Sequential soil washing techniques using hydrochloric acid and sodium hydroxide for remediating arsenic-contaminated soils in abandoned iron-ore mines

Chemosphere 66 (2007) 8–17 www.elsevier.com/locate/chemosphere

Sequential soil washing techniques using hydrochloric acid and sodium hydroxide for remediating arsenic-contaminated soils in abandoned iron-ore mines Min Jang a

a,*

, Jung Sung Hwang b, Sang Il Choi

b

Department of Civil and Environmental Engineering, The Pennsylvania State University, 205 Sackett Building, State College, Pennsylvania 16802, USA b Department of Environmental Engineering, Kwangwoon University, 447-1 Wolgye-Dong Nowon-Gu, Seoul, Republic of Korea Received 7 August 2005; received in revised form 19 May 2006; accepted 24 May 2006 Available online 10 July 2006

Abstract Sequential washing techniques using single or dual agents [sodium hydroxide (NaOH) and hydrochloric acid (HCl) solutions] were applied to arsenic-contaminated soils in an abandoned iron-ore mine area. We investigated the best remediation strategies to maximize arsenic removal efficiency for both soils and arsenic-containing washing solution through conducting a series of batch experiments. Based on the results of a sequential extraction procedure, most arsenic prevails in Fe–As precipitates or coprecipitates, and iron exists mostly in the crystalline forms of iron oxide. Soil washing by use of a single agent was not effective in remediating arsenic-contaminated soils because arsenic extractions determined by the Korean standard test (KST) methods for washed soils were not lower than 6 mg kg1 in all experimental conditions. The results of X-ray diffraction (XRD) indicated that iron-ore fines produced mobile colloids through coagulation and flocculation in water contacting the soils, containing dissolved arsenic and fine particles of ferric arsenate-coprecipitated silicate. The first washing step using 0.2 M HCl was mostly effective in increasing the cationic hydrolysis of amorphous ferrihydrite, inducing high removal of arsenic. Thus, the removal step of arsenic-containing flocs can lower arsenic extractions (KST methods) of washed soils. Among several washing trials, alternative sequential washing using 0.2 M HCl followed by 1 M HCl (second step) and 1 M NaOH solution (third step) showed reliable and lower values of arsenic extractions (KST methods) of washed soils. This washing method can satisfy the arsenic regulation of washed soil for reuse or safe disposal application. The kinetic data of washing tests revealed that dissolved arsenic was easily readsorbed into remaining soils at a low pH. This result might have occurred due to dominant species of positively charged crystalline iron oxides characterized through the sequential extraction procedure. However, alkaline extraction using NaOH was effective in removing arsenic readsorbed onto the surface of crystalline minerals. This is because of the ligand displacement reaction of hydroxyl ions with arsenic species and high pH conditions that can prevent readsorption of arsenic. Ó 2006 Elsevier Ltd. All rights reserved. Keywords: Arsenic; Sequential soil washing; Iron ore; Sodium hydroxide; Hydrochloric acid

1. Introduction About 894 of 900 abandoned metal mines are creating significant environmental problems in Korea. Mining

*

Corresponding author. Tel.: +1 814 865 9425; fax: +1 814 863 7304. E-mail address: [email protected] (M. Jang).

0045-6535/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2006.05.056

waste and acid mine drainage produced from these abandoned metal mines have released several toxic metalloids or heavy-metals into ground-water, surface-water, and geological environments because of their solubility and mobility (Mulligan et al., 2001). There are two categories of processes that mostly control arsenic mobilities: (1) adsorption and desorption, and (2) solid-phase precipitation and dissolution. These processes can be mainly controlled by

M. Jang et al. / Chemosphere 66 (2007) 8–17

pH, redox reactions, competing anions, and microbial activities (Kim et al., 2002). Among these factors, the pH and redox reactions may be the most important parameters to control the arsenic mobility through both processes, even though potential rates of the two processes are different. Arsenic and iron oxides have a redox-sensitive nature. Therefore, transfer of large amounts of arsenic between the solid phases and neighboring water may result from redoxfacilitated precipitation and dissolution reactions (Camm et al., 2004). Arsenic dissolution can occur due to changes in the geochemical environment into a reductive condition. A high pH condition can also induce desorption of arsenic due to the negative net surface charge of iron oxide (Pfeifer et al., 2004). Since arsenic can be transported to other areas through these processes and can create secondary arseniccontamination sites, arsenic-contaminated soils must be treated in a rapid and safe way. USEPA (1997) and Mulligan et al. (2001) described several available remediation technologies for heavy-metal contaminated soils. Among remediation technologies, soil washing cannot only extract heavy-metals or metalloids adsorbed or precipitated into soils, but it can also reduce the volume of contaminated soils. Soil washing can also be applied to large contaminated areas because of its rapid kinetics, ease, and economic efficiency (USEPA, 2001). For washing techniques, the selection of extractants (or agents) is the most important step because the extraction effectiveness of each extractant is different depending on its target contaminants, bonding strength, and soil characteristics. Up until now, several types of extractants (e.g., inorganic salts, inorganic acids, organic acids, and alkaline agents) have been studied for extracting heavy-metals or metalloids from tailings or soils. Among them, sodium hydroxide and hydrochloric acid have been known to be economical and effective for extracting arsenic from soils (Van Benschoten et al., 1994; Jang et al., 2005). In this study, sequential washing techniques using single or dual agents (NaOH and hydrochloric acid) were applied for arsenic-contaminated soils in an abandoned iron-ore mine area. Arsenic-containing iron-ore fines in this area can produce mobile colloids through coagulation and flocculation in natural waters contacting the soils (Pandey et al., 2004). The mobile colloids can be easily transported to other areas and can create secondary arsenic-contamination sites. Thus, a rapid and effective remediation is needed to satisfy the arsenic regulation for reuse or safe disposal of washed soils. Through establishing the following objectives, an effective washing strategy was found for treating arsenic-contaminated soils in floc-forming iron ore, as well as arsenic in the washing effluents. The specific objectives are as follows: (1) to observe the physico-chemical properties and arsenic partitioning into different compartments of soils through the sequential extraction procedure, (2) to determine parameters of soil washing such as effective physical sizing, types and concentration of washing agent, and ratio of agent volume to soil mass, (3) to find washing efficiencies with different types of sequential washing steps

9

using single or dual agents and floc removal, and finally (4) to investigate remediation strategies to enlarge arsenic removal efficiencies for both soils in abandoned iron-ore mines and arsenic-containing washing effluents, while meeting the regulatory limit for washed soils. 2. Materials and methods 2.1. Soil selection and characteristics Soil samples (designated as DC soils) were collected from arsenic-contaminated areas located at the Dal-Chun abandoned mine (Ulsan, Kyungsannamdo, South Korea) that had been developed for iron ore. Iron sources of the iron-ore mine mainly consist of the iron oxide minerals: magnetite (Fe3O4) and hematite (Fe2O3), goethite (Fe2O3 Æ H2O), and limonite (a mixture of hydrated iron oxides). Magnetite is a naturally occurring metallic mineral that is the dominant species of magnetic compartments, while hematite and goethite are the main species of crystalline iron oxides, which are nonmagnetic minerals. The DC soils include both nonmagnetic and magnetic minerals (Fig. 1). The arsenic concentration of this area was much higher than 15 mg kg1 (mg of arsenic per kg1 of soils), which is the concern level of the Soil Environment Conservation Act of Korea (MOE, 2003) legislated by Korean Ministry of Environment (MOE). The Korean standard test (KST) methods were utilized to estimate arsenic concentrations extracted from soils. The total volume of arsenic-contaminated soils for this area was about 1 780 000 m3. Soils taken from this area were sieved through a 4.75-mm opening sieve (no. 4) to remove large particles and allow a homogeneous soil size distribution. Characteristics of DC soils

Fig. 1. Photo of DC soils: magnetic separation of magnetite.

10

M. Jang et al. / Chemosphere 66 (2007) 8–17

such as pH, organic content, particle density, and uniformity coefficient (D60/D10) were measured by Methods of Soil Analysis (Page et al., 1986). The measurement of the cation-exchange capacity (CEC) was conducted by EPA Method 9080 (USEPA, 1986a,b). The total arsenic and iron concentrations of soil samples were measured by EPA 3050B (USEPA, 1986a,b), which is a hot nitric acid digestion method for soil. The detailed method of EPA 3050B adopted in this study was well described by Jang et al. (2005). As a disposal or reuse criterion of arsenic-contaminated soils, the KST methods for soils were adopted from the Soil Environment Preservation Act (MOE, 2002). The strictest regulation of arsenic concentration (6 mg kg1) extracted by the KST methods was selected as a strategy of soil remediation. The KST method is as follows: (1) add 50 ml of HCl (1 M) to each 10 g of soil sample, (2) shake the suspension at a speed of 100 rpm and 30 °C for 30 min, (3) centrifuge 10 ml of suspension at 3200 rpm for 20 min, and (4) filter the supernatant with a 0.6-lm micropore filter, dilute the filtrate, and acidify the filtrate with conc. HNO3 before the arsenic analysis. Arsenic concentrations of filtrates were measured using inductively coupled plasma spectrometry (ICP-1000VI, Shimadzu Company, Japan) at a concentration range of 0.02–20 mg l1. For the arsenic precipitation tests of washing effluents, arsenic concentration was measured by atomic absorption spectrometry (AA-6401F, ShimadzuÒ, Japan) connected with a continuous hydride generator (HVG-1, ShimadzuÒ, Japan) that has a detection limit of 0.5 lg l1. Duplicates, blank and EPA reference standards were analyzed with each set of samples as a quality control check on the analysis.

2.2. Arsenic and iron sequential extraction procedure, soil sieve analysis and arsenic extraction using the KST methods for different sizes of DC soils Although sequential extractions are operationally defined and not fully specific in extracting the element bound to a given fraction, they can provide comparative information to elucidate the relative contribution of the target compound and aid in the predictions of elemental mobility (Keon et al., 2001; Pueyo et al., 2003). Based on the chemical properties of the target binding phases, apportions of arsenic in mg kg1 can be quantified with sufficient sensitivity (Keon et al., 2001). Arsenic and iron species were analyzed for each extraction step of the sequential extraction procedure that is well described by Jang et al. (2005). Cumulative mass percentages for each size of DC soils were analyzed with the sieve analysis method, and arsenic extractions (KST methods) for different-size soils acquired from the sieve analysis were obtained. These results support the essential information for the following experimental results (especially soil size and washing agent effects on washing efficiency).

2.3. Effects of washing agents, concentrations, soil sizes, ratio of solution volume (ml) to soil mass (g), and sequential washing by use of single agent In this study, HCl and NaOH were selected as washing agents because they have been known to be effective in extracting arsenic and are economical (Van Benschoten et al., 1994; Jang et al., 2005). To estimate washing efficiencies of different concentrations of each agent on different sizes of DC soils, the following batch-scale washing tests were conducted. Fifty grams of different sizes of DC soils (0.25–4.7, 0.15–4.7, 0.09–4.7, and <4.7 mm) were washed with 250 ml of predetermined concentrations (0.1, 0.2, 0.5, 1, or 2 M) of HCl or NaOH. The suspension was then mixed at 20 ± 0.5 °C in a shaker at 300 rpm for 6 h. This mixing condition was identically applied for all tests. After mixing, arsenic in washing solutions was analyzed to calculate the extracted arsenic concentrations based on the mass of DC soils. Suspended solids were separated by filtration (0.45-lm micropore filter) and dried at 105 °C for 2 h before arsenic extractions (KST methods). Through this method, we evaluated whether arsenic-contaminated soils are satisfactory for on-site disposal or reuse after washing. Regarding the experimental results for soil size effects, nonhomogeneous soils (2.0–4.7 mm) and soils of target physical sizing (<0.15 mm) were excluded from the following experiments. The ratio of solution volume (ml) to soil mass (g) (designated as ml g1) is a significant parameter in soil washing. Different concentrations of HCl (0.1, 0.2, or 1 M) or NaOH (0.2 or 1 M) solution were used with different ratios (1, 3, 5, or 10 ml g1), in which soil mass was fixed at 20 g. Five series of sequential washing tests were conducted with the same concentration of single agent each time to find the applicability of sequential washing. For these tests, different concentrations of HCl (0.1, 0.2, or 1 M) or NaOH (0.2 or 1 M) were applied at a fixed ratio (5 ml g1) of solution volume to soil mass (20 g). 2.4. Characterization of flocculation To characterize the flocculation that occurred through the washing tests, the following procedure was carried out: the DC soils (0.15–2.0 mm) were washed using deionized water at 5 ml g1 of solution volume to soil mass for 30 min and floc formed was taken with a micro pump (Masterflux, Cole-Parmer) to a funnel filtration unit containing a 0.45-lm micropore filter. Then, the filtered solids were dried at 105 °C for 2 h. After cooling, the mass of the solids was recorded before X-ray diffraction analysis and arsenic extractions using the KST methods. Using an XRD (DMX-ULTIMA series, Rigaku) equipped with Cu Ka radition (40 kV, 25 mA), X-ray diffraction patterns in a long-range (0–70°) were obtained for dried solids. Several washing tests using different concentrations of HCl or NaOH were conducted to find out the degree of floc formation and arsenic loadings. Mass portions and arsenic

M. Jang et al. / Chemosphere 66 (2007) 8–17

extractions (KST methods) of each dried floc were analyzed along with the pH of resultant washing solution. 2.5. Effects of floc removal and sequential washing by use of dual agents To find the floc removal effect on the efficiencies of sequential washing, five series of sequential washing tests including floc removal steps were conducted with 1 M of HCl or NaOH solution. Flocs formed in suspension were removed at each washing step by the method previously described in Section 2.4. In the second trials, three sequential washing tests were conducted with different concentrations of dual agents (HCl and NaOH) in an alternative method. For these tests, 0.2 M HCl was used during the first washing step because of its highest efficiency of flocculation and arsenic removal that have been shown in previous experimental results. Arsenic extractions (KST methods) were carried out for washed soils produced at each combination. 2.6. Kinetic study Among several combinations, the sequential washings using 0.2 M HCl, followed by 1 M HCl and then 1 M NaOH, were most likely to have the highest efficiency to meet the disposal or reuse criterion (6 mg As kg1) for washed soils. In order to find the effect of the washing time of the second and third washing steps on arsenic extractions into the liquid phase and arsenic extractions (KST methods) of washed soils, the following kinetic tests were conducted. As previously described, several duplicated batch samples were prepared and washed with 0.2 M HCl at the fixed ratio (5 ml g1) of solution volume to soil mass (20 g), and then flocs were removed in the same manner. These pretreated samples were sequentially washed with 1 M HCl, followed by 1 M NaOH for 5, 10, 15, 20, 30, 60, 120, 180, 240, 360, 720, or 1440 min, respectively. The flocs formed at each washing step were removed using

11

the previously described method. Arsenic concentrations in each washing solution and arsenic extractions (KST methods) of each washed sample were analyzed. 3. Results and discussion 3.1. Soil characteristics and arsenic sequential extraction procedure The pH of the DC soils was 8.4. The organic content (5.4%) and particle density (2.38 g cm3) were higher than other previously reported values of arsenic-contaminated soils in our study (Jang et al., 2005). The uniformity coefficient (5.5) was also higher than the reported values of arsenic-contaminated tailings or soils (Jang et al., 2005), representing a more heterogeneous size distribution in the DC soils. Cation-exchange capacity (CEC) was 15.4 meq (100 g)1. Total arsenic and iron concentrations were 1710 ± 250 and 44 400 ± 4240 mg kg1, respectively. Arsenic KST extractions of the DC soils were 321 ± 32 mg kg,1 which is about 54 times higher than the regulation (6 mg kg1). Table 1 shows arsenic and iron concentrations for each step of the sequential extraction procedure. The total arsenic (1410 ± 124 mg kg1) and iron (31 700 ± 1650 mg kg1) concentrations in the sequential extraction procedure were about 82.7% and 71% of the total arsenic and iron determined by EPA 3050B, respectively. This result might be due to the fact that the aqua regia of the sequential extraction procedure is not strong enough and can extract only about 70–90% of the total iron or arsenic. Based on the sequential extraction procedure, arsenic (87.5%) and iron (99.7%) were most abundant in crystalline minerals (seventh step). In general, arsenic extracted from the first four steps is associated with sorption sites of each specific compartment, while arsenic extracted in the seventh step is related to dissolution of iron precipitates (Fe–As compound or coprecipitates) (Kim et al., 2003). Therefore, it can be thought that most arsenic prevails in

Table 1 Sequential extraction concentrations of arsenic and iron for DC soil (2.5 g) Step

Fractions

Extraction method for each fraction

Arsenic conc., mg kg1 (portion, %)

1 2 3

Soluble Adsorbed Carbonate

3.20 ± 0.2 (0.2%) 63.4 ± 1.8 (4.5%) 45.3 ± 4.4 (3.2%)

1.0 ± 0.0 (0.003%) 6.1 ± 2.6 (0.019%) 19.6 ± 0.3 (0.062%)

4

Organic matters

37.9 ± 3.4 (2.7%)

8.8 ± 1.0 (0.028%)

5

Easily reducible oxides

6

Amorphous oxides

7

Crystalline mineralsb Total (mg kg1)

0.2 M KCl (25 ml) STa (2 h) 0.1 M Na2HPO4 (25 ml, pH 8.0), ST (20 h) 1 M CH3COONa (25 ml), ST (5 h), then 0.1 M Na2HPO4 (25 ml, pH 8.0), ST (20 h) 5% NaOCl (10 ml, pH 9.5), then heating at 70 ± 0.5 °C, ST (30 min) 0.1 M NH2OH (25 ml, pH 2.0), ST (30 min), then 0.1 M KOH (25 ml), ST (20 h) 0.25 M NH2OH/HCl (25 ml, 50 ± 0.5 °C), ST (30 min), then 0.1 M KOH (25 ml), ST (20 h) Aqua regia (30 ml HCl and 10 ml HNO3), ST (1 h)

a b

Iron conc., mg kg1 (portion, %)

6.4 ± 1.8 (0.5%)

22.1 ± 1.3 (0.07%)

19.3 ± 1.4 (1.4%)

24.6 ± 1.6 (0.078%)

1240 ± 108 (87.5%) 1410 ± 124 (100%)

31 600 ± 1640 (99.74%) 31 700 ± 1650 (100%)

ST (stirring time). Crystalline minerals: the operationally defined crystalline mineral fraction of crystalline oxide and amorphous aluminosilicates.

12

M. Jang et al. / Chemosphere 66 (2007) 8–17

Fe–As precipitates or coprecipitates, and iron exists largely in crystalline forms such as goethite (Fe2O3 Æ H2O, 63% Fe), hematite (Fe2O3, 70% Fe), and magnetite (Fe3O4, 72% Fe). The primary forms in iron-ore mines are hematite and magnetite (Jorgenson, 2004). Most arsenic in crystalline minerals exists in the oxidized form (arsenate) (Legiec et al., 1997). Foster et al. (1998) studied the quantitative arsenic speciation in mine soils using X-ray absorption spectroscopy. The results of their X-ray absorption near edge structure (XANES) analysis indicated that arsenate is the dominant oxidation state in mine soils. In arsenic mobility characterization, there are two main processes: desorption and dissolution. Desorption of arsenate is related to the pH value of the water. As the most important minerals of soils, aluminum or iron oxides provide high numbers of sorption sites for anionic species. When the pH of water contacting soils is above the zeropoint-of-charge (ZPC), the net surface charge of aluminum or iron oxide becomes negative, resulting in the repulsion of negatively charged ions such as arsenate (H2 AsO 4 or HAsO2 4 ). Since arsenic extracted from the first four steps (10.6%) is associated with the sorption sites of each specific compartment, the high pH of DC soils can induce some mobility of arsenic. Meanwhile, the redox reaction can control not only the speciation, but also dissolution of arsenic and iron. Iron oxides dissolve under reducing conditions, dissociating iron–arsenic complexes. Arsenate also reduces to arsenite that is present as the uncharged species (H3AsO3). This reduced form is more mobile than the charged species. 3.2. Soil sieve analysis and arsenic extraction using KST methods for each soil size Fig. 2 shows the cumulative mass percentages and arsenic extractions (KST methods) for different sizes of DC soils obtained by sieve analysis. As the soil sizes decrease, arsenic concentrations increase. Arsenic concen-

100 <0.075mm

1200 80

60

900

0.075~0.09mm

0.09~0.15mm

40

600

0.15~0.25mm 0.25~0.42mm

20

0.42~0.83mm 0.83~2.0mm

300

2.0~4.7mm >4.7mm

0 0.1

1

Extracted As, KST Method [mg/kg]

Cumulative Mass Percentages [%]

1500

0 10

Avg. Opening Size [mm] Fig. 2. Mass fractions and arsenic extractions (KST methods) for each size of DC soils.

trations exponentially increased from about 400 mg kg1 (at 0.15–0.25 mm of soil size) to about 1200 mg kg1 (at <0.075 mm). This result was caused by the fact that the smaller particles had higher arsenic concentrations due to their higher surface areas. The cumulative mass percentage of this portion (<0.15 mm) was less than 10%. Although these fine particles are a small fraction of the total mass of soils, they could perform poorly in the washing technique (Zhang et al., 2003). The cleaning effect of soil washing can be, to some extent, accomplished by the removal of a fraction of fine particles. As an advantage in the soil washing technique, the physical sizing determined through washing tests for different sizes of soils is an important process to increase the extraction efficiency. The physical sizing can be successfully achieved by means of a hydrocyclone (Kawatra, 1985; Trawinski, 1985). On the other hand, the soil particles larger than 4.7 mm show much smaller arsenic concentrations (about 10 mg kg1) compared to other sizes. 3.3. Effects of washing agents, concentrations, soil sizes, ratio of solution volume (ml) to soil mass (g), and sequential washing by use of a single agent Different sizes of DC soils were washed with several predetermined concentrations of NaOH or HCl. Fig. 3(A) and (C) show arsenic extractions in washing solutions, while (B) and (D) are arsenic extractions (KST methods) of washed soils. Arsenic extractions in washing solutions increased with an increase of small-size soil portions and washing agent concentrations, even though their trends were different for each agent. At low concentrations of washing agent (<0.5 M), the arsenic extracted was lower than 50 mg kg1 for HCl. In the case of NaOH, it was much higher, with arsenic extractions of 50–120 mg kg1. It is especially notable that arsenic extractions (<5 mg kg1) at 0.1 and 0.2 M HCl were much lower than those at the same concentrations of NaOH. These results might be affected by amorphous ferrihydrite [Fe(OH)3] formed in flocs, which could effectively remove arsenic extracted in solution. More details are described in Section 3.4. At concentrations higher than 0.5 M, arsenic extractions with NaOH were almost the same or decreased, while they increased largely for HCl. Arsenic extractions (KST methods) for washed soils decreased with an increase of washing agent concentrations, suggesting that higher concentrations of washing agent are more effective in removing arsenic from the DC soils. However, they increased with an increase in the portions of small-size soils. These results indicate that higher arsenic concentrations remained for smaller particles of soils. Arsenic extractions (KST methods) were similar for both sizes of 0.25–4.7 mm and 0.15–4.7 mm in the washing results of HCl, while alkaline washing showed larger differences in which washing efficiency was better for larger soil particles. According to previous results (Fig. 2), arsenic extractions (KST methods) for different sizes of DC soils increased exponentially at sizes smaller than 0.15 mm (sieve

M. Jang et al. / Chemosphere 66 (2007) 8–17

250

Extracted As, KST Method [mg/kg]

Extracted As [mg/kg]

300 0.25-4.7 mm 0.15-4.7 mm 0.09-4.7 mm < 4.7 mm

200 150 100 50 0 0.0

0.5

A

1.0

1.5

60 40 20 Soil Remediation Standard, As = 6 mg/kg

0 0.0

0.5

1.0

1.5

2.0

HCl Conc. [M] 100

Extracted As, KST Method [mg/kg]

Extracted As [mg/kg]

0.25-4.7 mm 0.15-4.7 mm 0.09-4.7 mm < 4.7 mm

80

B

300 0.25-4.7 mm 0.15-4.7 mm 0.09-4.7 mm < 4.7 mm

250 200 150 100 50

C

100

2.0

HCl Conc. [M]

0 0.0

13

0.5

1.0

1.5

2.0

NaOH Conc. [M]

D

0.25-4.7 mm 0.15-4.7 mm 0.09-4.7 mm < 4.7 mm

80 60 40 20

Soil Remediation Standard, As = 6 mg/kg

0 0.0

0.5

1.0

1.5

2.0

NaOH Conc. [M]

Fig. 3. Soil washing tests using different concentrations of HCl or NaOH for different sizes of DC soils, (A) and (C) arsenic extractions in washing solutions, (B) and (D) arsenic extractions (KST methods) of washed soils.

no. 100). Thus, the target physical sizing was determined as 0.15 mm. Nonhomogeneous soil sizes (2.0–4.7 mm) were also excluded for the following experiments to obtain reliable data from soil washing. However, the washing step by use of a single agent is not effective in remediating the DC soils for on-site disposal or reuse application because the arsenic extractions (KST methods) were not lower than 6 mg kg1 even at 2 M of HCl or NaOH. Fig. 4 shows arsenic extractions (KST methods) for soils washed with different ratios of washing solution to soil mass: (A) HCl and (B) NaOH. With an increase in ratios and concentrations of HCl, arsenic extractions (KST methods) for washed soils decreased. In the case of NaOH, arsenic extractions (KST methods) decreased linearly for 0.2 M, but they were almost the same in the whole range of ratios for 1 M. The control of the ratio is also not effective in remediating the DC soils because the arsenic extractions (KST methods) did not meet the regulation (6 mg kg1), even at the highest ratio (9 ml g1) of solution volume to soil mass. Considering the efficiencies of arsenic extraction and economical use of agent solution, the ratio (5 ml g1) of washing solution to soil mass is thought to be the ideal, as shown in our previous study (Jang et al., 2005). Arsenic extractions (KST methods) for DC soils sequentially washed with different concentrations of single agent are shown in Fig. 4(C) and (D). The trends of arsenic extractions with different concentrations of agents in the first step were similar to those of previous washing tests

at 5 ml g1, suggesting that batch washing tests are reliable to show reproducibility. Most cases decrease linearly as the number of washings increases. Thus, the sequential washing procedure is effective in removing arsenic from the DC soils (Legiec et al., 1997). At the same concentration of 0.2 M, NaOH is more effective than HCl. Although five sequential washing steps were conducted with either HCl or NaOH, none of the data from the arsenic extractions (KST methods) was 6 mg kg1. Therefore, new approaches including floc removals and sequential washing steps using dual agents have been developed and tried to remediate the arsenic-contaminated soils that existed in abandoned iron ores in an economical and effective way. 3.4. Flocculation Fig. 5 shows the X-ray diffraction pattern of dried flocs that has three dominant phases: silicate, ferric arsenate (Fe8As10O23), and two-line ferrihydrite [Fe(OH)3]. Among them, two-line ferrihydrite was identified with two broad diffraction peaks at 35° and 62°, corresponding to d-spacing of 0.25 and 0.15 nm, respectively (Van der Giessen, 1966; Hofmann et al., 2004; Rhotona and Bighamb, 2005). Through charge neutralization of negatively charged colloidal particles (e.g., silicate and dissolved organic matter) in soils, coagulation and flocculation occurred by cationic hydrolysis of amorphous precipitates of metal

M. Jang et al. / Chemosphere 66 (2007) 8–17

Extracted As, KST Method [mg/kg]

B

0

Extracted As, KST Method [mg/kg]

A

100

0 3 6 9 -1 Ratio [solution volume to soil mass (ml g )]

C

80 60 0.1 M HCl 40

0.2 M HCl 1 M HCl

20

Soil Remediation Standard, As = 6 mg/kg

100

0

0 3 6 9 Ratio [solution volume to soil mass (ml g-1)]

D

80 60

0.2 M NaOH

40 1 M NaOH 20 Soil Remediation Standard, As = 6 mg/kg

0.1 M HCl 0.2 M HCl 1 M HCl

80

60

40

20 Soil Remediation Standard, As = 6 mg/kg

0

1

Extracted As, KST Method [mg/kg]

Extracted As, KST Method [mg/kg]

14

2 3 4 5 Sequential Washing Number

80

0.2 M NaOH 1 M NaOH

60

40

20 Soil Remediation Standard As = 6 mg/kg

0

1

2 3 4 5 Sequential Washing Number

Intensity [cps]

Fig. 4. Ratio [solution volume to soil mass (ml g1)] effects on soil washing: arsenic extractions (KST methods) of washed soils using different concentrations of HCl (A) or NaOH (B), sequential washing effects of different concentrations of HCl (C) or NaOH (D) at 5 ml g1.

Fe8As10O23

0.25 nm 0.15 nm

SiO2

10

SiO2

20

30

40

50

60

2 Theta [deg.]

Fig. 5. X-ray diffraction patterns of dried flocs.

hydroxide (Gregory and Duan, 2001; Duan and Gregory, 2003). Primary particles of amorphous ferrihydrite formed by Fe–O–Fe polymerization of iron oxyhydroxyl monomers and dimers can be aggregated to form large flocs (Dousma and De Bruyn, 1976; Waychunas et al., 1993). As an oxygen consuming process, magnetic iron oxide such as magnetite (Fe3O4) can be dissolved to produce amorphous ferrihydrite (Pratt et al., 1996; Cumbal et al., 2003): 1 9 Fe3 O4 ðSÞ þ O2 ðaqÞ þ H2 O $ 3FeðOHÞ3 ðSÞ 4 2 Amorphous ferrihydrite [Fe(OH)3] has been known to have high affinities for adsorbing both arsenite and arsenate because of its high surface area and the reactivity of surface functional groups (Jackson and Miller, 2000). As shown in the XRD results, flocs can contain not only dissolved ar-

senic, but also fine particles of ferric arsenate (FeAsO4)coprecipitated with silicate. After the washing tests using different concentrations of HCl or NaOH, the flocs formed were collected and dried using the method of Section 2.4. The arsenic extractions (KST methods) were applied to the dried flocs, and the mass ratios of dried flocs to soils were obtained. Arsenic extractions (KST methods) of each dried floc were obtained with the pH of resultant washing solutions because the formation of amorphous ferrihydrite and arsenic adsorption depend strongly on pH (Fig. 6). 0.2 M HCl showed the highest arsenic extraction (1086 mg kg1) of dried floc at a pH of 5.9. This extraction value is about 4.4 or 7.4 times higher than those of 1 M HCl (245 mg kg1, pH 0.6) or NaOH (147 mg kg1, pH 13.3). Mass ratios of dried flocs for soils applied were similar for all cases (0.102–0.125), even though it is not possible to separate pure amorphous ferrihydrite formed in flocs. However, washing by use of 0.2 M HCl showed the highest value (0.125) of mass ratios of dried flocs. Additional studies are required to reveal the arsenic speciation and compartments for components of floc. Adsorbing colloid flotation (ACF) by use of ferric hydroxide as the coprecipitant and sodium dodecyl sulfate (SDS) as negatively charged colloids has been developed to be effective in removing arsenic (Peng and Di, 1994). Peng and Di showed the optimum pH of 4–5 to achieve 99.5% arsenic removal efficiency. Generally, the best pH condition for iron hydrolysis has been known to be 4–5. Therefore, it can be suggested that pH adjustment in the first washing

M. Jang et al. / Chemosphere 66 (2007) 8–17 0.130 0.125

1200

0.2 M HCl 0.120

0.1 M HCl

900

0.115 600

0.2 M NaOH 1 M HCl

300

1 M NaOH 0

0.110

0

3

6

9

12

0.105

A

0.100

40

Extracted As, KST Method [mg/kg]

As extractions (KST methods) mass portion of dried floc

Mass ratio of dried flocs to soils

Extracted As, KST Method [mg/kg]

1500

15

pH

step using HCl is most important to increase cationic hydrolysis of amorphous ferrihydrite, inducing the high sorption capacity for dissolved arsenic and ferric arsenate-containing fine SiO2 particles. The removal step of arsenic-containing flocs in the soil washing technique is thought to lower arsenic loading of washed soils. Since iron-ore fines in this area can produce mobile colloids containing high concentrations of arsenic in natural waters contacting the soils, arsenic can be easily transported to other areas and can create secondary arsenic-contaminated sites. 3.5. Effects of floc removal and sequential washing by use of dual agents

30

20

10 Soil Remediation Standard, As = 6 mg/kg

0

1

2

4

5

B

10

Soil Remediation Standard, As = 6 mg/kg

5

0 1M

0.2M

3rd Washing : NaOH HCl

To study the floc removal effect on arsenic extractions (KST methods) of washed soils, five sequential washing tests using a single agent of 1 M HCl or NaOH were conducted with floc removal steps at each washing [Fig. 7(A)]. Floc removal was more effective for washing tests using 1 M HCl than NaOH. This result might be due to the fact that the arsenic extraction (KST methods) (245 mg kg1) and mass portion (0.123) of floc formed by washing with 1 M HCl were higher than those with 1 M NaOH (147 mg kg1 and 0.114) observed in previous tests. Arsenic extractions (KST methods) of soils washed with 1 M HCl were lower than 6 mg kg1 in the fourth step, while those with 1 M NaOH were higher than 6 mg kg1 even in the fifth step. Although four sequential washings using 1 M HCl can lower arsenic concentrations below the regulation, high volumes of chemicals are still needed and resultant acidic washing effluents should be neutralized and treated in an effective and safe way. Fig. 7(B) shows arsenic extractions (KST methods) of soils washed through three sequential washing steps using single or dual agents. For these trials, 0.2 M HCl was used in the first washing step to enhance the remove of arsenic contained in flocs. Among all combinations, alternative sequential washings using 1 M HCl in the second step

3

Sequential Washing Number

Extracted As, KST Method [mg/kg]

Fig. 6. Arsenic extractions (KST methods) and mass ratio of dried flocs to soils obtained after washing by use of different concentrations and types of washing agent.

1 M HCl, with flocs 1 M HCl, without flocs 1 M NaOH, with flocs 1 M NaOH, without flocs

2nd Washing :

1 M NaOH

1st Washing :

1M 1M 0.2M HCl NaOH HCl 0.2 M HCl

1M HCl

0.2M 0.2M 1M HCl NaOH NaOH 1 M HCl

0.2 M HCl

Fig. 7. Sequential washing and floc removal effects of single or dual agents: arsenic extractions of washed soils.

and NaOH (0.2 or 1 M) in the third step showed reliable and lower values of arsenic extractions (KST methods) of washed soils. The lowest value was 1.5 ± 0.3 mg kg1 obtained using 1 M NaOH in the third step. 3.6. Kinetic study Fig. 8 shows arsenic extractions in washing solutions (A) and arsenic extractions (KST methods) (B) of soils washed for each predetermined time in the second and third washing steps. For the second washing step by use of 1 M HCl, arsenic extractions in the washing solution increase to 2.6 mg l1 until 360 min, then decrease linearly to 1 mg l1. Meanwhile, the arsenic extractions (KST methods) of soils washed in the second step decrease steeply to about 2.4 mg kg1 at 60 min, and increase steadily to 3.6 mg kg1 at 360 min, and then increase more to 5.5 mg kg1. Thus, these observations were related to the readsorption of

16

M. Jang et al. / Chemosphere 66 (2007) 8–17 5

Concentration of As (mg/l)

4

Third step (1 M NaOH)

3

Second step (1 M HCl) 2

1

0

0

300

600

900

1200

1500

Time [min]

A 7

Extracted As, KST Method (mg/kg)

Soil Remediation Standard, As = 6 mg/kg 6

4. Conclusion 5

4

Second step (1 M HCl)

3

2

1

0

B

minerals. This is because of the ligand displacement reaction of hydroxyl ions with arsenic species and high pH conditions that can prevent readsorption of arsenic because predominant negatively charged crystalline oxides do not attract the negatively charged oxyanions. In our previous research, arsenic in resultant washing solutions was effectively removed by pH adjustment or ferric chloride addition (Jang et al., 2005). Along with high washing efficiencies, there is another advantage to use dual agents in the soil washing technique. Mixing both resultant washing solutions characterized with acidic and basic conditions can neutralize the pH so that dissolved arsenic can be effectively removed by the hydrolysis of dissolve metal species (especially iron species) (Dousova et al., 2005; Yong Gan et al., 2005). In this study, the resultant solutions of each washing step produced at the ending time (1440 min) were mixed for 10 min. The pH of the mixed solution was 9.53. Then, the solution was passed through a 0.45-lm filter, and the arsenic of filtrate was 37 lg l1 which is below the current arsenic regulation (50 lg l1).

Third step (1 M NaOH) 0

300

600

900

1200

1500

Time [min]

Fig. 8. Kinetics of arsenic extractions into resultant solutions (A) and arsenic extractions (KST methods) of washed soil (B).

dissolved arsenic into soils that have dominant species of crystalline iron oxides. In this study, the DC soils were characterized with a high portion of crystalline iron oxide through the sequential extraction procedure. At a low pH, the presence of positively charged crystalline iron oxide such as goethite or hematite can easily readsorb soluble arsenic (Jackson and Miller, 2000). Gruebel et al. (1988) showed 100% of readsorption of arsenate and selenate in the presence of goethite that were dissolved from the amorphous iron phase by hydroxylamine extract (pH <1). In the third washing step using 1 M NaOH, arsenic extracted into washing solution increased to about 2.8 mg l1 at 360 min and steadily increased to 4.2 mg l1. In the case of arsenic extractions (KST methods) of soils washed in the third step, they decreased rapidly to about 1 mg kg1 within 15 min and decreased reliably more to 0.2–0.4 mg kg1 at 360–1500 min. Thus, based on these results, alkaline extraction using NaOH is effective in removing arsenic readsorbed in the surface of crystalline

In this study, a soil washing technique was applied to remediate arsenic-contaminated soils in an abandoned iron-ore mine area to meet the current arsenic regulation for safe disposal or reuse application. The sequential extraction procedure showed most arsenic exists in Fe–As precipitates or coprecipitates and iron exists mostly in the crystalline forms of iron oxide. Soil washing by use of a single agent was not effective in remediating the soils at any washing condition. Iron-ore fines existing in the DC soils can flocculate, containing dissolved arsenic and fine particles of ferric arsenate-coprecipitated silicate. The removal step of arsenic-containing flocs by use of 0.2 M HCl was important to lower arsenic extractions (KST methods) of washed soils. Among all combinations, alternative sequential washing using 1 M HCl (second washing step) and NaOH (third washing step) shows reliable and lower values of arsenic extractions (KST methods). Kinetic studies showed dissolved arsenic was easily readsorbed into residual soils, in which crystalline iron oxides positively charged at a low pH dominantly existed. The subsequent alkaline extraction was effective in removing arsenic readsorbed in the surface of crystalline minerals. The ligand displacement reaction of hydroxyl ions with arsenic species and high pH conditions that can prevent readsorption of arsenic are the main reasons for these results. Based on preliminary tests in this study, sequential washing using dual agents has excellent potential for remediating the DC soils, satisfactory for on-site disposal or reuse application of washed soils. Acknowledgements We thank three anonymous individuals for critical review of the manuscript. The present study was supported

M. Jang et al. / Chemosphere 66 (2007) 8–17

by research grants from Ministry of Environment (MOE), Republic of Korea, and Kwangwoon University in 2005. References Camm, G.S., Glass, H.J., Bryce, D.W., Butcher, A.R., 2004. Characterisation of a mining-related arsenic-contaminated site, Cornwall, UK. J. Geochem. Explor. 82, 1–15. Cumbal, L., Greenleaf, J., Leun, D., SenGupta, A.K., 2003. Polymer supported inorganic nanoparticles: characterization and environmental applications. React. Funct. Polym. 54, 167–180. Dousma, J., De Bruyn, P.L., 1976. Hydrolysis-precipitation studies of iron solutions. I. Model for hydrolysis and precipitation from Fe(III) nitrate solutions. J. Colloid Interface Sci. 56, 527–539. Dousova, B., Kolousek, D., Kovanda, F., Machovic, V., Novotna, M., 2005. Removal of As(V) species from extremely contaminated mining water. Appl. Clay Sci. 28, 31–42. Duan, J., Gregory, J., 2003. Coagulation by hydrolysing metal salts. Adv. Colloid Interfac., 475–502. Foster, A.L., Brown, G.E., Tingle, T.N., Parks, G.A., 1998. Quantitative arsenic speciation in mine tailings using X-ray absorption spectroscopy. Am. Mineral. 83, 553–568. Gregory, J., Duan, J., 2001. Hydrolyzing metal salts as coagulants. Pure Appl. Chem. 73, 2017–2026. Gruebel, K.A., Davis, J.A., Leckie, J.O., 1988. The feasibility of using sequential extraction techniques for arsenic and selenium in soils and sediments. Soil Sci. Soc. Am. J. 52, 390–397. Hofmann, A., Pelletier, M., Michot, L., Stradner, A., Schurtenberger, P., Kretzschmar, R., 2004. Characterization of the pores in hydrous ferric oxide aggregates formed by freezing and thawing. J. Colloid. Interf. Sci. 271, 163–173. Jackson, B.P., Miller, W.P., 2000. Effectiveness of phosphate and hydroxide for desorption of arsenic and selenium species from iron oxides. Soil Sci. Soc. Am. J. 64, 1616–1622. Jang, M., Hwang, J.S., Choi, S.I., Park, J.K., 2005. Remediation of arsenic contaminated tailing or soils. Chemosphere 60, 344–354. Jorgenson, J.D., 2004. Minerals Yearbook: Iron Ore. USGS, Reston, VA, pp. 28. Kawatra, S.S.K., 1985. Calculating the particle size distribution in a hydrocyclone overflow product for simulation purposes. Miner. Metall. Proc. 2, 152–154. Keon, N.E., Swartz, C.H., Brabander, B.J., Harvey, C., Hemond, H.F., 2001. Validation of an arsenic sequential extraction method for evaluating mobility in sediments. Environ. Sci. Technol. 35, 2778– 2784. Kim, M.J., Ahn, K.H., Jung, Y., 2002. Distribution of inorganic arsenic species in mine tailings of abandoned mines from Korea. Chemosphere 49, 307–312. Kim, J.Y., Davis, A.P., Kim, K.W., 2003. Stabilization of available arsenic in highly contaminated mine tailings using iron. Environ. Sci. Technol. 37, 189–195. Legiec, I.A., Griffin, L.P., Walling Jr., P.D., Breske, T.C., Angelo, M.S., Isaacson, R.S., Lanza, M.B., 1997. DuPont soil washing technology program and treatment of arsenic contaminated soil. Environ. Prog. 16, 29–34. MOE, 2002. The Korean Standard Test (KST) Methods for Soils. Korean Ministry of Environment, Gwachun, Kyunggi, pp. 225 (in Korean).

17

MOE, 2003. Examination of Soil Contamination Caused by Abandoned Heavy-Metal Mines in Korea. Korean Ministry of Environment, Gwachun, Kyunggi. Mulligan, C.N., Yong, R.N., Gibbs, B.F., 2001. Remediation technologies for metal-contaminated soils and groundwater: an evaluation. Eng. Geo. 60, 193–207. Page, A.L., Miller, R.H., Keeney, D.R., 1986. Methods of Soil AnalysisAgronomy Series 9. American Society of Agronomy, Madison, WI. Pandey, P.K., Nair, S., Bhui, A., Pandey, M., 2004. Sediment contamination by arsenic in parts of central-east India and analytical studies on its mobilization. Curr. Sci. India 86, 190–197. Peng, F.F., Di, P., 1994. Removal of arsenic from aqueous solution by adsorbing colloid flotation. Ind. Eng. Chem. Res. 33, 922–928. Pfeifer, H.-R., Gueye-Girardet, A., Reymond, D., Schlegela, C., Temgouaa, E., Hesterberg, D.L., Chou, J.W., 2004. Dispersion of natural arsenic in the Malcantone watershed, Southern Switzerland: field evidence for repeated sorption–desorption and oxidation–reduction processes. Geoderma 122, 205–234. Pratt, A.R., Nesbitt, H.W., Mycroft, J.R., 1996. Increased reactivity of pyrrhotite and magnetite phases in sulphide mine tailings. J. Geochem. Explor. 56, 1–11. Pueyo, M., Sastre, J., Hernandez, E., Vidal, M., Lopez-Sanchez, J.F., Rauret, G., 2003. Prediction of trace element mobility in contaminated soils by sequential extraction. J. Environ. Qual. 32, 2054– 2066. Rhotona, F.E., Bighamb, J.M., 2005. Phosphate adsorption by ferrihydrite-amended soils. J. Environ. Qual. 34, 890–896. Trawinski, H.H.F., 1985. Practical hydrocyclone operation. Filtr. Separat. 22, 22–27. USEPA, 1986a. Atomic Absorption Methods: Test Methods for Evaluating Solid Waste. SW846, Method 3050B. USEPA, Office of Solid Waste and Emergency Response, Washington, DC. USEPA, 1986b. Cation-Exchange Capacity of Soils (Ammonium Acetate): Test Methods for Evaluating Solid Waste. SW846, Method 9080. USEPA, Office of Solid Waste and Emergency Response, Washington, DC. USEPA, 1997. Engineering Bulletin: Technology Alternatives for the Remediation of Soils Contaminated with As, Cd, Cr, Hg, and PbEPA 540-S-97-500. USEPA, Office of Emergency and Remedial Response, Washington, DC. USEPA, 2001. Region 2 news and speeches: EPA to modify cleanup plan for arsenic-contaminated soil at superfund site in Vineland, New Jersey. Van Benschoten, J.E., Reed, B.E., Matsumoto, M.R., McGarvey, P.J., 1994. Metal removal by soil washing for an iron oxide coated sandy soil. Water Environ. Res. 66, 168–174. Van der Giessen, A.A., 1966. The structure of iron (III) oxide-hydrate gels. J. Inorg. Nucl. Chem. 28, 2155–2159. Waychunas, G.A., Rea, B.A., Fuller, C.C., Davis, J.A., 1993. Surface chemistry of ferrihydrite: Part I. EXAFS studies of the geometry of coprecipitated and adsorbed arsenate. Geochim. Cosmochim. Ac. 57, 2251–2269. Yong Gan, W., Selomulya, C., Tapsell, G., Amal, R., 2005. Densification of iron(III) sludge in neutralization. Int. J. Miner. Process 76, 149– 162. Zhang, M.K., He, Z.L., Calvert, D.V., Stoffella, P.J., Yanga, X.E., Li, Y.C., 2003. Phosphorus and heavy metal attachment and release in sandy soil aggregate fractions. Soil Sci. Soc. Am. J. 67, 1158–1167.