Adsorption and heterogeneous oxidation of arsenite on modified granular natural siderite: Characterization and behaviors

Applied Geochemistry 48 (2014) 184–192

Contents lists available at ScienceDirect

Applied Geochemistry journal homepage: www.elsevier.com/locate/apgeochem

Adsorption and heterogeneous oxidation of arsenite on modified granular natural siderite: Characterization and behaviors Kai Zhao a,b,1, Huaming Guo a,b,⇑, Xiaoqian Zhou b a b

State Key Laboratory of Biogeology and Environmental Geology, China University of Geosciences, Beijing 100083, PR China School of Water Resources and Environment, China University of Geosciences, Beijing 100083, PR China

a r t i c l e

i n f o

Article history: Available online 23 July 2014 Editorial handling by M. Kersten

a b s t r a c t Although natural siderite has been investigated to remove both As(III) and As(V), it has relatively low adsorption rate and capacity. It is crucial to enhance its adsorption characteristics for As removal prior to being used in practical application. Modified granular natural siderite (MGNS) was fabricated through addition of organic binder, extrusion granulation and calcination, and evaluated for adsorption characteristics by means of batch and column tests. Results showed that MGNS had higher adsorption rate and capacity for As(III) in comparison with natural siderite. Arsenic(III) adsorption achieved equilibrium at 24 h, with adsorption capacity of 9.43 mg/g estimated from Langmuir isotherm at 25 °C. Column tests showed that there was less difference in total As loads in MGNS-packed filters for As(III)-spiked deionized water, As(III)-spiked tap water, and real-world high-As groundwater. The coexistence of anions had no significant effect on As adsorption in both batch and column experiments. Results of XRD, SEM and BET analysis indicated that MGNS, as an Fe(II)/(III) hybrid system, had a much larger specific surface area relative to the pristine natural siderite due to massive spherical aggregates attaching to the siderite matrix. XANES spectra showed that As(V) was the major species in the adsorbent after As(III) adsorption. Its proportion in total As slightly increased with the increase in contact time. Adsorption and heterogeneous oxidation of As(III) were believed to be the main mechanisms of As(III) removal by MGNS. This study suggested that MGNS is a potential adsorbent for effectively removing As from As-contaminated groundwater in filter application. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Due to its high toxicity and carcinogenicity, increasing attention has been focused on drinking water As in the last decades. Arsenic poisoning episodes have been reported all over the world. Most are the results of exposure to high As drinking water, especially groundwater, which has affected over 100 million people worldwide, including America, Argentina, Bangladesh, Chile, China, Hungary, India, Mexico, Romania, and Vietnam (Viraraghavan et al., 1999; Nordstrom, 2002; Smedley and Kinniburgh, 2002; Tchounwou et al., 2003; Hughes et al., 2011; Guo et al., 2014). Considering its potential health risk, the World Health Organization, 1996 has set a recommendation guideline limit of 10 lg/L for As in drinking water. It is notable that the new standard for drinking water has been executed after risk assessment in China since 2007 ⇑ Corresponding author at: School of Water Resources and Environment, China University of Geosciences, Beijing 100083, PR China. Tel.: +86 10 8232 1366; fax: +86 10 8232 1081. E-mail addresses: [email protected] (K. Zhao), [email protected] (H. Guo). 1 Mobile: +86 134 6633 7455; fax: +86 10 8232 1081. http://dx.doi.org/10.1016/j.apgeochem.2014.07.016 0883-2927/Ó 2014 Elsevier Ltd. All rights reserved.

(Ministry of Health of PR China, 2006; Liu et al., 2009). The maximum contaminant level (MCL) of As in drinking water has been reduced from 50 lg/L to 10 lg/L in urban areas, although 50 lg/L is still adopted in rural areas. Overall, it is an emerging issue to improve novel treatment technologies and to develop new materials for As removal from As-contaminated drinking water. Although As is mostly found in inorganic form as oxyanions of trivalent As(III) or pentavalent As(V) in natural waters, As-enriched groundwater was generally dominated by As(III), up to 96%, in the form of the uncharged As(III) species H3AsO3 at neutral pH (Smedley and Kinniburgh, 2002; Berg et al., 2007; Guo et al., 2008). Thus, As(III) is more difficult to be removed from water by means of adsorption and coprecipitation due to the lack of electrostatic attraction, in comparison with As(V) (Mohan and Pittman, 2007). In recent years, Fe-based sorbents have been developed and shown good adsorption behaviors for As removal. Some synthetic materials, containing Fe-(oxyhydr)oxides, such as granular ferric hydroxide (GFH) and Bayoxide E33 (Thirunavukkarasu et al., 2003a; Mohan and Pittman, 2007), have been used for As removal in practice as commercial adsorbents. However, the synthesized adsorbents would induce a significant hydraulic

K. Zhao et al. / Applied Geochemistry 48 (2014) 184–192

obstruction during water treatment processes, and are usually more expensive than naturally occurring Fe minerals, such as magnetite, goethite, hematite and siderite, which have already been studied for As removal (Ohe et al., 2005; Giménez et al., 2007; Guo et al., 2007a,b; Jönsson and Sherman, 2008). Previous studies have shown that siderite, both natural and synthetic, efficiently removed aqueous As species, including both As(III) and As(V) (Guo et al., 2007a,b, 2010, 2011). Guo et al. (2007a) reported that As adsorption on natural siderite with the particle size of 0.10–0.25 mm achieved equilibrium at a contact time of 3 days, and the estimated maximum adsorption capacities were 1.04 and 0.52 mg/g for As(III) and As(V), respectively. In contrast, due to the small size and the great reactivity of the adsorbents, As adsorption on ferrous carbonate, synthesized by mixing Fe(II) with HCO 3 , reached equilibrium at only 3 h, with the estimated adsorption capacities for As(III) and As(V) being up to 10 mg/g (Guo et al., 2010, 2011). However, the complex preparation processes, high cost and low hydraulic conductivity of the synthetic siderite limited its application in rural areas. To overcome these limitations, the natural siderite with the fine grain size of <0.10 mm, simultaneously generated in the process of mineral crushing, was a good alternative as a principal component. However, there is no information on As removal by using the fine grain size of natural siderite. In addition, it was recently observed that the bi-mineral coexistence during partial mineral transformation from siderite to goethite greatly enhanced As adsorption capacity of synthetic siderite under oxic conditions (Guo et al., 2013), which provides the preliminary idea for modification of natural siderite in enhancing As adsorption (Zhao and Guo, 2014). This study was carried out to investigate the possibility of using the fine grain size of natural siderite as the major components in fabricating modified granular natural siderite (MGNS) for As(III) removal. The modified granular adsorbents were synthesized by addition of organic binder, extrusion granulation, and calcination, which greatly improved the adsorption capacity of the material. The adsorbent was evaluated for adsorption characteristics in terms of batch and column tests. Effects of contact time, reaction temperature, solution pH, background electrolytes and coexisting anions on As(III) removal were intensively investigated in batch tests. Arsenic-spiked water and real-world high As groundwater were used to assess the performance of MGNS-packed columns in removing As. Additionally, removal mechanisms were expounded by means of BET surface area measurement, XRD, scanning electron microscopy, and X-ray adsorption spectroscopy. 2. Materials and methods All reagents used were of analytical grade. The As(III) stock solution was prepared with deionized (DI) water using sodium arsenite (NaAsO2, Fluka Chemical). Diverse As(III) solutions were freshly prepared by diluting As(III) stock solution with DI water. 2.1. Preparation of MGNS adsorbent Natural siderite was obtained from a mineral company in Guizhou Province, China. Prior to experiments, the siderite was sieved with a 150-mesh sieve to get the fine grain size particles, which were used for adsorbent synthesis. Polyanionic cellulose (PAC) (Luzhou North Chemical Industries Co., Ltd., China) was used as the binder. The mixture with solid–liquid ratio of 5 g siderite to 1 mL DI water (containing 2% PAC) was adequately stirred and then extruded into strip-like adsorbent with a diameter of about 1.5 mm under a pressure of about 5.5 MPa. These strips were modified in a muffle furnace at 50–600 °C for 0.5–4.0 h. The maximum As removal efficiency was observed for the modified siderite being

185

calcinated at 350 °C for about 2 h. Finally, the calcinated strips were manually cut into particles with length of 1–2 mm, which were assigned as MGNS and used in both batch and column experiments. The average crushing strength reached up to 50 N per single particle. 2.2. Batch experiments Batch tests were carried out at a fixed dose of 10 g/L, with 0.5 g adsorbent in 50 mL solution in 100 mL high density polyethylene (HDPE) bottles under oxic conditions. These bottles were placed in a shaking water bath with a temperature controller, at a shaking rate of 150 rpm. Adsorption isotherm studies were performed with As(III)spiked DI water by varying initial As(III) concentrations (1–100 mg/L) at different temperatures (i.e., 25, 35, and 45 °C) with contact time of 48 h. Other batch experiments were carried out with initial As(III) concentration of 5 mg/L at 25 °C with contact time of 48 h. Adsorption kinetic studies were conducted with contact time from 0.5 to 96 h using As(III)-spiked DI water and natural high-As groundwater obtained from the Hetao basin, Inner Mongolia. Chemical compositions of natural high-As groundwater are shown in Table 1, indicating that it was of Na–HCO3–Cl type. Effect of initial solution pH on As(III) adsorption was investigated in As(III)-spiked DI water with initial pH between 2 and 10, which were adjusted by using 0.05 M (mol/L) HCl and 0.01 M NaOH. Effect of background electrolytes was tested using As(III)-spiked DI water with sodium chloride and sodium bicarbonate concentrations between 1 and 100 mM. In addition, effect of co-existing anions was studied using As(III)-spiked DI water containing individual anions with 0.5, 1, 2, 5,10 and 20 mg/L of N as NO 3 , or S 2 3 as SO2 4 , or Si as SiO3 , or P as PO4 . It is worth noting that all batch experiments were conducted in duplicate and reported as a mean value. 2.3. Column experiments Plexiglass columns, with an inner diameter of 30 mm, a height of 150 mm and a working volume of about 100 mL, were used in the column study as fixed-bed up-flow reactors. Arsenic(III)-spiked DI water, As(III)-spiked tap water, and realworld high-As groundwater, were used as influent, and pumped through the MGNS-packed column filters at an up-flow rate of 1.8 mL/min with a peristaltic pump (BT100-1F, Longerpump) under oxic conditions. Quartz sand with grain size of 0.5–1.0 mm was packed at the bottom of the column (10 mm height), and the MGNS in the upper (115 mm height, about 118 g weight), presenting the mean bed porosity of 0.55. The empty bed contact time (EBCT) was about 45 min. In order to evaluate the feasibility of MGNS in practical application, four columns were setup to treat different types of high-As water. Column A was used for treating 1.0 mg/L As(III)-spiked tap water, column B for 1.0 mg/L As(III)-spiked DI water, column C for 2.0 mg/L As(III)-spiked DI water, and column D for real-world high-As groundwater. Chemical compositions of tap water and real-world high-As groundwater are shown in Table 1. Effluent solutions from the column filters were collected at regular intervals and analyzed for residual As concentration. 2.4. Analytical methods Solution pH was monitored by a standard pH meter (PB-10, Sartorius). Dissolved Fe and As concentrations were analyzed by ICPMS (7500C, Agilent). Multi-element standard solutions (GBW 081532 and GBW 081533) from National Institute of Metrology (China) were tested every 5 samples for quality control. The relative standard deviation (RDS) was less than ±2%, with the

186

K. Zhao et al. / Applied Geochemistry 48 (2014) 184–192

Table 1 Chemical compositions of natural high-As groundwater and tap water used in batch and column experiments.

Natural groundwater Tap water

Natural groundwater Tap water

Total As (lg/L)

pH

Na+ (mg/L)

Mg2+ (mg/L)

K+ (mg/L)

Ca2+ (mg/L)

Fe (lg/L)

Mn (lg/L)

666 0.35

8.38 7.31

314 14.3

21.0 14.1

3.89 2.42

16.0 39.8

564 8.40

37.1 4.35

As(III) (lg/L)

As(V) (lg/L)

Cl (mg/L)

HCO 3 (mg/L)

NO 3 (mg/L)

SO2 4 (mg/L)

Si (mg/L)

P (mg/L)

580 N.D.

85.4 N.D.

232 31.1

617 26.2

N.D. 66.0

N.D. 70.9

2.63 N.D.

N.D. N.D.

Note: N.D. = non-detectable.

analytical precision of 0.5%. Arsenic species were determined using an HPLC–ICP–MS, with the relative standard deviation (RDS)< ±5%, with the analytical precision of 2.0% (Guo et al., 2012). Ferrous Fe and total Fe concentrations in solids were analyzed by chemical extractions (Andradae et al., 2002) and determined by the 1,10phenanthroline method (Amonette and Templeton, 1998). The specific surface area was measured by BET N2 adsorption/ desorption isotherm method, and pore volume and size were calculated by the BJH method using the automatic analyzer (3H2000PS2, Beishide). The mineral composition was determined by X-ray diffraction analysis (XRD), using a URD-6 powder diffractometer (Cu Ka radiation, graphite monochromator, 2h range 2.6–70°, step 0.01°, counting time 5 s per step). Morphological analysis was performed by field emission scanning electron microscopy (FE– SEM) using Zeiss SUPRA 55 microscope (at 20 kV) with energy-dispersive X-ray analyses. Arsenic K-edge X-ray absorption near-edge structure (XANES) spectra were recorded at the beamline BL14W, at 3.5 GeV and 300 mA, in Shanghai Synchrotron Radiation Facility (SSRF), China. A Si(1 1 1) monochromator was used. All spectra were acquired in the energy range from 150 to +400 eV relative to As K-edge, and were collected in fluorescent mode with silicon drift fluorescence detector at room temperature. Prior to XANES analysis, samples were dried and ground to fine powder (<200 mesh) in a glovebox (Coy Lab, USA) with O2 levels <1.0 mg/L. In order to avoid oxygen from contacting the samples and restrain the beaminduced oxidation during measurement, samples were tightly sealed by Kapton tape (Guo et al., 2013). Repeated measurement of samples showed no evidence of As(III) oxidation during beam exposure in this case. Linear combination fitting (LCF) of XANES data was used to quantitatively determine As oxidation state with standards of XANES spectra of As(III)- and As(V)-loading synthetic siderite, which contacted As(III) solution and As(V) solution, respectively, with initial As concentration of 500 mg/L and dosage of 2 g/L for 96 h under anoxic conditions. 3. Results and discussion 3.1. Characterization of MGNS adsorbent According to XRD analysis (Fig. 1), it was noted that siderite is the major mineralogical component in natural pristine material, with quartz and aluminosilicate clay minerals as minor minerals. However, MGNS contains both siderite and hematite, being considered as an Fe(II)/(III) hybrid system. After calcinated at 350 °C for 2 h, siderite was partially transformed into hematite (Fig. 1b), although almost all of siderite was converted to crystalline hematite in the materials calcinated at 600 °C for 2 h (Fig. 1c). Mineral transformation was expected to be crucial factors for increasing in both specific surface area and pore volume. The transformation occurred as Eq. (1).

4FeCO3 þ O2 ! 2Fe2 O3 þ 4CO2

ð1Þ

SEM images clearly show that natural siderite was mainly in the form of dense lamellar siderite matrix (Fig. 2a), while large

Fig. 1. XRD patterns of pristine natural siderite (a), MGNS (b) and calcinated adsorbent at 600 °C (c). Peaks showed D, Q, H, and S for dolomite, quartz, hematite, and siderite, respectively.

quantities of spherical aggregates with small size particle (100 nm diameter) were presented after modification (Fig. 2b). It was inferred that these aggregates consist of activated siderite particles and Fe oxide minerals (i.e., hematite confirmed by XRD analysis) generated from Fe(II) oxidation. The nitrogen adsorption–desorption isotherm and corresponding mesoporous size distribution of MGNS are shown in Fig. 3. It illustrated that N2 adsorption–desorption isotherm (Fig. 3a) was of IUPAC Type IV, indicating that pore size was in the range of mesopore. The hysteresis loop occurred in the range of relative pressure P/P0 = 0.5–1.0, corresponding to capillary condensation in voids of inter-particles. On the other hand, the pore volume was up to 0.102 cm3/g and most of mesoporous diameters ranged from 3.0 to 8.0 nm according to Fig. 3b. The BET specific surface area of MGNS was 50.24 m2/g, being approximately 25 times greater than that of pristine natural siderite (2.02 m2/g). Besides, the total pore volume became much larger (from 0.0065 to 0.0696 mL/g), while the average pore size got smaller (from 10.61 to 3.66 nm) after modification. 3.2. Batch tests 3.2.1. Adsorption kinetics Time dependence for As(III) adsorption on MGNS is shown in Fig. 4a. It was clearly seen that residual As concentration significantly decreased with an increase in contact time, especially during the first 12 h. The equilibrium time of As(III) adsorption on MGNS in As(III)-spiked DI water was about 24 h, much shorter than that on natural siderite at 25 °C (72 h) (Guo et al., 2007a), indicating that the adsorption kinetic rate on MGNS was higher than that on natural siderite. The same adsorption kinetics was also observed in real-world high-As groundwater (Fig. 4a). Thus,

K. Zhao et al. / Applied Geochemistry 48 (2014) 184–192

187

Fig. 2. SEM images and energy-disperse spectra of pristine natural siderite (a and c) and MGNS (b and d).

Fig. 3. The nitrogen adsorption–desorption isotherm and corresponding mesoporous size distribution of MGNS.

a reaction time of 48 h was chosen for further batch studies to make sure that adsorption equilibrium was reached. Kinetic data for As(III) adsorption on MGNS in As(III)-spiked DI water and real-world high-As groundwater were both well fitted to Langergren pseudo-second-order rate model (R2 > 0.99), suggesting that the overall rate of As(III) adsorption should be controlled by chemical processes. Arsenic(III) adsorption amounts at equilibrium calculated through the pseudo-second-order model (0.50 and 0.078 mg/g in As(III)-spiked DI water and natural high-As groundwater, respectively) were closer to adsorption capacities at equilibrium obtained from experiments (0.49 and 0.077 mg/g) than those calculated through the pseudo-first-order model (0.35 and 0.036 mg/g). It indicated that the adsorption process could be better described by the pseudo-second order kinetics model. Generally, adsorption on granular, porous metal oxides involves a sequential progression through four steps: diffusion through the bulk liquid, film diffusion, intraparticle diffusion, and adsorption on the solid surface. Moreover, the bulk liquid diffusion and adsorption steps are rapid and thus are not rate-limiting (Badruzzaman et al., 2004). Therefore, to identify mass transport

Fig. 4. Effect of contact time (a) and the intraparticle rate of As(III) adsorption onto MGNS based on Weber-Morris model (b) (initial As(III) concentration in As(III)-spiked DI water = 5.0 mg/L; total As concentration of 666 lg/L and As(III) concentration of 580 lg/L in natural high-As groundwater; adsorbent dosage = 10 g/L).

188

K. Zhao et al. / Applied Geochemistry 48 (2014) 184–192

processes limiting the rate of As(III) adsorption on MGNS, the intraparticle diffusion model (Weber and Morris, 1963) was used to describe the kinetics data, as given in Eq. (2).

qt ¼ K W t 0:5 þ b

ð2Þ

where qt is the amount of As(III) adsorbed at time t (mg/g), KW is intraparticle diffusion rate constant (mg/(g h0.5)), b represents boundary layer diffusion. According to this model, the plot of qt against t0.5 should yield a straight line passing through the origin if intraparticle diffusion is the rate-controlling factor. Weber and Morris plots of As(III) adsorption on MGNS are shown in Fig. 4b. The plots showed multi-linearity, suggesting the complexity of the adsorption processes for both As(III)-spiked DI water and natural high As groundwater. At the first 4 h, the plots were linear and passed through the origin, which indicated that initial stage might be controlled by the intraparticle diffusion. Afterwards, formation of inner-sphere complexes precluded surface diffusion, leading to declining the external mass transfer due to the decrease in concentration gradient. The film diffusion gradually became the rate-determining step until the diffusion processes reached equilibrium, which could be verified by the increase in the value of intercept b. The same phenomena had been observed for As(V) adsorption on GFH (Badruzzaman et al., 2004) and synthetic siderite (Guo et al., 2010). 3.2.2. Adsorption isotherms Adsorption isotherm data at various temperatures were analyzed using Langmuir (Eq. (3)) and Freundlich (Eq. (4)) adsorption isotherms for parameter estimation via non-linear regression. Langmuir isotherm model indicates the reversible phenomenon and monolayer adsorption with a finite number of identical sites, while Freundlich model is always applied to multilayer adsorption as well as non-ideal adsorption on heterogeneous surfaces (Foo and Hameed, 2010).

qe ¼

bqm ce 1 þ bce

qe ¼ K F ce1=n

ð3Þ ð4Þ

where ce (mg/L) is the equilibrium As concentration, qe (mg/g) is the amount of As adsorbed at equilibrium, qm (mg/g) is the maximum adsorption capacity, b is the Langmuir constant being related to the binding energy or net enthalpy of adsorption, and KF and n are the Freundlich constants indicative of the adsorption capacity and adsorption intensity, respectively. Results showed that Freundlich isotherm model fitted As(III) adsorption data at different temperatures with correlation coefficients between 0.9864 and 0.9882, better than Langmuir isotherm model (r2 between 0.9376 and 0.9675) (Fig. 5), which was consistent with As(III) adsorption on synthesized siderite under oxic conditions (Guo et al., 2011, 2013). Therefore, it indicated that the multilayer adsorption resulting from the changes in the surface of MGNS adsorbent would be involved in the process of As(III) removal. Langmuir parameters (b, qm) and Freundlich constants (KF, 1/n) obtained from isotherm model fittings are shown in Table 2. The calculated 1/n was about 0.4 (<1.0), indicating favorable adsorption of As(III) on MGNS. In addition, the maximum adsorption capacity (qm) and Freundlich constant KF varied from 9.43 mg/g to 10.2 mg/g and from 3.01 to 4.75 when the experimental temperature increased from 25 °C to 45 °C, respectively. It was suggested that high temperature would strengthen As(III) adsorption on MGNS. Comparison of As(III) adsorption capacity of MGNS with other adsorbents is shown in Table 2. Since pH, adsorbent dosage and initial concentration range would affect As adsorption, specific experimental conditions are included in Table 2. Langmuir isotherm

Fig. 5. Langmuir isotherm (dashed lines) and Freundlich isotherm (solid lines) of As(III) adsorption on MGNS at different temperatures (initial As(III) concentration in As(III)-spiked DI water = 1–100 mg/L; adsorbent dosage = 10 g/L; contact time = 48 h; initial pH = 7.0).

fitting and parameter calculation showed that As(III) adsorption capacity of MGNS at 25 °C was 9.43 mg/g in this study, while that of natural siderite with the grain size of 0.10–0.25 mm at an adsorbent dosage of 2 g/L and initial concentrations between 0.25 and 2.0 mg/L was 1.04 mg/g (Guo et al., 2007a). Therefore, the modification has greatly improved As adsorption on the siderite material. Besides, as shown in Table 2, MGNS has much higher adsorption capacity for As(III) removal than other granular Fe-bearing adsorbents, such as goethite, hematite, magnetite (Giménez et al., 2007), granular ferric hydroxide (Thirunavukkarasu et al., 2003a), and Fe oxide-coated cement (Kundu and Gupta, 2007). Although synthetic siderite (Guo et al., 2011) seems to perform better than MGNS, MGNS has the advantage of separation in practical application due to the large grain size and high hydraulic conductivity and appears to be a promising adsorbent. 3.2.3. Effect of initial pH on As(III) adsorption Results of As(III) adsorption on MGNS at various initial pH (2– 10) are presented in Fig. 6. Obviously, As(III) adsorption generally kept constant (0.494–0.498 mg/L) at the pH range investigated. This trend somewhat agreed with the others obtained on ferrihydrite (Raven et al., 1998), HFO and goethite (Dixit and Hering, 2003), zero-valent Fe (Su and Puls, 2001; Kanel et al., 2005), and synthetic siderite (Guo et al., 2011). Dissolved Fe concentration was up to 1.7 mg/L at initial pH of 2.0, while several lg/L at other initial pH (data not shown). It was speculated that the adsorbent was partially dissolved at initial solution pH of 2.0, which slightly reduced adsorption sites for As removal. When the initial pH varied from 3.0 to 10.0, equilibrium solution pH had the range between 7.0 and 7.9, showing a ‘‘buffering’’ effect in the system with the presence of MGNS. It could be explained by the amphoteric nature of Fe oxides/oxyhydroxides with Eqs. (5) and (6) (Zhang et al., 2004). The adsorbent was capable of maintaining solution pH between 7.0 and 8.0. Therefore, it could be concluded that adsorption on MGNS was less affected by initial solution pH between 3.0 and 10.0.

 FeOH þ Hþ  FeOHþ2

ð5Þ

 FeOH  FeO þ Hþ

ð6Þ

3.2.4. Effects of background electrolytes and coexisting anions Results of As(III) adsorption on MGNS in the presence of background electrolytes (NaCl and NaHCO3) and coexisting anions

Chowdhury and Yanful (2011)

Guo et al. (2011)

Kanel et al. (2005)

Thirunavukkarasu et al. (2003a) Kundu and Gupta (2007)

Giménez et al. (2007)

3.005 3.469 4.746 13.10 – – – 18.00 0.2200 3.500 5.060 2.560 2.370 2.180 5.750 – 0.9882 0.9871 0.9864 0.9930 – – – 0.9900 0.9940 – – – 0.9980 0.9930 0.9900 – 0.4799 0.7746 1.6284 1.900 4.267 0.7333 0.3333 – 0.4000 – – – 0.2707 0.2787 1.228 6.600

0.3848 0.3772 0.3575 0.5850 – – – 0.4348 0.9259 0.3058 0.4950 0.5848 0.3900 0.4700 0.4300 –

Guo et al. (2007a)

189

Fig. 6. Effect of initial pH on As(III) adsorption onto MGNS (initial As(III) concentration in As(III)-spiked DI water = 5 mg/L; adsorbent dosage = 10 g/L; contact time = 48 h).

2 2 3 (NO 3 , SO4 , SiO3 , and PO4 ) are shown in Fig. 7. Although bicarbonate or silicate/phosphate showed a greater negative effect on As(III) removal than chloride or nitrate/sulfate, the adsorption capacities for all batches ranged between 0.485 and 0.498 mg/g. It indicated that background electrolytes or the presence of coexisting anions generally had no significant effect on As removal. Furthermore, it is worth noting that phosphate had adverse effect on As(III) adsorption as much as silicate at element concentrations <10 mg/L, while it had more significant effect than silicate at concentration of 20 mg/L due to the similarity in chemical structure and reactivity between As(V) and phosphate (Luengo et al., 2007; Mihaljevic et al., 2009). Although Meng et al. (2002) and Stachowicz et al. (2008) reported that the affinity of As(III) to Fe hydroxides was much weaker than As(V), As(III) adsorption responded less strongly to the change in phosphate concentration in comparison with As(V) adsorption due to a different electrostatic interaction (Stachowicz et al., 2008). Since As(V) was the major species on the surface of As(III)-treated MGNS, As(III) was believed to be firstly adsorbed and then oxidized to As(V) on the surface based on As speciation in both solids and solutions, which is discussed later. The less competition of phosphate with As(V) in this study than that in others (Meng et al., 2002; Stachowicz et al., 2008) would be due to the similar complexation nature between As(V) oxidized from As(III) on the surface and solids to that between As(III) and solids, although it needs further investigations. Natural high-As groundwater was generally dominated by As(III), and mostly had high total dissolved solid (TDS), with Cl and HCO 3 concentrations up to 70.7 and 23.8 mM, respectively (Guo et al., 2008). Besides, concentrations of other oxyanions showed significant variations, for example, SO2 concentration 4 ranged from <0.1 to 434 mg/L and Si from 3.5 to 16.9 mg/L (Guo et al., 2008; Deng et al., 2009). Thus, the experimental data suggested that As(III) adsorption on MGNS should be generally independent of the background electrolytes and coexisting anions with the ranges of concentrations observed in natural high-As groundwater.

1.8–7.5 0.71–3.0 0.4 Magnetite nanoparticles (20 nm)

‘‘–’’ No related data.

1.0–20 Synthetic siderite (powder <200 mesh)

2

2.0–40

1.010 1

1.0–10

– 0.50–10 – 30

22 27 25 35 45 25 35 45 22

0.90–9.0 0.017–0.33

0.015–15 5 25

0.075–75

0.13–1.0 0.25–2.0 2

Natural siderite (0.10–0.25 mm) Goethite (0.25 mm) Hematite (0.25 mm) Magnetite (0.10 mm) Granular ferric hydroxide (0.8–1.2 mm) Iron oxide coated cement (0.212 mm) Nanoscale Zero-Valent Fe (fine powder)

1.0–100 10

25 35 45 20 MGNS (1.0–2.0 mm)

0.10–10

0.9376 0.9519 0.9675 0.9820 0.9980 0.9990 0.9960 0.9700 0.9990 – – – 0.9550 0.9790 0.9820 0.9700

9.429 9.545 10.15 1.040 0.3767 0.2715 0.2069 0.1120 0.6900 1.800 2.470 1.560 9.980 10.01 14.51 7.962

KF Freundlich

R2 R2

Ratio of initial As to adsorbent (mg/g) Range of initial concentrations (mg/L) Adsorbent dosage (g/L) T (°C) Adsorbent

Table 2 Isotherm parameters for As(III) adsorption on MGNS and other granular ferric adsorbents at different temperatures.

Langmuir

qm (mg/g)

b (L/mg)

1/n

This study

References

K. Zhao et al. / Applied Geochemistry 48 (2014) 184–192

3.3. Column studies Breakthrough curves of MGNS-packed columns for As(III) removal are illustrated in Fig. 8, which showed that the removal efficiency of Column A for treating 1.0 mg/L As(III)-spiked tap water appeared to be approximately equal to that of Column B

190

K. Zhao et al. / Applied Geochemistry 48 (2014) 184–192

Fig. 7. Effect of background electrolytes(a) and coexisting anions (b) on As(III) adsorption onto MGNS (initial As(III) concentration in As(III)-spiked DI water = 5 mg/L; adsorbent dosage = 10 g/L; contact time = 48 h).

Fig. 8. Breakthrough curves of MGNS-packed columns for treating 1.0 mg/L As(III)-spiked tap water (Column A), 1.0 mg/L As(III)-spiked DI water (Column B), and 2.0 mg/L As(III)-spiked DI water (Column C) (a), and natural high-As groundwater with total As concentration of 666 lg/L and As(III) concentration of 580 lg/L (Column D) (b) (flow rate = 1.8 mL/min; EBCT = 45 min).

for treating 1.0 mg/L As(III)-spiked DI water, although tap water contained some coexisting anions (Table 1). It indicated that effects of coexisting anions in the tap water on As(III) adsorption onto MGNS were not significant, which was consistent with the result of batch experiments. Effluent As concentration of Column A increased rapidly after about 2200 bed volumes (BV). Both Columns A and B treated over 1600 BV of 1.0 mg/L As(III) solutions before the Chinese drinking water standard in rural areas of 50 lg/L As was broken through, and approximately 1200 BV prior to breakthrough of 10 lg/L As (Fig. 8a). A mass balance calculation indicated that total As loads in Column A and B were around 0.83 mg/g before 10 lg/L of As breakthrough, and 1.10 mg/g before 50 lg/L of As breakthrough. The MGNS presented a comparable performance to Fe oxide-coated sand (1403 BV with 0.5 mg/L As(III)-spiked Regina tap water, and total As load of 0.65 mg/g before 10 lg/L of As breakthrough) (Thirunavukkarasu et al., 2003b) and GAC-based Fe-containing adsorbent (7500 BV with influent As(III) concentration of 56 lg/L, and total As load of 0.82 mg/g before 10 lg/L of As breakthrough) (Gu et al., 2005). Column C for treating 2.0 mg/L As(III)-spiked DI water was setup to investigate effect of initial As(III) concentration on As removal by MGNS. As depicted in Fig. 8a, the drinking water standard of 10 lg/L was exceeded after 730 BV and 50 lg/L MCL

was surpassed after 930 BV. The As load was 1.00 mg/g before 10 lg/L of As breakthrough in the Column C, and 1.28 mg/g before 50 lg/L of As breakthrough. Similar As loads in Columns A and C indicated that the adsorption capacity of MGNS before 50 lg/L of As breakthrough was basically constant with initial As(III) concentrations between 1.0 and 2.0 mg/L. After having demonstrated the high removal efficiency of MGNS columns for synthesized As(III) solutions, Column D, with the same operating conditions, was utilized to evaluate the performance for treating natural high-As groundwater with total As concentration of 666 lg/L and As(III) concentration of 580 lg/L (Table 1). Breakthrough curve is shown in Fig. 8b. The breakthrough of 10 lg/L was not observed after over 1400 BV total throughputs (approximately 115 L in bulk). The results distinctly demonstrated that MGNS, as filter fillings, was expected to effectively remove As from realworld groundwater in practical treatment processes. 3.4. Mechanism of As(III) removal Solute adsorption largely depends on the surface area and other characteristics of the adsorbent. In this study, the surface area of MGNS was approximately 25 times greater than that of pristine natural siderite, and the total pore volume became much larger,

K. Zhao et al. / Applied Geochemistry 48 (2014) 184–192

while the average pore size got smaller after modification. All these changes are more favorable to As adsorption. On the other hand, XRD results indicated that partial mineral transformation from siderite to hematite occurred in MGNS (Fig. 1), being regarded as an Fe(II)/Fe(III) hybrid system. The Fe(II)/(III) hybrid system with a much larger specific surface area was believed to the major contribution to high adsorption capacity for As (Roberts et al., 2004; Muñiz et al., 2009; Amstaetter et al., 2010; Ona-Nguema et al., 2010). It was reported that reactive Fe(III) intermediate formed from oxidation of Fe(II) by electron transfer into the bulk Fe(III) phase played a vital role in As(III) adsorption and oxidation (Amstaetter et al., 2010). Meanwhile, Guo et al. (2013) observed that the calculated As adsorption on synthetic siderite under oxic conditions was around 11 times higher than that under anoxic conditions, which was attribute to the mineral transformation from siderite (Fe(II)) to amorphous needle-like goethite (Fe(III)) under oxic conditions. For those reasons, we believed that the adsorption mainly took place at the amorphous Fe(III) species formed by the oxidation of siderite during the preparation process of MGNS.

Fig. 9. Arsenic K-edge XANES spectra of As(III)-treated MGNS after different contact time with initial As concentrations of 5.0 mg/L, adsorbent dosage of 10 g/L at 25 °C, and reference compounds: sodium arsenite (NaAsO2), sodium arsenate (Na2HAsO47H2O), As(III)- and As(V)-adsorbing synthetic siderite under anoxic conditions (Siderite-As(III) (anoxic) and Siderite-As(V) (anoxic), respectively). Dotted plots: XANES raw data; solid lines: results of linear combination fitting; dashed lines: contribution of model compounds for the fitting procedure.

191

Arsenic K-edge XANES analysis was used to determine As redox state in the solids after adsorption. The reference spectra in the bottom of Fig. 9 exhibit a well-resolved edge structure with adsorption maximum at 11871.5 eV and 11875.0 eV, corresponding to As(III) and As(V), respectively. Arsenic K-edge XANES spectra in Fig. 9 showed that As occurred as both As(V) and As(III) in all As(III)-adsorbing MGNS from batches with different contact time and initial As(III) concentration of 5.0 mg/L at 25 °C. Similarly, previous studies observed that As occurred in forms of both As(III) and As(V) on As(III)-adsorbing synthetic siderite (Guo et al., 2011, 2013) and the Fe0 surface (Su and Puls, 2001) after As(III) adsorption. Linear combination fitting (LCF) curves of As(III)-treated MGNS with standards of As(III)- and As(V)-adsorbing synthetic siderite under anoxic conditions are shown in Fig. 9. It indicated that As(V) was the major species in all As(III)-treated MGNS at different contact time, and its proportion in total As slightly increased from 79.3% to 90.3% with the increase in contact time from 1 h to 48 h. It was suggested that fast oxidation occurred in the system of As(III)MGNS under aerobic conditions, with 79% As(III) being oxidized into As(V) within 1 h. Zhao et al. (2011) found that adsorbed-As(III) on the surface of ferrihydrite was gradually converted to As(V) with 31% As(III) being oxidized into As(V) within 192 h under aerobic conditions. Although the produced As(V) was detected in the solid phase, As(V) concentration was below the detection limit in the aqueous phase during experiments (<1.0 lg/L, data not shown). It was inferred that As(III) oxidation would occur on the surface of MGNS after adsorption. Therefore, removal of As(III) by MGNS was believed to include As(III) adsorption on the surface as the first step, followed by a fast heterogeneous oxidation step from As(III) to As(V). This finding was in agreement with other previous studies, which showed that the oxidation of As(III) by mixed Fe(II)/ Fe(III) occurred under both anoxic conditions (Amstaetter et al., 2010) and oxic conditions (Katsoyiannis et al., 2008; OnaNguema et al., 2010). However, the oxidation mechanism was different. Under anoxic conditions, As(III) was oxidized via the reactive Fe(III) intermediate formed from oxidation of Fe(II) by electron transfer into the bulk Fe(III) phase (Amstaetter et al., 2010). Whereas, As(III) oxidation occurred by pH-dependent reactive oxygen species formed during reduction of O2 by Fe(II) under oxic conditions (Hug and Leupin, 2003; Katsoyiannis et al., 2008; Ona-Nguema et al., 2010). In this study, under oxic and neutral pH conditions, the reaction of Fe(II) with O2 was expected to occur through a series of 1-electron transfer (Eqs. (7)–(9)) (Pang et al., 2011). It was verified by Fe(II) and total Fe contents of MGNS before and after adsorption, which indicated that proportion of Fe(II) in matrix Fe of the solid decreased from 33% to around 10% after adsorption. Hydroxyl radical (HO) was believed to be the crucial Fenton intermediate to oxidize As(III). Thus, As(III) oxidation on MGNS was likely due to Fenton-like reactions during adsorption. Furthermore, the importance of the heterogeneous Fe surface reactions and the trace elements like Mn in natural materials cannot be ruled out in this process.

FeðIIÞ þ O2 ! FeðIIIÞ þ O 2

ð7Þ

þ FeðIIÞ þ O 2 þ 2H ! FeðIIIÞ þ H2 O2

ð8Þ

FeðIIÞ þ H2 O2 ! FeðIIIÞ þ HO þ HO

ð9Þ

4. Conclusions Modified granular natural siderite (MGNS) adsorbent was fabricated using an extrusion granulation method. Batch and column results showed that MGNS effectively removed As(III) from the aqueous environment. Arsenic(III) adsorption on MGNS was

192

K. Zhao et al. / Applied Geochemistry 48 (2014) 184–192

favorable and increased with increasing reaction temperature. Solution initial pH between 3.0 and 10.0 showed no significant effect on As(III) adsorption. Background electrolytes and the presence of coexisting anions less affected As(III) adsorption on MGNS. Column tests demonstrated that MGNS, as filter fillings, effectively removed As from both synthesized As(III) water and natural highAs groundwater. The adsorbent, having a high specific surface area, was considered to be an Fe(II)/(III) hybrid system with amorphous spherical aggregates attached to the siderite matrix. Adsorption and heterogeneous oxidation of As(III) would be involved in the removal process, which was believed to be the main mechanisms of As(III) removal by MGNS. Therefore, the low-cost MGNS (i.e., $200/t natural siderite, relative to $1800/t synthetic siderite) appeared to be a potential and promising adsorbent for remediating As(III)-contaminated groundwater. Acknowledgements The study has been financially supported by National Natural Science Foundation of China (Nos. 41222020 and 41172224), the Fundamental Research Funds for the Central Universities (No. 2652013028), and the Fok Ying-Tung Education Foundation, China (Grant No. 131017). The authors would like to thank the Shanghai Synchrotron Radiation Facility (Beamline BL14W) and its staff (Y. Huang and Z. Jiang) for allowing us to perform the XANES analysis. References Amonette, J.E., Templeton, J.C., 1998. Improvements to the quantitative assay of nonrefractory minerals for Fe(II) and total Fe using 1,10 phenanthroline. Clays Clay. Miner. 46, 51–62. Amstaetter, K., Borch, T., Larese-Casanova, P., Kapppler, A., 2010. Redox transformation of arsenic by Fe(II)-activated goethite (a-FeOOH). Environ. Sci. Technol. 44, 102–108. Andradae, S., Hypolito, R., Ulbrich, H.H., Silva, M.L., 2002. Iron(II) oxide determination in rocks and minerals. Chem. Geol. 182, 85–89. Badruzzaman, M., Westerhoff, P., Knappe, D.R.U., 2004. Intraparticle diffusion and adsorption of arsenate onto granular ferric hydroxide (GFH). Water Res. 38, 4002–4012. Berg, M., Stengel, C., Trang, P.T.K., Viet, P.H., Sampson, M.L., Leng, M., Samreth, S., Fredericks, D., 2007. Magnitude of arsenic pollution in the Mekong and Red River Deltas — Cambodia and Vietnam. Sci. Total Environ. 372, 413–425. Chowdhury, S.R., Yanful, E.K., 2011. Arsenic removal from aqueous solutions by adsorption on magnetite nanoparticles. Water Environ. J. 25, 429–437. Deng, Y.M., Wang, Y.X., Ma, T., Gan, Y.Q., 2009. Speciation and enrichment of arsenic in strongly reducing shallow aquifers at western Hetao Plain, northern China. Environ. Geol. 56, 1467–1477. Dixit, S., Hering, J.G., 2003. Comparison of arsenic(V) and arsenic(III) sorption onto iron oxide minerals: implications for arsenic mobility. Environ. Sci. Technol. 37, 4182–4189. Foo, K.Y., Hameed, B.H., 2010. Insights into the modeling of adsorption isotherm systems. Chem. Eng. J. 156, 2–10. Giménez, J., Martínez, M., de Pablo, J., Rovira, M., Duro, L., 2007. Arsenic sorption onto natural hematite, magnetite, and goethite. J. Hazard. Mater. 141, 575–580. Gu, Z.M., Fang, J., Deng, B.L., 2005. Preparation and evaluation of GAC-based iron-containing adsorbents for arsenic removal. Environ. Sci. Technol. 39, 3833–3843. Guo, H.M., Li, Y., Zhao, K., 2010. Arsenate removal from aqueous solution using synthetic siderite. J. Hazard. Mater. 176, 174–180. Guo, H.M., Li, Y., Zhao, K., Ren, Y., Wei, C., 2011. Removal of arsenite from water by synthetic siderite: behaviors and mechanisms. J. Hazard. Mater. 186, 1847–1854. Guo, H.M., Ren, Y., Liu, Q., Zhao, K., Li, Y., 2013. Enhancement of arsenic adsorption during mineral transformation from siderite to goethite: mechanism and application. Environ. Sci. Technol. 47, 1009–1016. Guo, H.M., Stüben, D., Berner, Z., 2007a. Adsorption of arsenic(III) and arsenic(V) from groundwater using natural siderite as the adsorbent. J. Colloid Interface Sci. 315, 47–53. Guo, H.M., Stüben, D., Berner, Z., 2007b. Removal of arsenic from aqueous solution by natural siderite and hematite. Appl. Geochem. 22, 1039–1051. Guo, H.M., Wen, D.G., Liu, Z.Y., Jia, Y.F., Guo, Q., 2014. A review of high arsenic groundwater in Mainland and Taiwan, China: distribution, characteristics and geochemical processes. Appl. Geochem. 41, 196–217. Guo, H.M., Yang, S.Z., Tang, X.H., Li, Y., Shen, Z.L., 2008. Groundwater geochemistry and its implications for arsenic mobilization in shallow aquifers of the Hetao Basin. Inner Mongolia. Sci. Total Environ. 393, 131–144.

Guo, H.M., Zhang, Y., Xing, L.N., Jia, Y.F., 2012. Spatial variation in arsenic and fluoride concentrations of shallow groundwater from the Shahai town of the Hetao basin, Inner Mongolia. Appl. Geochem. 27, 2187–2196. Hug, S.J., Leupin, O., 2003. Iron-catalyzed oxidation of arsenic(III) by oxygen and hydrogen peroxide: pH-dependent formation of oxidants in the Fenton reaction. Environ. Sci. Technol. 37, 2734–2742. Hughes, M.F., Beck, B.D., Chen, Y., Lewis, A.S., Thomas, D.J., 2011. Arsenic exposure and toxicology: a historical perspective. Toxicol. Sci. 123, 305–332. Jönsson, J., Sherman, D.M., 2008. Sorption of As(III) and As(V) to siderite, green rust (fougerite) and magnetite: implications for arsenic release in anoxic groundwaters. Chem. Geol. 255, 173–181. Kanel, S.R., Manning, B., Charlet, L., Choi, H., 2005. Removal of arsenic(III) from groundwater by nanoscale zero-valent iron. Environ. Sci. Technol. 39, 1291–1298. Katsoyiannis, I.A., Ruettimann, T., Hug, S.J., 2008. PH dependence of Fenton reagent generation and As(III) oxidation and removal by corrosion of zero valent iron in aerated water. Environ. Sci. Technol. 42, 7424–7430. Kundu, S., Gupta, A.K., 2007. Adsorption characteristics of As(III) from aqueous solution on iron oxide coated cement (IOCC). J. Hazard. Mater. 142, 97–104. Liu, Y., Zheng, B.H., Fu, Q., Meng, W., Wang, Y.Y., 2009. Risk assessment and management of arsenic in source water in China. J. Hazard. Mater. 170 (2–3), 729–734. Luengo, C., Brigante, M., Avena, M., 2007. Adsorption kinetics of phosphate and arsenate on goethite. A comparative study. J. Colloid Interface Sci. 311, 354–360. Meng, X.G., Korfiatis, G.P., Bang, S.B., Bang, K.W., 2002. Combined effects of anions on arsenic removal by iron hydroxides. Toxicol. Lett. 133, 103–111. Mihaljevic, M., Ettler, V., Sisr, L., Sebek, O., Strnad, L., Vonaskova, V., 2009. Effect of low concentrations of phosphate ions on extraction of arsenic from naturally contaminated soil. Bull. Environ. Contam. Toxicol. 83, 422–427. Ministry of Health of PR China, 2006. Standard for drinking water quality, GB57492006 (in Chinese). Mohan, D., Pittman Jr., C.U., 2007. Arsenic removal from water/wastewater using adsorbents – a critical review. J. Hazard. Mater. 142, 1–53. Muñiz, G., Fierro, V., Celzard, A., Furdin, G., Gonzalez-Sanchez, G., Ballinas, M.L., 2009. Synthesis, characterization and performance in arsenic removal of iron doped activated carbons prepared by impregnation with Fe(III) and Fe(II). J. Hazard. Mater. 165, 893–902. Nordstrom, D.K., 2002. Worldwide occurrences of arsenic in ground water. Science 296, 2143–2145. Ohe, K., Tagai, Y., Nakamura, S., Oshima, T., Baba, Y., 2005. Adsorption behavior of arsenic(III) and arsenic(V) using magnetite. J. Chem. Eng. Jpn. 38, 671–676. Ona-Nguema, G., Morin, G., Wang, Y.H., Foster, A.L., Juillot, F., Calas, G., Brown, G.E., 2010. XANES evidence for rapid arsenic(III) oxidation at magnetite and ferrihydrite surfaces by dissolved O2 via Fe2+-mediated reactions. Environ. Sci. Technol. 44, 5416–5422. Pang, S.Y., Jiang, J., Ma, J., 2011. Oxidation of sulfoxides and arsenic(III) in corrosion of nanoscale zero valent iron by oxygen: evidence against ferryl ions (Fe(IV)) as active intermediates in Fenton reaction. Environ. Sci. Technol. 45, 307–312. Raven, K.P., Jain, A., Loeppert, R.H., 1998. Arsenite and arsenate adsorption on ferrihydrite: kinetics, equilibrium, and adsorption envelopes. Environ. Sci. Technol. 32, 344–349. Roberts, L.C., Hug, S.J., Ruettimann, T., Billah, M., Khan, A.W., Rahman, M.T., 2004. Arsenic removal with iron(II) and iron(III) waters with high silicate and phosphate concentrations. Environ. Sci. Technol. 38, 307–315. Smedley, P.L., Kinniburgh, D.G., 2002. A review of the source, behaviour and distribution of arsenic in natural waters. Appl. Geochem. 17, 517–568. Stachowicz, M., Hiemstra, T., van Riemsdijk, W.H., 2008. Multi-competitive interaction of As(III) and As(V) oxyanions with Ca2+, Mg2+, PO43, and CO32 ions on goethite. J. Colloid Interface Sci. 320, 400–414. Su, C.M., Puls, R.W., 2001. Arsenate and arsenite removal by zerovalent iron: kinetics, redox transformation, and implications for in situ groundwater remediation. Environ. Sci. Technol. 35, 1487–1492. Tchounwou, P.B., Patlolla, A.K., Centeno, J.A., 2003. Carcinogenic and systemic health effects associated with arsenic exposure – a critical review. Toxicol. Pathol. 31, 575–588. Thirunavukkarasu, O.S., Viraraghavan, T., Subramanian, K.S., 2003a. Arsenic removal from drinking water using granular ferric hydroxide. Water SA 29, 161–170. Thirunavukkarasu, O.S., Viraraghavan, T., Subramanian, K.S., 2003b. Arsenic removal from drinking water using iron oxide-coated sand. Water, Air, Soil Pollut. 142, 95–111. Viraraghavan, T., Subramanian, K.S., Arduldoss, J.A., 1999. Arsenic in drinking water problems and solutions. Water Sci. Technol. 40, 69–76. Weber, W.J., Morris, J.C., 1963. Kinetics of adsorption on carbon solution. J. Sanit. Eng. Div. Proc. Am. Soc. Civil Eng. 89, 31–59. World Health Organization (WHO), 1996. Guidelines for drinking-water quality, health criteria and other supporting information, second ed., vol. 2, WHO, Geneva, Switzerland, pp. 940–949. Zhang, W., Singh, P., Paling, E., Delides, S., 2004. Arsenic removal from contaminated water by natural iron ores. Miner. Eng. 17, 517–524. Zhao, K., Guo, H.M., 2014. Behavior and mechanism of arsenate adsorption on activated natural siderite: evidences from FTIR and XANES analysis. Environ. Sci. Pollut. Res. 21, 1944–1953. Zhao, Z.X., Jia, Y.F., Xu, L.Y., Zhao, S.L., 2011. Adsorption and heterogeneous oxidation of As(III) on ferrihydrite. Water Res. 45, 6496–6504.