Arsenic removal from aqueous solutions by adsorption onto hydrous iron oxide-impregnated alginate beads

G Model

JIEC-2785; No. of Pages 10 Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec

Arsenic removal from aqueous solutions by adsorption onto hydrous iron oxide-impregnated alginate beads Abinashi Sigdel, Jeongwon Park, Hyoeun Kwak, Pyung-Kyu Park * Department of Environmental Engineering, Yonsei University, 1 Yonseidae-gil, Wonju 220-710, Gangwon-do, Republic of Korea

A R T I C L E I N F O

Article history: Received 8 September 2015 Received in revised form 4 January 2016 Accepted 4 January 2016 Available online xxx Keywords: Adsorption Alginate Arsenic Impregnation Iron oxide

A B S T R A C T

Hydrous iron oxide impregnated alginate beads were developed for effective arsenic removal from water. As(III) adsorption was maximized at neutral pH while As(V) adsorption was higher in acidic conditions. Adsorption efficiency for both As(III) and As(V) mostly increased with increasing iron loading, but As(V) adsorption slightly decreased at high iron loading. Phosphate showed a pronounced interfering effect, especially at high concentration. Kinetics data fitted to pseudo-second-order and intraparticle diffusion model suggested chemisorption and intra-particle diffusion might mainly govern As(III) and As(V) adsorption, respectively. Beads were regenerated using NaOH solution and successfully reused for multiple cycles. ß 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

Introduction Arsenic, which is ubiquitous in nature, is a serious worldwide concern due to its high toxicity and carcinogenicity [1]. Long-term exposure to arsenic can cause skin lesions, peripheral neuropathy, gastrointestinal symptoms, diabetes, renal damage, cardiovascular disease, and various types of cancer [2]. Humans are exposed to arsenic mainly through ingestion of arsenic-contaminated water and foods from natural and anthropogenic sources [3]. Naturally occurring arsenic contamination of groundwater has been discovered in at least 70 countries, and an estimated 140 million people are at risk of consuming contaminated water [4]. Arsenic usually occurs in inorganic form as arsenite and arsenate in natural water, which are the oxyanions of trivalent As(III) and pentavalent As(V), respectively. Arsenite is 20 to 60 times more toxic than arsenate and more mobile in the environment [5]. However, typical arsenic treatment techniques such as anion exchange and coagulation with ferric salts are less effective to remove As(III) because As(III) mostly occurs as uncharged arsenious acid, H3As(III)O3, in the natural environment [6]. When water contains appreciable amount of As(III), the contaminated water is usually pretreated with oxidants to convert As(III) to As(V) before applying the treatment techniques. Among

* Corresponding author. Tel.: +82 33 760 2890; fax: +82 33 760 2571. E-mail address: [email protected] (P.-K. Park).

various arsenic treatment technologies, adsorption is one of the most effective methods for removing As(III) and/or As(V) [6,7]. Many previous studies have revealed that iron oxides and hydrous iron oxides (iron oxide-hydroxide) have high affinity toward both As(III) and As(V) species [8,9]. Mayo et al. [10] found that the adsorption capacities of magnetite particles for both As(III) and As(V) was increased nearly 200 times when the particle size decreased from 300 to 12 nm. The decrease in particle size increased specific surface area of the magnetite, prevented agglomeration of magnetite particles in water and improved solute transport into the interior surface of the magnetite. Iron oxides in the amorphous phase are known to exhibit higher adsorption capacity than other phases due to their large surface areas [11]. The small particle size and high reactivity of iron oxides, however, make it difficult to separate iron particles from reactors after adsorption in full-scale applications [12,13]. Such limitations have been overcome by coating, doping, or packing iron oxides and hydrous iron oxides on support materials. The use of iron oxidecontaining sand, zeolite, and activated carbon have been well documented [9,14,15,16,17]. Polymer-supported metal oxide based composite adsorbents have recently gained considerable attention for heavy metal removal due to their biocompatibility, water permeability, and ability to load large amounts of solid particles. Cho et al. [18] encapsulated akaganeite in alginate beads for arsenic, Lv et al. [19] immobilized Fe(0)–Fe3O4 in polyvinyl alcohol-alginate for chromium(VI), Cho et al. [20] immobilized clay-magnetite in chitosan for arsenic(V) and copper(II), and Pan

http://dx.doi.org/10.1016/j.jiec.2016.01.005 1226-086X/ß 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

Please cite this article in press as: A. Sigdel, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.01.005

G Model

JIEC-2785; No. of Pages 10 2

A. Sigdel et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

et al. [21] impregnated hydrous iron oxide in a cation exchange resin for heavy metal adsorption. Ocin´ski et al. [22] immobilized iron oxide in polystyrene-divinylbenzene for arsenic removal, but observed a low As(V) adsorption rate for the supported beads. Major limitations of these studies include low adsorption capacity due to, for instance, low iron loading, and little information on the regenerability and reusability of adsorbents considering the potential for practical applications. The objective of this study was to develop hydrous iron oxideimpregnated alginate beads (HIO-alginate beads) for effective removal of inorganic arsenic from water. Amorphous hydrous iron oxides prepared by chemical precipitation method were homogeneously dispersed into alginate gels. Composite beads were then synthesized from the gel matrix and finally dried on air. Beads were characterized with scanning electron microscopy/energy dispersive X-ray spectroscopy (SEM-EDX), transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS) and N2 adsorption-desorption analysis. The arsenic adsorption efficiency of beads was investigated as a function of contact time, pH, iron loading, and concentrations of competing ions in batch experiments. Studies were performed to investigate the effects of interfering ions on arsenic adsorption. Kinetic and equilibrium studies were applied to analyze the mechanisms of arsenic adsorption by the prepared beads. The capability for regeneration and reusability of HIO-alginate beads were also examined.

Experimental Chemical reagents All chemicals were of reagent grade. Sodium meta-arsenite (NaAs(III)O2, >98%), sodium arsenate dibasic heptahydrate (Na2HAs(V)O47H2O, >98%), and sodium alginate (medium viscosity grade) were purchased from Sigma-Aldrich, USA. Stock solutions of arsenite and arsenate were prepared by dissolving their respective salts in deionized water. Ferric nitrate nonahydrate (Fe(NO3)39H2O, >98%, Junsei, Japan) was used to synthesize iron oxides. Other reagents were purchased from Duksan, Republic of Korea. Synthesis of hydrous iron oxide-impregnated alginate beads Hydrous iron oxide was prepared by chemical precipitation of ferric nitrate according to the method described by Wilkie and Hering [23]. Hydrous iron oxide precipitate was formed at pH 7.5– 8.0 by slowly adding 0.5 M NaOH to 300 mL of 0.02 M ferric nitrate. The precipitate was washed 3 times with deionized (DI) water to remove water soluble ions from the precipitate. The produced hydrous iron oxide was immobilized using alginate with calcium ion via entrapment. Briefly, 3 wt% alginate gel matrix was prepared. A suspension of hydrous iron oxide precipitate containing 0.34 g of Fe was added into the 100 mL of the alginate gel matrix, and then DI water was added until the final volume of the mixture was 150 mL. The mixture was uniformly mixed with a magnetic stirrer for 4 h. The mixture contained sodium alginate and iron in a proportion of 9:1 (3 g alginate/0.34 g Fe) for 10 wt% iron loading. Finally, HIO-alginate beads were produced from this homogeneous mixture through crosslinking of sodium alginate with calcium chloride (3%) solution. Beads were cured in the calcium chloride solution for 6 h to complete gelation, washed several times with DI water to remove water soluble compounds, and finally air-dried for 48 h. Only dry beads were used in the subsequent adsorption experiments. HIO-alginate beads with a higher proportion of iron (up to 30 wt%) were produced using a similar procedure.

Characterization and analytical techniques The SEM images and approximate elemental composition of HIO-alginate beads were obtained with SEM-EDX analysis (Ametek, Inc., USA). The morphology of beads was observed with a TEM (JEM-ARM200F, JEOL), which was operated at 200 kV. Characterization of porous structure of HIO-alginate beads was performed by the method of N2 adsorption-desorption at 77 K using an ASAP 2010 (Micromeritics, USA). XPS spectra of HIOalginate beads were acquired using an X-ray photoelectron spectroscopy (Thermo VG, UK) with a monochromatized AlKa Xray source at 12 kV (1486.6 eV). Arsenic concentrations in solutions were determined in terms of As(III) or As(V) using inductively coupled plasma (ICP) spectrometer (IRIS/AP, ThermoJarrel Ash Corp., USA). Adsorption experiments Batch experiments were carried out separately on As(III) or As(V) solutions at 23  1 8C in a 50-mL glass vial. The initial pH was adjusted by adding 0.1 M HCl or 0.1 M NaOH to arsenic solution. A total of 40 mL arsenic solution was shaken with 0.040 g of HIOalginate beads (1.0 g/L) on an orbital shaker at 100 rpm. Aliquots of the samples were taken at certain time intervals to investigate As(III) or As(V) adsorption onto the beads. Effect of pH The effect of pH on As(III) and As(V) adsorption onto beads was investigated. Initial pH was varied within the range of 2–12 by adding 0.1 M NaOH or HCl dropwise to 10 mg/L of As(III) or As(V) solutions. No visible change in the color of the solution was observed, indicating that immobilized iron should not leach from beads. Effect of iron loading in beads Beads with different iron loadings from 10 to 30 wt% were also tested to elucidate the role of hydrous iron oxides in uptake of arsenic species. Initial solution pH was adjusted to 6.0  0.2. Effect of competing ions Various common anions found in surface and ground water may compete with arsenic for adsorption sites. As(III) and As(V) adsorption onto HIO-alginate beads was investigated in the presence of sulfate, silicate, phosphate, bicarbonate, chloride, or nitrate ion at different concentrations of 0.2 and 2.0 mM. Kinetic study A kinetics experiment was conducted to investigate the adsorption rate of As(III) or As(V) onto HIO-alginate beads. Aqueous solutions of As(III) and As(V) with initial concentrations of 2.5 mg/L or 10 mg/L were contacted with HIO-alginate beads for 3–240 h. The solution concentration was monitored over time by collecting samples of each bottle at fixed time intervals. For a blank, the same experiment was performed using bare alginate beads impregnating no iron oxides. Data were fitted with pseudofirst- and pseudo-second-order models [24,25]. An intra-particle diffusion model proposed by Weber and Morris [26] was also applied to analyze experimental data. Isotherm study An equilibrium adsorption experiment was conducted by varying the initial concentration of As(III) or As(V) in a range of 7–500 mg/L. A plot of arsenic concentration in solid (bead) phase at equilibrium, qe, versus arsenic concentration in solution at equilibrium, Ce, after 7 d was obtained. This plot was further analyzed using Langmuir and Freundlich adsorption isotherms [27,28].

Please cite this article in press as: A. Sigdel, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.01.005

G Model

JIEC-2785; No. of Pages 10 A. Sigdel et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

Regeneration and reuse Eight cycles of regeneration and reuse were carried out to evaluate the reusability of HIO-alginate beads. For initial adsorption, 10 mg/L arsenic solution with initial pH value adjusted to 6 was contacted with 1 g/L of beads for 7 d. For regeneration, beads that had adsorbed arsenic were removed from solution, washed with DI water, and then agitated in 0.05 M NaOH solution for 24 h to desorb arsenic from the beads. NaOH concentration and desorption time were determined after the preliminary studies with 0.01–0.2 M NaOH solution for 10–48 h. The adsorption efficiency being highly dependent on pH, beads were neutralized with dilute mineral acid (0.01 M H2SO4). Regenerated beads were washed with DI water and dried again for reuse in the next adsorption experiment under the same conditions as the initial adsorption. Regeneration and reuse cycles were repeated for the next 8 consecutive cycles. Adsorption efficiency after each regeneration and reuse cycle was compared to the efficiency of fresh beads at the initial adsorption experiment.

3

hysteresis was found in N2 adsorption-desorption isotherms at P/ P0 of 0.4 to 0.6, and the average pore size was 2.6 nm (Fig. S2 in Supporting information), which implies the hydrous iron oxides prepared in this study has mesoporous structure [32,33]. Effect of various parameters on arsenic adsorption Contact time The adsorption of arsenic on HIO-alginate beads was investigated as a function of contact time at two different initial arsenic concentrations (2.5 and 10 mg/L as As). Fig. 4 and Table 2 show that As(III) removal was higher than As(V) removal with the same contact time. Similar results were reported for adsorptive removal of arsenic by amorphous iron oxides in the near-neutral pH range [13,30]. Equilibrium was reached around 7 d for As(III), and the rate of increase in the solid phase concentration, qe, was steeply reduced after 7 d for As(V). Various researchers have recognized that many previous experiments with shorter equilibration time may not have reached equilibrium [34,35]. Contact time was fixed as 7 d in the following studies.

Results and discussion pH Characterization of HIO-alginate beads The surface morphology of dry beads was observed by SEM (Fig. 1). Bare alginate and HIO-alginate beads were nearly spherical in shape with approximate diameters of 1.3 and 1.1 mm, respectively. The bare alginate beads had relatively smooth surfaces (Fig. 1a), whereas the HIO-alginate beads were rough (Fig. 1b). Rough surfaces are favorable for molecular diffusion [29] and could provide a larger surface area for the adsorption of contaminants. As shown in Fig. 1c, the elemental analysis of HIOalginate beads confirmed that the beads contained iron and calcium but a very small amount of sodium. These results imply that most sodium ions included in the sodium alginate matrix were released into water during the crosslinking reaction of sodium alginate with calcium chloride. The TEM image of a 10 wt% HIO-alginate bead is shown in Fig. 2. Hydrous iron oxides, which appear bright in the image, were homogeneously dispersed in alginate matrix in the bead. The size of the oxide particles ranged approximately from 30 to 100 nm. The selected area electron diffraction (SAED) pattern of HIO-alg bead was also measured (Fig. 2b). The SAED pattern showed two bright rings that were spaced at the interplanar spacings of 0.15 nm and 0.25 nm. The measured spacings were typical of 2-line ferrihydrite, indicating the hydrous iron oxide in the beads existed as 2-line ferrihydrite. X-ray photoelectron spectroscopy was used to study the surface chemical composition of HIO-alginate beads. XPS survey scan confirmed the presence of iron on the beads (Fig. 3) (see Fig. S1 in Supporting information for the XPS spectra for survey scan and calcium). The high resolution XPS spectra in the region of Fe 2p showed two peaks at 711.4 and 725.0 eV, which can be assigned to Fe 2p3/2 and Fe 2p1/2 of Fe(III) in hydrous ferric oxides, respectively, as depicted in Fig. 3 because Fe 2p3/2 and Fe 2p1/2 peak of mixedvalent iron oxides are generally lower located [30]. Therefore, the oxidation state of iron in HIO-alginate beads was confirmed to be +3, and thus the hydrous iron oxide could be classified as 2-line ferrihydrite. The BET surface area of hydrous iron oxides was 348 m2/g (Table 1), which is very comparable to those of amorphous iron oxides reported by Lin et al. [31]. The high surface area means the hydrous iron oxides existed in amorphous to poorly crystalline phase and had nano-sized particles, leading to high adsorption capacity. The BET surface area of 10 wt%, 20 wt% and 30 wt% HIOalginate beads was 0.26, 1.04 and 0.93 m2/g, respectively. A small

Fig. 5 shows the arsenic adsorption of HIO-alginate beads as a function of the initial and final pH values of the solution. Final pH, measured after 7 d of adsorption, shifted toward neutral pH to some degree in all experiments because the carboxyl group of alginate tends to be protonated at low pH while the iron oxide surface tends to be deprotonated at high pH [36]. When initial pH was near neutral, the pH drift during adsorption experiments was minimized. As shown in Fig. 5a, As(III) adsorption steadily increased with higher initial pH up to pH 4, remained nearly constant in the pH range of 4–10, and then decreased with further increases in pH. On the other hand, As(V) adsorption was high with acidic conditions but decreased with increased pH (Fig. 5b). To confirm the stability of beads, the amount of iron released into the experimental solution through the beads was also measured especially at pH 2. After 7 days of contact at pH 2, only 1.5% and 1.3% of the immobilized iron were released from 10 wt% and 30 wt% HIO-alginate beads, respectively. Solution pH has a significant influence on adsorption because the surface charge properties of adsorbent and solute speciation are linked to the pH value of solution [37]. As(V) ion occurs mainly as H2AsO4 in the pH range 3–6, HAsO42 occurs at pH 8–10.5, and AsO43 (arsenate) occurs at higher pH [38]. At pH values above pzc, HIO-alginate beads would have a net negative surface charge. Since the pzc for HIO-alginate beads during adsorption experiments was around 7 (Fig. S3 in Supporting information), lower As(V) adsorption at pH higher than 7 may be linked to repulsion between the negatively charged surface and As(V) anion species [39]. As mentioned above, As(III) adsorption decreased at both low and high pH values. A similar behavior was observed for As(III) adsorption on amorphous iron oxide [40] and nanostructured Fe– Cu binary oxide [39]. Stumm [37] reported that adsorption of acid anions by metal oxides reaches a maximum at pH values similar to the pKa1 of the acid, i.e., 9.2 for arsenious acid (H3As(III)O3). For polyprotic weak acid anions having high pKa such as arsenious acid, ligand adsorption onto metal oxides gradually increases with increasing pH, reaching a maximum when solution pH approximately equals the pKa [41]. Cumbal and SenGupta [42] reported that non-ionized arsenious acid (H3AsO30) was most favorably adsorbed onto non-ionized sites. Non-ionized sites of variable charged oxide surface increases with increasing pH and reaches a maximum at point of zero charge (pzc), which corresponds to pH value of 7 to 9 for hydrous iron oxides. The sharp decrease of As(III) adsorption at pH over 9.2 may be due to coulombic repulsion of

Please cite this article in press as: A. Sigdel, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.01.005

G Model

JIEC-2785; No. of Pages 10 4

A. Sigdel et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

Fig. 1. (a) SEM image of a bare alginate bead, and (b) SEM image and (c) SEM-EDX spectrum of an HIO-alginate bead.

negatively charged predominant As(III) species such as H2AsO3, HAsO32, and AsO33 on deprotonated HIO-alginate bead surfaces [39]. Considering the acceptable adsorption of both of As(III) and As(V), as well as minimization of the pH shift between initial and final pH values, initial pH was fixed at 6 for further experiments.

Iron loading in beads To further investigate the role of iron oxide in arsenic adsorption, beads with different iron loadings of 0–30 wt% were evaluated under the same experimental conditions (initial arsenic (As(III) or As(V)) concentration = 10 mg/L as As; adsorbent dose = 1 g/L; contact time = 7 d; initial pH = 6.0  0.2;

Please cite this article in press as: A. Sigdel, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.01.005

G Model

JIEC-2785; No. of Pages 10 A. Sigdel et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

5

Table 1 BET surface area of adsorbents prepared in this study. BET surface area (m2/g)

Adsorbent *

HIO HIO-alg bead (10 wt% iron loading) HIO-alg bead (20 wt% iron loading) HIO-alg bead (30 wt% iron loading) *

348.09 0.26 1.05 0.93

Hydrous iron oxide without impregnation in alginate beads.

hydrous iron oxides as colloids responsible for the decrease in porosity of the calcium alginate at high concentration, which is in agreement with the results of decrease in BET surface area of beads with iron oxide loadings at very high loading. The reduction in porosity at 30 wt% iron loading finally caused hindrance to molecular diffusion of arsenic species inside of the beads.

Fig. 2. (a) TEM image and (b) SAED pattern of an HIO-alginate bead.

temperature = 23  1 8C). No arsenic adsorption by bare alginate beads indicates that alginate itself acts only as a support matrix for hydrous iron oxide particles. The results for As(III) indicate that higher removal efficiency was achieved by increasing iron loading, but the rate of increase in removal was slightly reduced when iron loading in beads was above 20 wt% (Fig. 6a). In addition, As(V) adsorption by 30 wt% iron loaded beads was slightly less than that on 20 wt% beads (Fig. 6b). These results imply that adsorption was hindered at higher loadings. The increase in adsorption with increasing iron oxide loading was attributed to enhanced amount of adsorption sites. The BET surface area of 30 wt% iron loaded beads, however, was less than that of 20 wt% as shown in Table 1, which is a clue to the reason why the As(V) adsorption by 30 wt% beads was slightly reduced. The hydrous iron oxides had high surface area, and thus the increase in their contents would result in the increase in surface area of whole HIO-alginate beads. The increase in the concentration of particles/colloids within porous structure, however, also would lead to the reduction in porosity [43]. We can consider the

Competing ions Both surface and ground water contain soluble ions, some of which may lower adsorption site density due to competitive effects. These anions are of particular importance for arsenic in water phase because they may occupy potential sites by forming inner-sphere complexes and reduce the interactive potential of arsenic species with iron oxide surfaces [44]. The effect of co-existing anions on arsenic adsorption onto beads was investigated as summarized in Fig. 7. Inorganic anions such as sulfate, bicarbonate, carbonate, chloride, and nitrate did not show any significant interfering effect on arsenic adsorption. However, phosphate anions showed a strong competitive effect in the adsorption process, reducing the amount of both adsorbed As(III) and As(V). Compared to experiment with arsenic-spiked DI water as a control process, As(III) and As(V) adsorption decreased by 14% and 12%, respectively, at initial phosphate concentration of 0.2 mM, and by 31% and 54%, respectively, at initial phosphate concentration of 2.0 mM. Silicate ions showed a slight competitive effect for only As(III) when initial silicate concentration was as high as 2.0 mM. Previous studies show competitive effect of phosphate and silicate on arsenic adsorption [45–47]. This interference effect is likely explained by the chemical similarity of arsenate and phosphate in aqueous solution, as both ions form stable innersphere complexes with surface hydroxyl groups of iron oxides.

10000

Fe2p1/2

Fe2p3/2

Counts /s

9000

8000

7000

6000

5000 740

730

720

710

700

Bind ing energy (eV) Fig. 3. XPS spectrum of an HIO-alginate bead in the region of Fe 2p.

Please cite this article in press as: A. Sigdel, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.01.005

G Model

JIEC-2785; No. of Pages 10 6

A. Sigdel et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

8

8

(a)

10 mg/L 2.5 mg/L

10 mg/L 2.5 mg/L

4

qe

4

(mg/g)

6

qe

(mg/g)

6

(b)

2

2

0

0 0

2

4

6

8

10

0

2

4

Time (d)

6

8

10

Time (d)

Fig. 4. Variation of (a) As(III) and (b) As(V) concentrations in solid phase, qe (mg/g of beads), during adsorption by HIO-alginate beads for initial concentration of 2.5 and 10 mg/ L as As (bead dosage = 1 g/L; iron loading in beads = 10 wt%; initial pH = 6.0  0.2; temperature = 23  1 8C) (n = 2; all the measurements were duplicated).

Table 2 Kinetics parameters for As(III) and As(V) adsorption onto HIO-alginate beads. 2.5 mg/L

Parameter

10 mg/L As(V)

As(III) A. Pseudo-first-order [ln(qe  qt) = ln qe  k1 t] k1 (h1) qe,experimental (mg/g) qe,model (mg/g) R2 B. Pseudo-second-order [t/qt = 1/(k2qe2) + t/qe] k2 (g/(mg h) qe,model (mg/g) t1/2 (h) R2 C. Weber and Morris intraparticle diffusion [q = KHt + C] K (mg/(g h0.5) C (mg/g) R2

0.009 1.53 1.42 0.985

0.022 6.21 5.56 0.995

0.012 4.03 4.44 0.954

0.028 2.06 17.4 0.999

0.007 1.74 79 0.995

0.005 6.95 26.6 0.999

0.001 6.65 159 0.962

0.103 0.543 0.867

0.105 0.117 0.996

0.378 1.083 0.917

0.303 0.586 0.996

10 Initial pH

qe (mg/g), As(V)

qe (mg/g), As(III)

8

As(V)

0.018 1.94 1.23 0.956

10

(a)

As(III)

Final pH

6 4 2

(b)

8

Initial pH Final pH

6 4 2 0

0 0

2

4

6

8

10

12

0

14

2

4

6

8

10

12

14

pH

pH

Fig. 5. Influence of solution pH on (a) As(III) and (b) As(V) adsorption to HIO-alginate beads. Solid line is for data plotted considering initial pH, and dotted line is for data considering final pH (initial arsenic (As(III) or As(V)) concentration = 10 mg/L as As; bead dose = 1 g/L; iron loading in beads = 10 wt%; contact time = 7 d; temperature = 23  1 8C).

100

(a)

% As(V) adsorption

% As(III) adsorption

100 80 60 40 20

(b)

80 60 40 20 0

0 0

10

20

% Fe(III) loading

30

0

10

20

30

% Fe(III) loading

Fig. 6. Effect of iron loading in HIO-alginate beads for (a) As(III) and (b) As(V) adsorption (initial arsenic (As(III) or As(V)) concentration = 10 mg/L as As; bead dose = 1 g/L; contact time = 7 d; initial pH = 6.0  0.2; temperature = 23  1 8C) (n = 2).

Please cite this article in press as: A. Sigdel, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.01.005

G Model

JIEC-2785; No. of Pages 10 A. Sigdel et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

60

80 0.2 mM 2.0 mM

40 20

60

(b)

0.2 mM 2.0 mM

40 20 0 3

D

Iw

3

D Iw at N er a 2S N O4 a 2S N iO aH 3 2P N O4 aH C O C 3 aC N l2 aN O N a 3 2C O

0

at N er a 2S N O4 a 2S N iO aH 3 2P N O4 aH C O C 3 aC N l2 aN O N a 3 2C O

(a)

As(V) removal (%)

As(III) removal (%)

80

7

Fig. 7. Effect of competing ions on (a) As(III) and (b) As(V) adsorption on HIO-alginate beads (initial arsenic (As(III) or As(V)) concentration = 10 mg/L as As; bead dose = 1 g/L; iron loading in beads = 10 wt%; contact time = 7 d; initial pH = 6.0  0.2; temperature = 23  1 8C).

Arsenate, phosphate, and silicate are all tetrahedral anions that compete with each other for active adsorption sites on bead surfaces, resulting in decreased adsorption capacity. Another previous study reported that As(III) and As(V) can form innersphere complexes as well as outer-sphere ones with HIOs. In contrast, most background electrolytes other than phosphate are generally held in outer-sphere region by electrostatic interactions, and thus their competition for functional sites of the adsorbent inhibits arsenic adsorption only via outer-sphere complexation mechanism [48]. In this study, As(III) and As(V) removal remained either constant or slightly increased in the presence of the noncompeting ions. The results imply that the main mechanism for arsenic adsorption may be inner-sphere complexation and thus the inhibition via outer-sphere complexation mechanism was not significant, leading the arsenic ions have a high affinity for hydrous iron oxides compared to other surrounding ions [49–51]. The increases in arsenate of As(V) adsorption by beads with increasing other non-competing ions (anions and electrolytes) may be attributed to increased ionic strength, leading to less negative potential on the adsorbent surface, which resulted for the slight enhancement of specific adsorption of negatively charged arsenate via inner-sphere complexation [52]. Similar results were obtained for oxyanion adsorption onto iron oxide-based adsorbents [22,53]. The predominant As(III) species being uncharged at neutral pH, their adsorption with diffusion of the uncharged As(III) species to inner-sphere region was less influenced by the background electrolytes and their ionic strength effect, leading neither of increase nor decrease in As(III) adsorption. Our results indicate that co-existing solutes found in natural waters except phosphate show insignificant competitive effects on arsenic removal by prepared HIO-alginate beads. Kinetics study of arsenic adsorption Pseudo-first-order and pseudo-second-order models were employed to analyze the kinetic data in Fig. 4. Kinetic parameters were obtained by fitting the experimental data to linearized kinetic models, as presented in Table 2 (see Figs. S4 and S5 in Supporting information for corresponding plots). According to R2 values, the pseudo-second-order model described the experimental data better than the pseudo-first-order model for both As(V) and As(III). The rate constant of the pseudo-second-order equation, k2, decreased with increased initial arsenic concentration, indicating that arsenic adsorption by beads was slower at high solute concentrations. The k1 and k2 for As(III) adsorption were higher than those for As(V) adsorption under the same experimental conditions, confirming that removal of As(III) by HIO-alginate beads was

faster than removal of As(V). This was also confirmed by comparing the time for half of equilibrium adsorption from the pseudo-second-order model, t1/2, at 10 mg/L initial arsenic concentration. For instance, t1/2 values were 26.6 and 159 h for As(III) and As(V), respectively (Table 2). The high R2 value of 0.999 for As(III) with the pseudo-secondorder model suggest that chemisorption should be the ratelimiting step for As(III) adsorption by the beads [25]. To further examine the mechanism of arsenic absorption, Weber and Morris intra-particle diffusion model [26] was applied to the kinetics data. If ffiffiintra-particle diffusion is the rate-limiting step, a plot of qt versus p t should be a straight line that passes through pffiffi the origin with a small value for parameter C. The qt versus t graph for As(III) adsorption by HIO-alginate beads was multi-linear (Fig. S6 in Supporting information), indicating that intra-particle diffusion was not only rate-limiting for As(III) adsorption. Conversely, kinetics data was well fitted to the intra-particle diffusion model equation with an R2 value of 0.996 for adsorption of both 2.5 and 10 mg/L As(V), suggesting that As(V) adsorption by the beads was mainly governed by intra-particle diffusion. This partly explains why As(V) adsorption was slower than As(III) adsorption: The intra-particle diffusion could be slow through HIO-alginate beads, which affected As(V) adsorption governed by intra-particle diffusion more significantly than As(III) adsorption. Isotherm study of arsenic adsorption To determine the adsorption capacity of HIO-alginate beads, As(III) and As(V) adsorption onto 1.0 g/L of HIO-alginate beads was investigated at different initial arsenic concentrations ranging from 7–430 mg/L and 10–550 mg/L, respectively (Fig. 8), and the results were compared with those obtained for unimpregnated original HIOs (Fig. 9). Adsorption capacity, qe, increased with initial equilibrium arsenic concentration, Ce, as shown in Figs. 8 and 9. The adsorption capacities for HIO-alginate beads were less than those for unimpregnated HIOs possibly because of relatively limited diffusion through the inside of the beads. HIO-alginate beads, however, would have advantages such as better stability of hydrous iron oxides in aqueous phase, easier separation during both operation and regeneration/reuse in real processes, and so on. The regeneration performance of the beads and advantage will be further discussed in Regeneration and reuse section. The adsorption capacity of both impregnated and unimpregnated adsorbents was higher for As(III) than As(V). This less As(V) adsorption on the amorphous iron oxides has previously shown and explained as follows [13,40]: negatively charged As(V) species at neutral pH are removed by iron oxides through ligand exchange by the formation of monodentate complexes at low surface

Please cite this article in press as: A. Sigdel, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.01.005

G Model

JIEC-2785; No. of Pages 10 A. Sigdel et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

8

50

(a) (mg/g), As(V)

40 30 Experimental data Langmuir isotherm Freundlich isotherm

20 10

qe

qe

(mg/g), As(III)

50

0

(b)

40 30 20

Experimental data Langmuir isotherm Freundlich isotherm

10 0

0

100

200

Ce

300

400

0

100

200

300

Ce

(mg/L)

400

500

600

(mg/L)

Fig. 8. Adsorption isotherms for (a) As(III) and (b) As(V) adsorption onto HIO-alginate beads. Solid and dotted lines indicate model fits to Langmuir and Freundlich isotherm equations, respectively (initial arsenic(III) concentration = 7–430 mg/L as As; initial arsenic(V) concentration = 10–550 mg/L as As; bead dose = 1 g/L; iron loading in beads = 10 wt%; contact time = 7 d; initial pH = 6.0  0.2; temperature = 23  1 8C).

400

qe (mg/g)

300

200 As(III) As(V) Langmuir isotherm Freundlich isotherm

100

0 0

100

200

300

400

Ce (mg/L) Fig. 9. Adsorption isotherms for As(III) and As(V) by unimpregnated HIOs. Solid and dotted lines indicate model fits to Langmuir and Freundlich isotherm equations, respectively (initial arsenic(III) concentration = 7–430 mg/L as As; initial arsenic(V) concentration = 10–550 mg/L as As; HIO dose = 1 g/L; contact time = 7 d; initial pH = 6.0  0.2; temperature = 23  1 8C).

coverage whereas, at high surface coverage, they bind to the oxides mainly by the formation of bidentate complexes occupying two adsorption sites at the same time. This reduces adsrotpion capacity for As(V) in the course of adsorption. On the other hand, As(III) is uncharged at neutral pH as discussed above, it may not undergo any charge effect. Figs. 8 and 9 also show Langmuir and Freundlich isotherms fitted to data points for As(III) and As(V) adsorption on either of HIO-alginate bead and unimpregnated HIO. Table 3 presents parameters for Langmuir and Freundlich models. The equilibrium of As(III) and As(V) adsorption onto the beads fitted to Langmuir and Freundlich model yielded a high R2 (>0.985), suggesting that both models adequately describe the adsorption isotherms. The maximum adsorption capacities, qe,max, of As(III) and As(V) onto

the unimpregnated HIO from Langmuir model were 393.7 and 200.4 mg/g, respectively. The As(III) and As(V) adsorption capacities correspond to be 0.45 and 0.23 mole per mole of Fe, respectively. These values are comparable to previously reported As(III) and As(V) adsorption capacity of amorphous iron oxides [40]. The qe,max, of As(III) and As(V) onto the HIO-alginate beads were 47.8 and 55.1 mg/g, respectively. Cho et al. [18] reported maximum adsorption capacity of 25.19 mg/g for As(V) adsorption by nano-akaganeite encapsulated alginate beads with a greater affinity for As(V) than As(III). To the best of our knowledge, the observed arsenic adsorption capacity is among the highest for alginate-based adsorbents. Amorphous iron oxide has high sorption site density for As(III) and As(V) than other iron oxides including goethite, hematite and magnetite. High sorption site density of the immobilized hydrous iron oxide and high iron oxide loading in the beads may result for higher As(III) and As(V) adsorption. HIO-alginate beads prepared in this study effectively remove As(III) and As(V) without any pre-treatment to oxidize As(III) to As(V). Regeneration and reuse In consideration of practical applications and economics, the regeneration potential of HIO-alginate beads was investigated. Alkaline solution was selected as a desorbing solution because both of As(III) and As(V) adsorption was steeply reduced in highly alkaline pH. Regeneration and reuse of beads for arsenic adsorption were studied for 8 cycles, as presented in Fig. 10. The removal efficiency of regenerated beads varied between 51.3% and 56.6% for As(III) and between 26.7% and 29.1% for As(V). These values did not significantly deviate from those of fresh HIO-alginate beads, which were 55.7% for As(III) and 30.5% for As(V). These effective regeneration results are in agreement with other studies that have reported alkaline solution as a favorable eluent for anion desorption from polymer-based composite adsorbents due to the high affinity of impregnated metal oxides for hydroxyl ions [54]. In high alkaline pH, most of the surface sites of alginate and HIO are

Table 3 Langmuir and Freundlich isotherm constants for As(III) and As(V) adsorption onto HIO-alginate beads and unimpregnated HIOs. Adsorbent

HIO-alginate beads Unimpregnated HIOs

Target

As(III) As(V) As(III) As(V)

Freundlich qe = Kf Ce1/n

Langmuir qe = qe,max b Ce/(1 + b Ce) 2

b (L/mg)

qe,max (mg/g)

R

Kf ((mg/g).(L/mg)1/n)

n

R2

0.0152 0.0028 0.1140 0.0596

47.8 55.1 393.7 200.4

0.989 0.993 0.994 0.985

4.02 0.67 85.11 54.34

2.49 1.59 3.37 4.63

0.991 0.991 0.988 0.979

Please cite this article in press as: A. Sigdel, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.01.005

G Model

JIEC-2785; No. of Pages 10

60 (a)

As(V) removal (%)

As(III) removal (%)

A. Sigdel et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx

40

20

60

9

(b)

40

20

0

0 0

1

2

3

4

5

6

7

8

0

Regeneration cycle

1

2

3

4

5

6

7

8

Regeneration cycle

Fig. 10. Removal of (a) As(III), and (b) As(V) by HIO-alginate beads regenerated using 0.05 M NaOH solution with up to 8 regenartion cycles. Regeneration cycle 0 indicates fresh HIO-alginate beads were used for the arsenic adsorption (initial arsenic(III) or arsenic(V) concentration = 10 mg/L as As; bead dose = 1 g/L; iron loading in beads = 10 wt%; contact time = 7 d; initial pH = 6.0  0.2; temperature = 23  1 8C) (n = 2).

negatively charged, and competition between complexed ligands and hydroxyl ions causes desorption of certain ligands [55]. In other words, coulombic repulsion of the desorbed ligand from beads promoted for desorption kinetics of the ligand. The losses in bead mass after 8 regeneration cycles were measured as approximately 15% and 20% for As(III) and As(V) loaded beads, respectively. These indicate that the arsenic loaded beads can be effectively desorbed and regenerated via NaOH treatment. Diffusion of adsorbate across a support material such as alginate in this study may require longer time and reduce adsorption capacity at a certain time. However, supported adsorbents are mechanically and chemically stronger and more amenable to regeneration for multiple cycles than unimpregnated ones. The granulated or unimpregnated iron oxides have not been regenerated [56] possibly due to their disintegration with corrosive actions of eluents (acidic or alkaline solution) and wettability and dispersibility of the particles in water. Immobilization of iron oxides into support material is very helpful to solve issues related to early deactivation of the particles, poor mechanical strength and separation of used particles from any flow-through system. For comparison of arsenic adsorption data for the regeneration, experiments were conducted with 10 wt% iron-loaded beads. However, arsenic removal efficiency by the beads can be improved by increasing iron loading in beads to 20 wt% or 30 wt% as shown in Fig. 6 or by increasing adsorbent dose in a solution. Cyclic regeneration might reduce mechanical strength, which may directly influence bead durability. Further studies will be conducted to enhance long-term stability in terms of mechanical strength. Conclusions HIO-alginate beads were efficient in treating arsenic-contaminated water. As(III) removal by adsorption onto HIO-alginate beads was maximized at pH of 6–9 while adsorption of As(V) was higher in acidic solution than alkaline solution. Adsorption efficiency for both As(III) and As(V) mostly increased with increased iron loading, but As(V) adsorption efficiency slightly decreased at high loading due to the reduction in surface area of beads. Phosphate showed a pronounced interfering effect on arsenic removal by beads, especially at high phosphate concentration. The competitive effects of other anions, such as sulfate, bicarbonate, chloride, and nitrate, were insignificant. Pseudo-second-order kinetic and Weber and Morris intra-particle diffusion models well described the experimental data for As(III) and As(V), respectively. These results indicate that chemisorption and intra-particle diffusion mainly governed adsorption for As(III) and As(V), respectively. The parameters of adsorption isotherms were calculated by fitting

equilibrium data, which suggested that beads effectively adsorbed both As(III) and As(V) compared to previous studies of alginatebased adsorbents. The regeneration studies showed that beads could be regenerated by NaOH solution and reused for multiple cycles. The overall results indicate that HIO-alginate beads possess potential for use as an effective adsorbent for arsenic removal from water. Acknowledgements This work was supported by the Korea Ministry of Environment as ‘‘The Eco-Innovation Project’’ (Global Top project) (No. GT-SWS11-02-007-4 and No. GT-SWS-11-02-007-6). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.jiec.2016.01.005. References [1] T.M. Clancy, K.F. Hayes, L. Raskin, Environ. Sci. Technol. 47 (2013) 10799. [2] WHO, Guidelines for Drinking Water Quality, fourth ed., World Health Organization, Geneva, 2011p. 315. [3] S. Shankar, U. Shanker, Sci. World J. 304524 (2014) 1. [4] UNICEF, Arsenic Contamination in Groundwater: Current Issues 2, The United Nations Children’s Fund, New York, NY, 2013. [5] C.K. Jain, R.D. Singh, J. Environ. Manage. 107 (2012) 1. [6] D. Mohan, C.U. Pittman, J. Hazard. Mater. 142 (2007) 1. [7] B. Petrusevski, S. Sharma, J.C. Schippers, K. Shordt, Arsenic in drinking water, IRC International Water and Sanitation Centre, Delft, 2007. [8] D.E. Giles, M. Mohapatra, T.B. Issa, S. Anand, P. Singh, J. Environ. Manage. 92 (2011) 3011. [9] K.B. Payne, T.M. Abdel-Fattah, J. Environ. Sci. Health 40 (2005) 723. [10] J.T. Mayo, C. Yavuz, S. Yean, L. Cong, H. Shipley, W. Yu, J. Falkner, A. Kan, M. Tomson, V.L. Colvin, Sci. Technol. Adv. Mater. 8 (2007) 71. [11] P. Trivedi, L. Axe, J. Colloid Interface Sci. 244 (2001) 221. [12] S.G. Chung, J.C. Ryu, M.K. Song, B. An, S.B. Kim, S.H. Lee, J.W. Choi, J. Hazard. Mater. 267 (2014) 161. [13] S. Dixit, J.G. Hering, Environ. Sci. Technol. 37 (2003) 4182. [14] B. Chen, Z. Zhu, J. Ma, M. Yang, J. Hong, X. Hu, Y. Qiu, J. Chen, J. Colloid Interface Sci. 434 (2014) 9. [15] E. Deliyanni, T.J. Bandosz, K.A. Matis, J. Chem. Technol. Biotechnol. 88 (2013) 1058. [16] V.K. Gupta, V.K. Saini, N. Jain, J. Colloid Interface Sci. 288 (2005) 55. [17] S.K. Maji, Y.H. Kao, C.W. Liu, Desalination 280 (2011) 72. [18] K. Cho, B.Y. Shin, H.K. Park, B.G. Cha, J. Kim, RSC Adv. 4 (2014) 21777. [19] X. Lv, G. Jiang, X. Xue, D. Wu, T. Sheng, C. Sun, X. Xu, J. Hazard. Mater. 262 (2013) 748. [20] D.W. Cho, B.H. Jeon, C.M. Chon, Y. Kim, F.W. Schwartz, E.S. Lee, H. Song, Chem. Eng. J. 200 (2012) 654. [21] B. Pan, H. Qiu, B. Pan, G. Nie, L. Xiao, L. Lv, W. Zhang, Q. Zhang, S. Zheng, Water Res. 44 (2010) 815. [22] D. Ocin´ski, I. Jacukowicz-Sobala, J. Raczyk, E. Kociołek-Balawejder, React. Funct. Polym. 83 (2014) 24. [23] J.A. Wilkie, J.G. Hering, Colloids Surf., A: Physicochem. Eng. Aspects 107 (1996) 97. [24] Y.S. Ho, J.C.Y. Ng, G. McKay, Sep. Purif. Methods 29 (2000) 189. [25] Y.S. Ho, G. McKay, Process Biochem. 34 (1999) 451. [26] W.J. Weber, J.C. Morris, J. Sanitary Eng. Div. 89 (1963) 31.

Please cite this article in press as: A. Sigdel, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.01.005

G Model

JIEC-2785; No. of Pages 10 10 [27] [28] [29] [30] [31] [32]

[33] [34] [35] [36] [37]

[38] [39]

A. Sigdel et al. / Journal of Industrial and Engineering Chemistry xxx (2016) xxx–xxx I. Langmuir, J. Am. Chem. Soc. 40 (1918) 1361. H.M.F. Freundlich, Z. Phys. Chem. 57 (1906) 385. B. Tansel, P. Nagarajan, Adv. Environ. Res. 8 (2004) 411. J. Yang, H. Zhang, M. Yu, I. Emmanuelawati, J. Zou, Z. Yuan, C. Yu, Adv. Funct. Mater. 24 (2014) 1354. H.Y. Lin, Y.W. Chen, W.J. Wang, J. Nanopart. Res. 7 (2005) 249. S. Lowell, J.E. Shields, M.A. Thomas, M. Thommes, Characterization of Porous Solids and Powders: Surface Area, Pore Size and Density, Kluwer Academic Publishers, Dordrecht, 2004p. 58. C. Peng, Q. Zhao, C. Gao, Colloids Surf., A: Physicochem. Eng. Aspects 353 (2010) 132. P.J. Swedlund, H. Holtkamp, Y. Song, C.J. Daughney, Environ. Sci. Technol. 48 (2014) 2759. D.A. Dzombak, F.M. Morel, J. Colloid Interface Sci. 112 (1986) 588. C. Escudero, N. Fiol, I. Villaescusa, J.C. Bollinger, J. Hazard. Mater. 164 (2009) 533. W. Stumm, Chemistry of the Solid-Water Interface: Processes at the MineralWater and Particle-Water Interface in Natural Systems, Wiley-Interscience, New York, NY, 1992p. 43. S.O. Kim, W. Chun Lee, H. Goo Cho, B.T. Lee, P.K. Lee, S. Hee Choi, Environ. Technol. 35 (2014) 251. G. Zhang, Z. Ren, X. Zhang, J. Chen, Water Res. 47 (2013) 4022.

[40] K.P. Raven, A. Jain, R.H. Loeppert, Environ. Sci. Technol. 32 (1998) 344. [41] M.E. Essington, Soil and Water Chemistry: An Integrative Approach, second ed., CRC Press, Boca Raton, FL, 2015. [42] L. Cumbal, A.K. SenGupta, Environ. Sci. Technol. 39 (2005) 6508. [43] P.X. Sheng, K.H. Wee, Y.P. Ting, J.P. Chen, Chem. Eng. J. 136 (2008) 156. [44] D.A. Sverjensky, K. Fukushi, Geochim. Cosmochim. Acta 70 (2006) 3778. [45] S. Lin, D. Lu, Z. Liu, Chem. Eng. J. 211 (2012) 46. [46] H. Zeng, M. Arashiro, D.E. Giammar, Water Res. 42 (2008) 4629. [47] H. Zhu, Y. Jia, X. Wu, H. Wang, J. Hazard. Mater. 172 (2009) 1591. [48] R. Xu, J. Jiang, C. Cheng, in: J. Xu, P.M. Huang (Eds.), Effect of Ionic Strength on Specific Adsorption of Ions by Variable Charge Soils: Experimental Testification on the Adsorption Model of Bowden et al., Springer, Berlin, Heidelberg, 2010, p. 78. [49] S. Goldberg, Soil Sci. Soc. Am. J. 66 (2002) 413. [50] M.B. McBride, Clays Clay Miner. 45 (1997) 598. [51] R. Xu, Y. Wang, D. Tiwari, H. Wang, J. Environ. Sci. (China) 21 (2009) 927. [52] J. Wang, W. Xu, L. Chen, X. Huang, J. Liu, Chem. Eng. J. 251 (2014) 25. [53] J. Lu, D. Liu, J. Hao, G. Zhang, B. Lu, Chem. Eng. Res. Des. 93 (2015) 652. [54] K.J. McDonald, R. Brandon, K.J. Reddy, Sci. Rep. 5 (2015) 1. [55] S. Sarkar, J.E. Greenleaf, A. Gupta, D. Ghosh, L.M. Blaney, P. Bandyopadhyay, R.K. Biswas, A.K. Dutta, A.K. SenGupta, Water Res. 44 (2010) 5813. [56] W.H. Ho¨ll, Environ. Geochem. Health 32 (2010) 287.

Please cite this article in press as: A. Sigdel, et al., J. Ind. Eng. Chem. (2016), http://dx.doi.org/10.1016/j.jiec.2016.01.005