Sb concentrations after the leaching experiment were 985 mg kg  1 in L1, 1016 mg kg  1 in L2, 1060 mg kg  1 in L3, and 1040 mg kg  1 in L4. The amounts by which the Sb concentrations in the contaminated soil in the lysimeters decreased are shown in Fig. 1a. Between 0.84 and 3.84 g Sb was leached from the contaminated soil in the lysimeters, and 91.1–98.1% of the Sb was retained (Table 1). The degree of leaching that occurred was high and the pH was low in L1, caused that more Sb was dissolved in the lysimeter in comparison to the other lysimeters (Lafond et al., 2013). More Sb was retained by the contaminated soil in each of the other lysimeters. The distributions of Sb in the different fractions in the contaminated soil in the lysimeters are shown in Fig. 1b. The contribution of the Sb in the exchangeable fraction to the total Sb concentration in the contaminated soil decreased from 8.27 to 10.07% before the leaching experiment to 2.39–2.88% after the leaching experiment. The exchangeable fraction included the water-soluble fraction, and was the main Sb component released from the contaminated soil. The Sb concentration in the exchangeable fraction decreased because the Sb migrated to the sublayers and was transformed into more stable forms. The Sb in the immobile fractions (Am-MnOx, Cr-FeOx, and residual factions) 1.0 L1 L2 L3 L4

120 100 80 60 40

Exchangeable Carbonate-bound Me-Org Re-MeOx H O -Org

0.8

Fractions proportion

Sb decrease concentration mg kg

-1

140

0.6

Am-MeOx Cr-FeOx Residual

0.4

0.2 20 0.0

0 L1

L2

L3

L1 L2 L3 L4 Before leaching

L4

--

L1 L2 L3 L4 After leaching

Lysimeters Fig. 1. (a) Decrease in the Sb concentration in the contaminated soil (0–5 cm deep) in each lysimeter and (b) the distribution of Sb in the different fractions in the contaminated soil (0–5 cm deep) in each lysimeter (L1 ¼contaminated soil þFerrosol, L2¼ contaminated soilþ Primosol, L3¼ contaminated soilþ Isohumosol, L4 ¼contaminated soilþ Sandy soil).

4

Y.-x. Shangguan et al. / Ecotoxicology and Environmental Safety 133 (2016) 1–9

Table 1 Mass balance of the Sb in the lysimeters.

Sb in contaminated soil before leachinga (g) Sb from rainfallb (mg) Sb released from the contaminated soilc (g) Sb adsorbed to surface soil (5–25 cm)d (g) Sb in soil solutione (mg) Sb in leachatef (mg) Plant shoots uptake of Sbg (mg) Percentage of Sb released from contaminated soilh (%) Percentage of Sb adsorbed by surface soili (5–25 cm) (%) Percentage of Sb outputj (%)

L1 (Ferrosol)

L2 (Primosol)

L3 (Isohumosol)

L4 (Sandy soil)

43.2 0.50 3.84 2.54 1.74 5.54 5.24 8.89 66.1 0.14

43.2 0.50 2.60 1.21 1.72 2.61 – 6.02 46.5 0.10

43.2 0.50 0.84 1.34 2.08 2.95 0.8 1.94 160 0.35

43.2 0.50 1.64 0.10 2.90 14.1 – 3.80 6.10 0.86

Three rainwater samples were collected, and the Sb concentrations in them were almost negligible compared with the Sb concentration in the contaminated soil. a

Sb in contaminated soil before leaching¼ Sb concentration  dry bulk density  volume. Sb from rainfall¼ average Sb concentration in rainfall  total precipitation. Sb released from the contaminated soil¼ (Sb concentration in the contaminated soil before leaching  Sb concentration in the contaminated soil after leaching)  dry bulk density  volume. d Sb adsorbed to surface soil (5–25 cm) ¼ (Sb concentration in the 5–25 cm deep soil after leaching  Sb concentration in the 5–25 cm deep soil before leaching)  dry bulk density  volume. e Sb in soil solution ¼ Sb concentration in the soil solution  soil solution volume. f Sb in leachate¼Sb concentration in the soil solution  leachate volume. g Plant shoots uptake of Sb ¼Sb concentration in the plants shoots  plant shoots weight. h Percentage of Sb released from contaminated soil¼ (Sb released from contaminated soil/Sb in contaminated soil before leaching)  100%. i Percentage Sb adsorbed by soil surface ¼ (Sb adsorbed to surface soil (5–25 cm)/Sb released from contaminated soil)  100%. j Percentage of Sb output ¼ (Sb in leachate/Sb released from contaminated soil)  100%. b c

1.0 5- 7 Exchangeable Carbonate-bound Me-Org Re-MeOx H O -Org

7- 9

Soil depth (cm)

9- 11 11-13 13-15 15-17 17-19 L1 L2 L3 L4

19-21 21-23

Fractions proportion

0.8

0.6

Am-MeOx Cr-FeOx Residual

0.4

0.2

23-25 0.0 0

10

20

30

40

50

60

70 -1

Sb increase concentration (mg kg )

L1

L2

L3

L4

Lysimeters

Fig. 2. (a) Increases in the Sb concentrations in the sublayers (5–25 cm deep) and (b) the distributions of Sb in the different fractions after the leaching experiment (L1 ¼ contaminated soil þFerrosol, L2 ¼contaminated soilþ Primosol, L3¼ contaminated soil þIsohumosol, L4¼ contaminated soil þSandy soil).

fractions accounted for more than 50% of the total Sb in the Ferrosol and Isohumosol after the leaching experiment. More Sb was in the mobile fractions in the Primosol and Sandy soil than in the Ferrosol and Isohumosol. In particular, the carbonate-bound fraction contained 31.2% and 28.7% of the Sb in the Sandy soil and Primosol, respectively. The Sb chemical fraction in the four soils was similar to the Sb chemical fraction in soil determined by Griggs et al. (2011), in that the Sb was mostly in the unstable fractions in the Sandy soil but in the stable fractions in the clay soils. The proportion of the Sb in the immobile fractions (Am-MnOx, Cr-FeOx, and residual fraction) after the leaching experiment decreased in the order Ferrosol4 Isohumosol4Primosol4Sandy soil. It has been previously reported (Murciego et al., 2007) that Sb is strongly adsorbed by clays and Fe/ Mn hydroxides. The Ferrosol contained more clays and Fe/Mn hydroxides than did the other soils (Table S1), so most of the Sb was retained in the surface layer (5–25 cm deep) in L1, and this agreed with the results of other studies (Hou et al., 2013; Yang et al., 2015). Similar to clays and Fe/Mn hydroxides, soil organic matter also adsorbed Sb in the Isohumosol. Relatively little artificial rainfall was applied to the Isohumosol, and this may cause more Sb to be retained in L3 than in the other lysimeters.

The Sb concentration in the soil solution decreased with the soil depth in each lysimeter during the leaching experiment (Fig. 3). The Sb concentrations in the soil solutions decreased from 327 μg L  1 to 6.16 μg L  1 in L1, from 268 μg L  1 to 33 μg L  1 in L2, from 192 μg L  1 to 6.6 μg L  1 at in L3, and from 609 μg L  1 to 6.4 μg L  1 in L4. The mean Sb concentrations between 35 and 115 cm deep were, in decreasing order, 39.9 μg L  1 in L2, 28.2 μg L  1 in L3, 21.4 μg L  1 in L1, and 7.28 μg L  1 in L4. The mean Sb concentrations in the soil solutions between 35 and 115 cm deep at different times during the experiment are shown in Fig. 4a. The Sb concentration in the leachate occurred at different times in different lysimeters, the earliest in L2 (16 July), the next in L3 (17 August), and the last in L1 (15 September). No clear peak concentration occurred in L4 because the Sb migrated through the Sandy soil rapidly, in soluble metal–colloid complexes in the soil solution. The maximum concentration in L2 decreased on 22 July because heavy rain (72 mm on 15–16 July) added a large amount of water to the soil solution and diluted the Sb. The Sb remained in the surface soil in L3 and L1 and was little affected by the heavy rain. There were two maxima in the Sb concentrations in the

Y.-x. Shangguan et al. / Ecotoxicology and Environmental Safety 133 (2016) 1–9

Soil depth (cm)

L1 10

10

15

15

25

25

35

35

55

55

85

85

115

115

35-115

35-115

0

100

200

300

400

500

600

5

L2

0

100

200

300

400

500

L3 10

10

15

15

25

25

35

35

55

55

85

85

115

115

35-115

35-115

0

100

200

300

400

500

600

600

L4

0

150

300

450

600

750

900

1050

-1

Sb concentration in the soil solution (ug L ) Fig. 3. Average Sb concentrations in the soil solution at different depths during the leaching experiment (L1 ¼contaminated soil þFerrosol, L2¼ contaminated soilþ Primosol, L3 ¼contaminated soilþ Isohumosol, L4 ¼contaminated soilþ Sandy soil).

L1 L2 L3 L4

140

120

300 250

-1

100 200 80

60

40 30

40

20 20

10

16-Oct

11-Nov

12-Oct

5 -Oct

10-Oct

23-Sep

19-Sep

8 -Sep

15-Sep

6 -Sep

29-Aug

25-Aug

22-Jul

7 -Aug

11-Nov

16-Ocp

2-Ocp

15-Sep

2-Sep

17-Aug

2-Aug

22-Jul

16-Jul

7-Jul

14-Jul

0

0

3 -Jul

Sb concentration ug L

L1 L2 L3 L4

350

Data Fig. 4. (a) Mean Sb concentrations in the soil solution from 35 to 115 cm deep during the experiment and (b) Sb concentrations in the soil leachates (L1 ¼ contaminated soilþ Ferrosol, L2 ¼contaminated soilþ Primosol, L3 ¼ contaminated soilþ Isohumosol, L4¼ contaminated soil þSandy soil).

leachate (Fig. 4b). The Sb might have diffused to the sublayers naturally during the equilibration process before the leaching experiment was started, leading to higher Sb concentrations being found than expected at the beginning of the experiment. The second maximum occurred during the leaching experiment. The maximum Sb concentrations in the leachates decreased in the order L4 (330 μg L  1), L2 (183 μg L  1), L3 (35.9 μg L  1), then L1 (29.9 μg L  1). These maximum concentrations were much higher than we found in a previous study (Hou et al., 2013), and were

similar to Sb concentrations found near arms firing ranges (Martin et al., 2013). The Sb concentrations were much higher than the quality standards for general groundwater resources (24 μg L  1) and potable groundwater resources (6 μg L  1) (Illinois Administrative, 2013). This suggests that Sb would pose high levels of risk if it was present in any of the four soils in the open environment. The mean Sb concentration was significantly higher (P¼ 0.03) in the acid-digested unfiltered leachate (28.3 75.02 μg L  1) than the filtered leachate (6.28 71.11 μg L  1) from L4. The total amounts of

6

Y.-x. Shangguan et al. / Ecotoxicology and Environmental Safety 133 (2016) 1–9

Sb in the L1, L2, L3, and L4 leachates were 5.54, 2.61, 2.95, and 14.06 mg, respectively, and these amounts accounted for 0.10– 1.05% of the Sb released from the contaminated soil. Metals mainly migrate through soil through two processes, (1) repeated adsorption and desorption (leading to slow stepwise movement) and (2) movement as soluble metal–colloid complexes in the soil solution (leading to rapid movement) (Hou et al., 2005). The Sb concentrations in the filtered and total (digested) L4 leachate were different. This results indicated that the Sb migrated in a markedly different way in the Sandy soil in comparison with the other soils, and Sb might associate with soil colloids in the Sandy soil as report by Zhu et al. (2014) and Klitzke et al. (2012), and our former research. In the other three soils, Sb was found at high concentrations in the surface soil, and the concentration decreased as the depth increased. The clay, Fe/Mn hydroxides, and organic matter in these three soils would have adsorbed Sb and affected its migration through the soil. This led us to conclude that Sb migrated through these three soils by repeatedly being adsorbed and desorbed (i.e., in slow stepwise movements). Combining the soil solution and leachate results, the Sb mobility in the soils was ranked in the decreasing order Sandy soil4Primosol4Isohumosol4Ferrosol. This order agrees with the results of a study by Hou et al. (2013). 3.3. Pearson correlations between the Sb concentration in the soil solution and relevant parameters Antimony migration through soil is affected by environmental factors such as the soil solution composition, soil water content, temperature, and Eh (Wilson et al., 2010). The Pearson correlation coefficients for the relationships between the Sb concentrations and relevant parameters were calculated (Table 2). The Sb concentration in the soil solution positively correlated with the Na concentration in all four soils (Po 0.01). It has been found that Sb reacts with Na2S to form Na3SbS3 (Anderson, 2012). The dissolution of Na3SbS3 will affect the Sb concentration in the soil solution. The Sb concentrations in L2 and L4 most strongly correlated with the environmental factors and the Ca concentration, while the Sb concentrations in L1 and L3 most strongly correlated with the Fe and Mn concentrations. Calcium in soil can coprecipitate with Sb to form Ca2(SbO3), which is extremely insoluble, and the dissolution of romeite (Ca[Sbx(OH)y]Z) will control the leaching of antimonate at pH 8–11 (Cornelis et al., 2012). The Sandy soil and the Primosol had pH values between 8.2 and 9.1. The Fe and Mn (which can adsorb Sb and retard its migration (Xi et al., 2013)) concentrations were higher in the Ferrosol than the other soils. The organic matter content and Fe concentration were higher in the Isohumosol than in the Primosol or Sandy soil. The Sb negatively correlated with the Fe and Mn in the four lysimeters (Table 2), and this may attributed to strongly adsorbed

ability of Fe and Mn compounds. Metal hydroxides may be the main compounds that adsorb Sb and retard its migration through soil (Wilson et al., 2010; Huang et al., 2012). Sb can form innersphere surface complexes with Fe oxides, and this could explain the effect of metal hydroxides on the migration of Sb, especially because of the Fe concentrations in the Ferrosol and Isohumosol. In addition to the strong influence of the soil solution composition, the soil temperature, water content, and Eh can also significantly affect Sb migration through soil. The Sb concentrations in the Primosol and Sandy soil were significantly affected by the water content (P o0.05). The Sandy soil and Primosol had the highest sand contents, making them very porous and easy for water to pass through them. However, the Sandy soil had the lowest heat capacity, causing more water to evaporate from it than from the other three soils (Sakai et al., 2011). The more water evaporates the lower will be the tendency for Sb to migrate downward from the surface contaminated soil. The soil temperature negatively correlated with the Sb concentration in the Sandy soil (P¼ 0.01). The soil Eh played an important role in Sb migration in the Primosol. The soil Eh was affected by the air permeability of the soil and the metal oxide content (Frohne et al., 2011; Messias et al., 2013). The Sandy soil was more permeable to air than were the other three soils, and the Ferrosol and Isohumosol had the highest iron oxide contents. The Ferrosol, Isohumosol, and Sandy soil had higher average Eh values during the whole experiment (285, 259, and 312 mV, respectively) than those in the Primosol. The low Primosol Eh (248 mV, average Eh values during the whole experiment) was associated with Sb migrating more efficiently through the Primosol than the other soils. A low Eh can facilitate Sb migration through soil. The Sb was originally bound to Fe and Mn oxyhydroxides, and a decrease in the Eh could cause Fe and Mn oxyhydroxides to dissolve, then associated substances (including organic matter) to be released (Grybos et al., 2007). Shaheen et al. (2014a) indicated that the direct metal reduction, the release of metal bound to reductively dissolved Fe- and Mn-oxides and OM, as well as the indirect changes of pH caused by the changes of Eh were probably the three reasons for the increasing concentrations of heavy metals in two groundwater lysimeters during the different flooding durations. The Sb has the different geochemical behaviors with the other heavy metals, which exist as an anion. Therefore, the direct metal reduction of Sb(Ⅴ) to Sb(Ⅲ) have the low mobility in the Sb migration. The decrease of pH caused by changes of Eh probably not the factor influences the Sb migration. The release of metal bound to reductively dissolved Fe- and Mnoxides and OM may be the most important factors influencing the Sb migration and consistent with our research. This could explain why the Sb concentration in the soil solution increased as the Eh decreased. The redox potential is also a crucial factor for

Table 2 Pearson correlation coefficients for the relationships between the Sb concentrations in the soil solutions and relevant environmental parameters for each lysimeter. Soil a

W Tb E hc Na Ca Fe Al Mn a

L1 (Ferrosol)  0.023 0.122  0.074 0.718**  0.018  0.215*  0.163  0.242*

Water content. Soil temperature. c Soil redox potential. * Significant at the 0.05 level. ** Significant at the 0.01 level. b

P 0.848 0.315 0.542 0.001 0.858 0.033 0.108 0.016

L2 (Primosol) *

0.287  0.041  0.345** 0.320**  0.224*  0.130  0.021  0.173

P 0.016 0.735 0.003 0.001 0.037 0.201 0.834 0.089

L3 (Isohumosol) 0.064  0.016 0.046 0.366**  0.149  0.322**  0.058  0.047

P 0.600 0.894 0.708 0.001 0.169 0.001 0.573 0.648

L4 (Sandy soil) **

0.388  0.307**  0.035 0.771**  0.358**  0.077 0.198  0.079

P 0.001 0.010 0.772 0.001 0.001 0.451 0.050 0.442

Y.-x. Shangguan et al. / Ecotoxicology and Environmental Safety 133 (2016) 1–9

0

L1

0

20

20

40

40

60

60

80

80

100

L1

100 Na = 1056 185×ln(D 4.17) R =0.624, n=98, P<0.001

120 140

0

500

1000

Sb=0.257 0.054×ln(D 4.74) R =0.621, n=98, P<0.001

120 140

1500

0

L2

20

40

40

60

60

80

80

100

0.0

0.3

0.9

1.2

1.5

L2

100 Na = 677 71.3×ln(D 4.98)

120 Soil depth (cm)

0.6

0

20

140

7

Sb=0.153 0.028×ln(D 4.98)

120

R =0.499, n=98, P<0.001

0

500

1000

140

1500

0

R =0.598, n=98, P <0.001

0.0

0.3

0.6

0.9

1.2

1.5

0

L3

L3 20

20

40

40

60

60

80

80

100

100

120

Sb=0.105 0.019×ln(D 4.99)

120

Na = 1214 199×ln(D 4.39)

R =0.104, n=98, P =0.001

R =0.346, n=98, P<0.001

140

140 0

500

1000

1500

0.0

0

L4

20

20

40

40

60

60

80

0.3

0.6

0.9

1.2

1.5

0

L4

80 Na = 865 171×ln(D 4.92)

100

Sb=0.285 0.063×ln(D 4.99)

100

R =0.699, n=98, P<0.001

120

120

Measured values Simulate cure

140 0

500

1000

1500

R =0.513, n=98, P <0.001

Measured values Simulate cure

140

2000 -1

Na concentration in the soil solution (mg L )

0.0

0.3

0.6

0.9

1.2

1.5 -1

Sb concentration in the soil solution (mg L )

Fig. 5. Sodium and antimony distributions in the soil solutions in the lysimeters (L1 ¼contaminated soilþ Ferrosol, L2 ¼contaminated soilþ Primosol, L3¼ contaminated soilþ Isohumosol, L4¼ contaminated soil þSandy soil).

8

Y.-x. Shangguan et al. / Ecotoxicology and Environmental Safety 133 (2016) 1–9

controlling the methylation of Sb. A series of sequential redox reactions when the redox status of the soil Eh changes were catalyzed by microorganisms, which will influence the methylation of Sb (Du Laing et al., 2009). As Wei et al. (2015) reported in the same contaminated soils, the inorganic Sb species were the major Sb species in the majority of the soil and plant extracts, while the methylation of Sb was observed in a few of the plant extracts at certain sites, primarily in the leaves and stems. Thus, the Sb methylation is an essential matter that we will discuss the contents in the future. Other factors such as DOC and soil pH will also influence the Sb behavior, the high content of the DOC will combine with the Sb and facility the Sb with long distance transport. And the soil Sb adsorption ability will also decrease with the soil pH by the effects of the soil point of zero charge. We concluded that Sb migration through the Primosol and Sandy soil was mainly influenced by the dissolution of Sb in the contaminated soil and by environmental factors. The Ca concentration was the only soil property that negatively correlated with the Sb concentrations in the soil solutions in the Isohumosol and Sandy soil. The low metallic oxide content in the Primosol and Sandy soil may have caused the environmental factors to have more effect on Sb migration in these soils than the other soils. The migration of Sb in the Ferrosol and Isohumosol was mainly influenced by the dissolution of Sb in the contaminated soil and the metallic oxide content. 3.4. Comparing Sb migration rates using the Sb/Na ratio The contaminated soil contained high Na concentrations (Table S1), caused by Sb mining and smelting processes. The Na and Sb might form Na3SbS3 in the contaminated soil as described by Anderson (2012) and our former investigation in the mining area that the Na3SbS3 may the main products during the mining process. The Na and Sb would have been able to dissolve in the soil solution and migrate with water down the soil profile. The Sb concentrations significantly correlated (P o0.01) with the Na concentrations in the soil solutions in the lysimeters (Table 2). To summarize, the solute distributions in the soils were influenced by migration in water and adsorption to the soil. It was difficult to identify the separate effects of diffusion in water and adsorption to soil because a different leaching volume was used in each lysimeter. However, the Sb/Na ratios (Supplementary material) could indicate the ability of each soil to adsorb Sb. The distributions of Sb and Na in the soil solutions in the soil profiles are shown in Fig. 5. The migration of the solutes in the soil was influenced by the convection and dispersion of the soil water. The solute concentration will also have been influenced by adsorption to soil. Sodium migration will not have been strongly affected by the retarding effect of adsorption to soil but would have been controlled by the migration of the soil water (Frohnert et al., 2014), whereas Sb migration would have been affected by adsorption to the soil (Xi et al., 2013). The changes in the Sb and Na concentrations in the soil solutions correlated with each other. The Na concentrations were quickly decreased (logarithmically) with the soil depth. The R2 values for the logarithmic relationships between the Sb and Na concentrations and the soil depth in the different soils were 0.104– 0.699. The poorest correlation was for L3 and L2, probably because of the heterogeneous vertical distributions of the properties (Table S1). The Na concentration decreased sharply in L4 because of its high sand content. The other soils had much higher clay and silt contents, the mean pore diameters were much lower (Table S1) and make the water diffused through slowly. The Sb concentration decreased more sharply than the Na concentration with depth because the adsorption of Sb by soil will have retarded Sb movement. The vertical Sb and Na distributions

were most similar in the Sandy soil. There were more marked differences between the vertical Sb and Na distributions in the other three soils because they had much higher adsorption capacities for Sb than did the Sandy soil. We could identify the separate effects of convection dispersion and adsorption onto soil more effectively using the Sb/Na ratio than from the Sb and Na migration data. The mean Sb/Na ratios were 0.160  10  3, 0.179  10  3, 0.170  10  3, and 0.220  10  3 for L1, L2, L3, and L4, respectively. A smaller Sb/Na ratio indicated that Sb was less able to migrate through the soil and easy to adsorb by the soils. The ability of Sb to migrate through the soils decreased in the order Sandy soil4Primosol4Isohumosol4Ferrosol. This order is consistent with the order for the proportions of stable Sb species in the soils and the order for the maximum Sb concentrations in the soil leachates. The Sb was mostly in immobile forms in the Ferrosol and Isohumosol, but a larger proportion of the Sb was mobile in the Primosol and Sandy soil. The Sb in the Ferrosol and Isohumosol was readily adsorbed by the soil, so it migrated slowly, but the Sb in the Primosol and Sandy soil was poorly adsorbed by the soil, so it was mobilized readily.

4. Conclusions Antimony in mining-contaminated soil was found to pose risks to the environment. After five months of leaching, the contaminated soil in the lysimeters had exchangeable Sb concentrations much higher than the EU waste acceptance criterion for landfills. The maximum Sb concentrations in the leachates were higher than the groundwater quality standards for general groundwater resources and potable groundwater resources. The Sb migration patterns were different in each of the four soils that were tested. The dissolution of Sb in the contaminated soil controlled Sb migration through the different soils. The soils metal oxide contents (Fe and Mn) and some factors (Eh and soil water content) also affected Sb migration. The soils with high metal oxide contents (Fe and Mn) retarded the migration of Sb, but Sb migration in the soils with low metal oxide contents was strongly affected by factors such as the Eh and soil water content. The adsorption ability of Sb in the four soils was decreased in the following order as: Ferrosol4Isohumosol 4Primosol 4Sandy soil. The amount of Sb used annually is increasing, so the amount of Sb released into the environment may also be increasing. Techniques for stabilizing and remediating soil contaminated with Sb will therefore need to be developed, and a better understanding of the behavior of Sb in soils will be required. Ameliorating and avoiding Sb pollution will be an important challenge for the near future, to allow potential risks posed by Sb entering groundwater (which may be used as drinking water) to be minimized.

Acknowledgments This study were funded by the National Natural Science Foundation of China, China (41271338), Science and Technology Innovation Talent Plan Innovation Team of Yunnan Province, China (2015HC018).

Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.ecoenv.2016.06. 030.

Y.-x. Shangguan et al. / Ecotoxicology and Environmental Safety 133 (2016) 1–9

References Anderson, C.G., 2012. The metallurgy of antimony. Chem. Erde – Geochem. 72, 3–8. Brady, N.C., Weil, R.R., 1996. The Nature and Properties of Soils. Prentice-Hall Inc, Upper Saddle River, NJ. Cornelis, G., Van Gerven, T., Vandecasteele, C., 2012. Antimony leaching from MSWI bottom ash: modelling of the effect of pH and carbonation. Waste Manag. 32, 278–286. Du Laing, G., Rinklebe, J., Vandecasteele, Meers, B.E., Tack, F.M.G., 2009. Trace metal behaviour in estuarine and riverine floodplain soils and sediments: A review. Sci. Total Environ. 407, 3972–3985. European Commission., 2003. 2003/33/EC. Frohne, T., Rinklebe, J., Diaz-Bone, R.A., Du Laing, G., 2011. Controlled variation of redox conditions in a floodplain soil: impact on metal mobilization and biomethylation of arsenic and antimony. Geoderma 160, 414–424. Frohnert, A., Apelt, S., Klitzke, S., Chorus, I., Szewzyk, R., Selinka, H.C., 2014. Transport and removal of viruses in saturated sand columns under oxic and anoxic conditions – potential implications for groundwater protection. Int. J. Hyg. Environ. Health 217, 861–870. Fu, Z.Y., Wu, F.C., Mo, C.L., Deng, J.J., Meng, W., Giesy, J.P., 2016. Comparison of arsenic and antimony biogeochemical behavior in water, soil and tailings from Xikuangshan, China. Sci. Total Environ. 539, 97–104. Griggs, C.S., Martin, W.A., Larson, S.L., O’Connnor, G., Fabian, G., Zynda, G., Mackie, D., 2011. The effect of phosphate application on the mobility of antimony in firing range soils. Sci. Total Environ. 409, 2397–2403. Grybos, M., Davranche, M., Gruau, G., Petitjean, P., 2007. Is trace metal release in wetland soils controlled by organic matter mobility or Fe-oxyhydroxides reduction? J. Colloid Interface Sci. 314, 490–501. Guo, X.J., Wu, Z.J., He, M.C., Meng, X.G., Jin, X., Qiu, N., Zhang, J., 2014. Adsorption of antimony onto iron oxyhydroxides: adsorption behavior and surface structure. J. Hazard. Mater. 276, 339–345. Hammel, W., Debus, R., Steubing, L., 2000. Mobility of antimony in soil and its availability to plants. Chemosphere 41, 1791–1798. Hou, H., Takamatsu, T., Koshikawa, M.K., Hosomi, M., 2005. Migration of silver, indium, tin, antimony, and bismuth and variations in their chemical fractions on addition to uncontaminated soils. Soil Sci. 170, 624–639. Hou, H., Yao, N., Li, J.N., Wei, Y., Zhao, L., Zhang, J., Li, F.S., 2013. Migration and leaching risk of extraneous antimony in three representative soils of China: lysimeter and batch experiments. Chemosphere 93, 1980–1988. Hu, S.P., Chen, X.C., Shi, J.Y., Chen, Y.X., Li, Q., 2008. Particle-facilitated lead and arsenic transport in abandoned mine sites soil influenced by simulated acid rain. Chemosphere 71, 2091–2097. Huang, Y., Chen, Z., Liu, W., 2012. Influence of iron plaque and cultivars on antimony uptake by and translocation in rice (Oryza sativa L.) seedlings exposed to Sb (III) or Sb (V). Plant Soil 352, 41–49. Ilgen, A.G., Trainor, T.P., 2011. Sb (III) and Sb (V) sorption onto Al-rich phases: hydrous Al oxide and the clay minerals kaolinite KGa-1b and oxidized and reduced nontronite NAu-1. Environ. Sci. Technol. 46, 843–851. Illinois Administrative., 2013. 2013/35/620. Klitzke, S., Lang, F., Kirby, J., Lombi, E., Hamon, R., 2012. Lead, antimony and arsenic in dissolved and colloidal fractions from an amended shooting-range soil as characterised by multi-stage tangential ultrafiltration and centrifugation. Environ. Chem. 9, 462–473. Lafond, S., Blais, J.F., Martel, R., Mercier, G., 2013. Chemical leaching of antimony and other metals from small arms shooting range soil. Water Air Soil Pollut. 224, 1–15. Leleyter, L., Probst, J.L., 1999. A new sequential extraction procedure for the speciation of particulate trace elements in river sediments. Int. J. Environ. Anal. Chem. 73, 109–128. Li, J.N., Wei, Y., Zhao, L., Zhang, J., Shangguan, Y.X., Li, F.S., Hou, H., 2014. Bioaccessibility of antimony and arsenic in highly polluted soils of the mine area and

9

health risk assessment associated with oral ingestion exposure. Ecotoxicol. Environ. Saf. 110, 308–315. Martin, W.A., Lee, L.S., Schwab, P., 2013. Antimony migration trends from a small arms firing range compared to lead, copper, and zinc. Sci. Total Environ. 463, 222–228. Messias, I.A.M., Couto, E.G., Amorim, R.S.S., Johnson, M.S., Pinto, O.B., 2013. Monitoring the soil redox potential and its interaction with associated variables in the Northern Pantanal. Rev. Bras. Cienc. Solo 37, 632–639. Murciego, A.M., Sanchez, A.G., González, M.R., Gil, E.P., Gordillo, C.T., Fernández, J.C., Triguero, T.B., 2007. Antimony distribution and mobility in topsoils and plants (Cytisus striatus, Cistus ladanifer and Dittrichia viscosa) from polluted Sbmining areas in Extremadura (Spain). Environ. Pollut. 145, 15–21. Okkenhaug, G., Gebhardt, K.A.G., Amstaetter, K., Bue, H.L., Herzel, H., Mariussen, E., Almas, A.R., Cornelissen, G., Breedveld, G.D., Rasmussen, G., 2016. Antimony (Sb) and lead (Pb) in contaminated shooting range soils: Sb and Pb mobility and immobilization by iron based sorbents, a field study. J. Hazard. Mater. 307, 336–343. Sakai, M., Jones, S.B., Tuller, M., 2011. Numerical evaluation of subsurface soil water evaporation derived from sensible heat balance. Water Resour. Res., 4. http://dx.doi.org/10.1029/2010WR009866. Sh, T., Liu, C.Q., Wang, L., 2012. Antimony coordination to humic acid: nuclear magnetic resonance and X-ray absorption fine structure spectroscopy study. Microchem. J. 103, 68–73. Shaheen, S.M., Rinklebe, J., Rupp, H., Meissner, R., 2014a. Lysimeter trials to assess the impact of different flood-dry-cycles on the dynamics of pore water concentrations of As, Cr, Mo and V in a contaminated floodplain soil. Geoderma 228, 5–13. Shaheen, S.M., Rinklebe, J., Rupp, H., Meissner, R., 2014b. Temporal dynamics of pore water concentrations of Cd, Co, Cu, Ni, and Zn and their controlling factors in a contaminated floodplain soil assessed by undisturbed groundwater lysimeters. Environ. Pollut. 191, 223–231. Sharifi, R., Moore, F., Keshavarzi, B., 2016. Mobility and chemical fate of arsenic and antimony in water and sediments of Sarouq River catchment, Takab geothermal field, northwest Iran. J. Environ. Manag. 170, 136–144. Shi, X.Z., Yu, D.S., Warner, E.D., Pan, X.Z., Petersen, G.W., Gong, Z.G., Weindorf, D.C., 2004. Soil database of 1:1,000,000 digital soil survey and reference system of the Chinese genetic soil classification system. Soil Surv. Horiz. (In Chinese) Tighe, M., Lockwood, P.V., Ashley, P.M., Murison, R.D., Wilson, S.C., 2013. The availability and mobility of arsenic and antimony in an acid sulfate soil pasture system. Sci. Total Environ. 463, 151–160. U.S. Geological Survey, 2015. Mineral Commodity Summaries 2015. U.S. Geological Survey, Washington, p. 19. Vithanage, M., Rajapaksha, A.U., Dou, X., Bolan, N.S., Yang, J.E., Ok, Y.S., 2013. Surface complexation modeling and spectroscopic evidence of antimony adsorption on iron-oxide-rich red earth soils. J. Colloid Interface Sci. 406, 217–224. Wei, C.Y., Ge, Z.F., Chu, W.S., Feng, R.W., 2015. Speciation of antimony and arsenic in the soils and plants in an old antimony mine. Environ. Exp. Bot. 109, 31–39. WHO, 1996. 2nd ed.Guidelines for Drinking-Water Quality 2. World Health Organization, Geneva, p. 937. Wilson, S.C., Lockwood, P.V., Ashley, P.M., Tighe, M., 2010. The chemistry and behaviour of antimony in the soil environment with comparisons to arsenic: a critical review. Environ. Pollut. 158, 1169–1181. Xi, J., He, M., Wang, K., Zhang, G., 2013. Adsorption of antimony (III) on goethite in the presence of competitive anions. J. Geochem. Explor. 132, 201–208. Yang, H., He, M., Wang, X., 2015. Concentration and speciation of antimony and arsenic in soil profiles around the world's largest antimony metallurgical area in China. Environ. Geochem. Health 37, 21–33. Zhu, Y., Ma, L.Q., Dong, X., Harris, W.G., Bonzongo, J.C., Han, F., 2014. Ionic strength reduction and flow interruption enhanced colloid-facilitated Hg transport in contaminated soils. J. Hazard. Mater. 264, 286–292.