magnetite composite for adsorption of Cu(II) and As(V)

Chemical Engineering Journal 200–202 (2012) 654–662

Contents lists available at SciVerse ScienceDirect

Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej

A novel chitosan/clay/magnetite composite for adsorption of Cu(II) and As(V) Dong-Wan Cho a,b, Byong-Hun Jeon a, Chul-Min Chon b, Yongje Kim b, Franklin W. Schwartz c, Eung-Seok Lee d, Hocheol Song e,⇑ a

Department of Environmental Engineering, Yonsei University, Wonju, Gangwon-do 220-710, South Korea Korea Institute of Geoscience and Mineral Resources, Daejeon 305-350, South Korea c School of Earth Sciences, Ohio State University, Columbus, OH 43210, USA d Department of Geological Science, Ohio University, Athenes, OH 45701, USA e Department of Environment and Energy, Sejong University, Seoul 143-747, South Korea b

h i g h l i g h t s " A composite composed of chitosan, nano-magnetite, and heulandite was prepared. " Properties of the composite were characterized using various instrumentation. " The composite was applied as an adsorbent for removal of Cu(II) and As(V). " The optimal mass ratio of chitosan, nano-magnetite, and heulandite was 1:1:2. " We demonstrated potential utility of the material in environmental cleanup.

a r t i c l e

i n f o

Article history: Received 30 March 2012 Received in revised form 27 June 2012 Accepted 27 June 2012 Available online 5 July 2012 Keywords: Copper Arsenic Chitosan Clay Magnetite Tripolyphosphate

a b s t r a c t A composite adsorbent was prepared by entrapping cross-linked chitosan and nano-magnetite (NMT) on heulandite (HE) surface to remove Cu(II) and As(V) in aqueous solution. The optimized mass ratio among chitosan, HE, and NMT was determined to be 1:1:2. Kinetics studies indicated the removal of both ions followed pseudo-second order kinetics, suggesting specific interaction with surface functional groups was the major route of the removal process. The composite gave the maximum equilibrium uptakes of Cu(II) and As(V) of 17.2 and 5.9 mg g1 in initial concentration ranges of 16–656 and 17–336 mg L1, respectively. The thermodynamic data showed that both adsorption processes were thermodynamically favorable, spontaneous, and endothermic nature. The adsorption capacity for Cu(II) increased continuously with an increase in initial solution pH (3–9), but adsorption of As(V) showed an opposite trend. The overall results demonstrated the potential utility of the composite for Cu(II) and As(V) removal from aqueous solutions. Ó 2012 Elsevier B.V. All rights reserved.

1. Introduction Copper is one of the essential elements to human body in trace quantities, but elevated levels of copper may cause various health problems in liver, kidney, and the central nervous system [1]. Also, arsenic is a priority pollutant posing various health threats associated with lung, skin, bladder, liver, and other cancers, and high levels of arsenic in drinking water have affected millions of people across the world [2]. Arsenic in the environment is mostly of natural origin, occurring in minerals in conjunction with other metals and sulfide. The predominant arsenic form in oxidizing conditions is arsenate (As(V)), while it readily converts to arsenite (As(III)) in reducing conditions. High levels of As(V) are frequently found in ⇑ Corresponding author. Tel.: +82 2 3408 3232; fax: +82 2 3408 4320. E-mail address: [email protected] (H. Song). 1385-8947/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.cej.2012.06.126

geological regions rich in arseno-metal sulfide minerals [3]. Industrially, arsenic has been used a main ingredient of copper alloy metals for providing material strength, and therefore, both metals are likely to be found in a single water body if contamination originates from such sources. Surface- and groundwaters contaminated with inorganic pollutants can be treated by various methods, including ion exchange, precipitation, adsorption, and membrane processes [4–8]. Adsorption is a relatively effective and economical method compared to other methods, which occasionally suffer from high operation cost and low treatment efficiency. During the past decade, there has been a great deal of growing interests and research advancement on the use of biomaterials for environmental clean-up. For example, chitosan-based materials have drawn significant attention for their favorable properties and versatility as a treatment medium for various contaminants [9,10].

D.-W. Cho et al. / Chemical Engineering Journal 200–202 (2012) 654–662

Chitosan is an alkaline deacetylated product of chitin, characterized by high hydrophilicity and large number of surface hydroxyl and amino groups, and is considered as an eco-friendly material with minial toxicity, biocompatibility, biodegradability, and great availability in nature [9,11]. Chitosan is soluble in acidic media and solidifies in alkaline media, with the latter offering compatibility with other supporting materials such as clay minerals, sand, activated carbon, and PVC bead [1,9,12,13]. The key technical issue in application of a chitosan-based adsorbent in aqueous media is prevention of chitosan dissolution in acidic pH conditions. Cross-linking agents such as epichlorohydrin, tripolyphosphate (TPP), and glutaraldehyde have been frequently employed for such a purpose of enhancing the physical stability of chitosan [14–16]. In addition, recent researches on chitosan application have focused on imparting magnetic property to chitosan-based adsorbents for facile recovery after treatment. Fan et al. [17,18] developed magnetic chitosan adsorbents by coating thioureachitosan with magnetic fluids and further enhanced adsorbent capacity with surface imprinting method. Klepka et al. [19] prepared Fe-crosslinked chitosan complex by co-precipitation technique. Yuwei and Jianlong [20] developed super-paramagnetic chitosan nanoparticles from mixture of chitosan and Fe(II)–Fe(III) ions under anaerobic condition. Zhang et al. [21] synthesized activated carbon fiber coated with nano-sized magnetite using chitosan as a film-forming material. Magnetic chitosan composite materials were alternatively prepared by micro-emulsion process using kaolin [16] and fly-ash [22] as a base material. This study aimed to preparation of a novel magnetic composite material, composed of chitosan, nano-magnetite (NMT), and clay (heulandite) using TPP as a chitosan stabilizer, for removal of As(V) and Cu(II) in aqueous solution. Heulandite (HE) is a zeolitetype mineral possessing sheet-like structure, and therefore it is expected to provide good anchoring sites for NMT and chitosan as well as serving as a direct adsorbent for metal ions [23]. In other words, clay would offer a large surface area for incorporating NMT and chitosan while preventing particle agglomeration for better contaminant adsorption, and also provide additional adsorption sites. The characteristics of the composite material were examined with various instrumental analyses, and its performance for adsorption of Cu(II) and As(V) was evaluated under various experimental conditions to demonstrate adsorption kinetics, isotherms, thermodynamics, and effect of pH.

655

by the interaction between phosphate groups of TPP and amino group of chitosan. The composites were kept in the solution for 12 h, washed several times with distilled water to remove the excess of TPP, and oven-dried for 24 h at 50 °C. After drying and grinding, the composites were milled through a 100-mesh screen. The adsorbent synthesized was referred to as chitosan/clay/Fe3O4 (CCM). A schematic plot of CCM preparation is shown in Fig. S1 in Supplementary Materials (SM). X-ray diffractometer (D8 Advance, Bruker-AXS) analyses were performed for CCM, HE, NMT, and chitosan, using Cu Ka radiation and a LynxEye position sensitive detector. The diffraction pattern was collect from 5° to 65° 2h, with a nominal step size of 0.01° and a time per step of 1 s, using a 0.3° divergence slit and 2.5° secondary Soller slit. The resulting peaks and intensities of each mineral were compared with the powder diffraction files published by the International Centre for Diffraciton Data. Field-emission scanning electron microscope coupled with energy dispersive spectroscopy (FE-SEM/EDS), zeta potential at solution pH 5, and BET surface area of HE and CCM were measured using a Sirion FE-SEM analyzer (Netherlands), Malvern Zetasizer nano-zs (UK), and Micromeritics ASAP 2020 (USA), respectively. Magnetic properties of HE and CCM were also analyzed using a Bartington magnetic susceptibility meter (UK) with 36 mm internal diameter calibration vol. 10 cc. 2.3. Adsorption experiments Cu(II) and As(V) adsorption experiments were carried out 25 mL high density polyethylene vials (Fisher Scientific, USA). Standard solutions of desired Cu(II) and As(V) concentrations were obtained by diluting 1000 mg L1 stock solution, followed by pH adjustment at pH 5 with 0.1 N HCl or NaOH. Adsorption kinetics experiments were performed by equilibrating 0.1 g adsorbent with 20 mL adsorbate solutions of 45.1 mg L1 Cu(II) and 39.5 mg L1 As(V) in the vials. Each mixture was shaken at 23 ± 2 °C and 150 rpm for 600 min. At specified intervals, the samples were collected and filtered with 0.45 lm filter (Whatman, USA) to determine the concentration of Cu(II) and As(V). Adsorption isotherm experiments were carried out by varying the concentrations of Cu(II) and As(V) solution from 16 and 17 to 656 and 336 mg L1, respectively, under the same conditions as the adsorption kinetics experiments. The experiments were performed in duplicate and the amount of Cu(II) and As(V) ions adsorbed per unit mass of CCM (qe) was calculated as follows:

2. Experimental 2.1. Materials

qe ðmg g1 Þ ¼

ðC o  C e ÞV W

ð1Þ

Chitosan with 75–85% degree of deacetylation and 190,000– 310,000 g mol1 viscosity molecular weight was purchased from Sigma–Aldrich (USA). HE was obtained from Donghae Chemical Co., South Korea, and pulverized and sieved through a 100-mesh screen prior to use. NMT (<50 nm), sodium tripolyphosphate (STPP), copper chloride, sodium arsenate, acetic acid (99%), sodium hydroxide, hydrochloric acid were purchased from Sigma–Aldrich and were of ACS reagent grade.

where Co and Ce are initial and equilibrium concentrations of adsorbate (mg L1), respectively, W is the dry mass of the adsorbent (g), and V is volume of the solution (L). The concentration of Cu(II) and As(V) in aqueous solution was determined by ICP-OES (Ultima 2C, Horiba-Yuvon, France). The pH of solution was measured using a pH meter (Horiba, Ltd. Kyoto, Japan).

2.2. Preparation of chitosan/heulandite/Fe3O4 composites

Adsorption experiments were conducted at different constant temperatures in the range of 25–45 °C under the same condition as adsorption isotherm experiments to investigate the effect of temperature on adsorption of Cu(II) and As(V) onto CCM. The effect of pH on Cu(II) and As(V) adsorption by CCM was also investigated in the initial pH ranges between 3 and 9. The reactors received 20 mL of 68.7 mg L1 Cu(II) or 66.9 mg L1 As(V), and solution pH was adjusted with 0.1 N NaOH and 0.1 N HCl. Adsorbent (0.1 g) was added in each reactor and reacted for 24 h at 25 °C.

Chitosan (2 g) was dissolved in 100 mL acetic acid (2%) to prepare chitosan solution. STPP solution was prepared by dissolving 13.3 g of the solid in 1 L of distilled water and adjusted to pH 4 with 1 N HCl. Known amount of NMT and HE were added into the chitosan solution under ultrasonic stirring for 30 min. Then the mixture was added drop wise using 10 mL syringe into 100 mL STPP solution. The cross-linked adsorbents were formed

2.4. Effect of temperature and pH

656

D.-W. Cho et al. / Chemical Engineering Journal 200–202 (2012) 654–662

imparted significant magnetism to HE. The measured magnetic susceptibility of 3.0  104 m3 kg1 of CCM is comparable to the values of magnetite (4.0  104–1.0  103 m3 kg1), suggesting CCM can be readily removed from water by applying magnetic force to the medium [26]. According to XRD analysis, the spectrum of CCM composites not only showed a typical diffraction pattern of NMT but included the peaks of a supporting material comprised mainly of heulandite with traces of plagiclase and quartz (Fig. 2a–c). Heulandite–clinoptilolite series (HEU-type zeolites) are the most common natural zeolites on the earth, occurring mainly in sedimentary and volcanic rocks and constituting extensive deposits all over the world [27]. Heulandite-group minerals consisting of heulandite–clinoptilolite are in diagenetically altered pyroclastic rocks, particularly in rhyoliterhyodacite tuffs [28]. Heulandites and clinoptilolites, ideally (Ca, Na, K)6Al6Si30O7224H2O, are isostructural and distinguished on the basis of their Si/Al ratios and the stabilities of their structures; heulandites have Si/Al < 4 and clinoptilolites have Si/Al > 4. Heulandite-Ca and clinoptilolite-K are taken as the type species of the series, and Ca and K are the most abundant single extra-framework cation in the range 0.71–0.80 and 0.80–0.84, respectively [29]. Because of their ion-exchange, adsorption, and molecular sieve properties, as well as their geographically widespread abundance, zeolite minerals have generated worldwide interest for use in a broad range of applications. Due to the favorable ion-exchange selectivity of natural zeolites for certain cations, such as Cs+, Sr2+ and NHþ 4 , these minerals have been studied for potential use in the treatment of nuclear wastewaters, municipal and industrial wastewaters and acid mine drainage waters [30,31].

3. Results and discussion 3.1. Characterization of adsorbent The characteristic properties of HE and CCM are presented in Table 1. The values of zeta potential at pH 5 for HE and CCM were 0.173 and 0.425 mV, respectively, indicating the clay surface became more negatively charged during modification process. It is presumed the negative charge development mainly resulted from incorporation of anionic TPP molecules into the surface, but the presence of NMT on CCM surface may have exerted a negating effect on the negative charge development because magnetite would exhibit positive surface charge under the same pH condition [24]. BET results shows modification resulted in a decrease of surface area from 22.1 to 5.1 m2 g1. It has been reported cross-linked chitosan tends to agglomerate on clay surface by interaction with hydroxylated edges of clay [25], and reduces surface area of clay by blocking the pores. In addition, stacking of magnetite particles on HE surface might have rendered a similar pore blocking effect. HE had a negligible magnetic susceptibility, but modification

Table 1 Characteristic parameters of HE and CCM.

PA CCM

Zeta potential (mV) at pH 5

BET surface area (m2 g1)

Magnetic susceptibility (m3 kg1)

0.173 0.425

22.1 5.1

1.2  107 3  104

M 7000

H,Pl 5000

M

a) H

H

M

Q

M

M

M

M

M

3000

Intensity (cps)

7000

M

b)

5000

M H Q H

Q H

2000

M

M

H

3000

M

M

M

HH

Pl H,Pl

H

H H Pl H

c)

12000 0

(040)

8000

(020)

d) 4000 0 5

10

15

20

25

30

35

40

45

50

55

60

65

2 theta (degree) Fig. 1. X-ray diffraction patterns for the reacted materials, (a) CCM composites after the cross-linking reaction using tripolyphosphate, (b) NMT, (c) supporting materal and (d) virgin chitosan before the cross-linking reaction. Abbreviations: M, magnetite; H, heulandite; Pl, plagioclase; Q, quartz.

657

D.-W. Cho et al. / Chemical Engineering Journal 200–202 (2012) 654–662

1007

after As adsorption after Cu adsorption

0.25

CCM Heulandite 1632

0.20

1028 1535

Absorbance

1057

0.15

1379

0.10 1150

1800

1600

1400 901

0.05

1632 1535 781 721

1379

2916 2848

0.00 3900 3600 3300 3000 2700 2400 2100 1800 1500 1200

900

600

-1

Wavenumbers (cm ) Fig. 2. FTIR spectra of Heulandite, CCM and CCM suspensions after the As(V) and Cu(II) adsorption.

The XRD spectrum of chitosan (Fig. 1d) has low crystallinity and the characteristic peaks at 2h = 11.2° and 20.0° are assigned to semi-crystalline structure at the (0 2 0) and (0 4 0) plane, respectively [32]. For the composites material with chitosan cross-linked by tripolyphosphate, the diffraction peaks which are typical fingerprints of semi-crystalline chitosan were disappeared, indicating that the degree of chitosan crystallinity was destroyed due to interruption of hydrogen through the cross-linking reaction. Additional characterization of heulandite and CCM (before and after As(V) and Cu(II) adsorption) was conducted with Fourier transform infrared spectroscopy (FTIR), and the spectral result is given in Fig. 2. The broad band at 3700–3000 cm1 observed for all the spectra is due to the adsorption of water molecules as result of an OAH stretching mode of hydroxyl groups of the samples. The spectrum of CCM and suspensions after As(V) adsorption clearly shows the asymmetric and symmetric ACH2 stretching modes of the chitosan at 2916 and 2848 cm1, respectively, in a complex spectral region where several bands have been observed due to symmetric and asymmetric stretching of ACH from the ring and from ACH2OH and ACH3 groups. The FTIR spectrum of chitosan is characterized by several amine and amide bands in the 1800–1200 cm1 range of wavenumbers. The band at 1632 cm1 of CCM spectrum (i.e. amide I band) is assigned to the absorption C@O group of chitosan. The amide II band due to NAH bending appears at 1535 cm1 and it is overlapped by amine bands [33]. In the spectrum of CCM, the amide and amine band appear at 1632 cm1 and 1535 cm1, respectively, as a clear absorption, while the decrease in intensity of the band at the spectral region was observed and the amide II band was absent in the spectrum of heulandite. The absorption bands at 1150 cm1 (anti-symmetric stretching of the CAOAC bridge), 1057 and 1028 cm1 (skeletal vibrations involving the CAOAC stretching) were also observed in the CCM spectrum.

In the heulandite spectrum, the strongest TAO stretching vibration appeared at 1007 cm1, which could be attributable to SiAOASi and SiAOAAl vibrations [34]. In addition, the isolated OH stretching (3620 cm1) band attributable to interaction of the water hydroxyl with the cations, to the hydrogen bonding of the water molecule to surface oxygen (3440 cm1), and to the bending mode of the water (1640 cm1) was observed. The only substantial absorption band of the magnetite is centered around 570(+5) cm1 [35]. The surface morphologies of HE and CCM analyzed by FE-SEM showed that the surface of HE was relatively clean, exhibiting sheet-like structure typical of zeolite minerals (Fig. 3a). On the other hand, CCM showed quite different morphology with clusters of layered nano-sized particles covering surface, while the presence of chitosan was not visually apparent. The EDS spectrum of HE revealed Si (32 wt%), O (31.7 wt%), Al (13.3 wt%), and C (11.8 wt%) as major components, with Ca (4.8 wt%), Na (4.8 wt%), Fe (0.6 wt%), K (0.6 wt%), and Mg (0.4 wt%) representing minor and trace components (Fig. 3a). The EDS spectrum of CCM showed various changes in elemental composition by the modification. The major changes in elemental composition include appearance of P, decrease of Al and Si, and increase of Fe, which respectively reflect presence of phosphate groups of TPP, surface coverage of clay by magnetite-chitosan complex, and incorporation of magnetite on clay surface. 3.2. Synthesis optimization In order to find optimal mass ratio of chitosan, NMT, and HE for better removal of Cu(II) and As(V), the adsorbents were synthesized under varying mass ratio of HE and chitosan at constant NMT mass, and of HE and NMT at constant chitosan mass. Adsorption experiments were carried out with 5 g L1 adsorbent dose and

658

D.-W. Cho et al. / Chemical Engineering Journal 200–202 (2012) 654–662

Fig. 3. FE-SEM images & EDX analysis of (a) HE and (b) CCM.

118.2 mg L1 Cu(II) or 10.6 mg L1 As(V) at pH 5. At constant NMT mass (HE:NMT = 1:0.33), the adsorbed amount of Cu(II) after 24 h reaction increased from 8.1 to 13 mg g1 with increasing ratio of HE:chitosan from 1:0 to 1:4, while there was little influence of chitosan content on the adsorption of As(V), which leveled off at 0.03 mg g1 (Fig. 4a). It has been reported chitosan is capable of removing As(V) via electrostatic attraction between amine functional groups of chitosan and As(V), which becomes more favorable under low pH condition due to protonation of amine groups [36,37]. The lack of As(V) adsorption on CCM in this study could be ascribable to preferential adsorption of TPP on the protonated amine functional groups, leaving little sites available for As(V) adsorption. The TPP mediated cross-linking of chitosan occurs through the ionic interaction between the negatively charged phosphate groups of TPP and protonated amine groups on the chitosan molecule [15]. As(V) removal could also occur via adsorption onto NMT considering great affinity of As(V) to iron oxide

minerals [38,39], but this appears to be insignificant, presumably due to relatively low mass of NMT and competition effect rendered by TPP. The increase Cu(II) adsorption with increasing chitosan content suggests chitosan imparted a favorable property to HE for Cu(II) adsorption (Fig. 4a). Cu(II) removal by unmodified HE gave removal efficiency of 37% and it has been demonstrated Cu(II) binds to HE by formation of surface complex or entering in the intra-crystalline cavities and channels of molecular dimensions in the structure of HE [40]. The major effect of chitosan on enhanced Cu(II) adsorption is to provide additional binding sites as Cu(II) readily binds to amine groups that were not involved in cross-linking by TPP, as well as to hydroxyl groups in the chitosan to form stable complexes [41–43]. In addition, phosphate groups in TPP present on the surface might play a beneficial role in Cu(II) removal because they could form metal-ion complex with transition metal ions [42]. Given the Cu(II) removal efficiencies with varying HE to

659

D.-W. Cho et al. / Chemical Engineering Journal 200–202 (2012) 654–662

75

10

50

5

25

0

0 HE

1:0.13

1:0.4

1:1

1:2

1:4

(a)

qe (mg g-1)

15

100

Adsorbed amount_Cu(II) Adsorbed amount_As(V) RE (%)_Cu(II) RE (%)_As(V)

75

10

50

5

25

Removal Efficiency (%)

20

0

0 1:0.33

1:1

1:2

1:4

NMT

(b) Fig. 4. Removal efficiencies (%) and adsorbed amount (mg g1) of Cu(II) and As(V) by composites with (a) varying ratio of chitosan to HE (HE: NMT = 1:0.33) at initial concentrations of 109.4 Cu(II) mg L1 and 10.3 As(V) mg L1, and (b) varying ratio of NMT to HE (HE: chitosan = 1:1) at initial concentrations of 118.2 Cu(II) mg L1 and 10.6 As(V) mg L1 (Adsorbent dose = 5 g L1, contact time = 24 h, initial pH = 5).

chitosan ratio and the As(V) removal efficiencies that had little variation with respect to the ratio, the mass ratio of 1:1 was chosen for further optimization of NMT of the composite. The results of As(V) and Cu(II) adsorption by CCM synthesized under varying mass ratios of HE to NMT at constant chitosan mass (HE:chitosan = 1:1) are shown in Fig. 4b. Increasing the relative mass ratio of NMT from 0.33 to 4 resulted in increase of As(V) adsorption from 0.1 to 0.6 mg g1, which corresponds to 4.6% to 25.8% As(V) removal in the solution. The increase of As(V) adsorption may be attributed to the increased mass of NMT entrapped on the surface to provide adsorption sites for As(V). Adsorption experiment with 5 g L1 NMT gave removal efficiencies of 97% and 7.8% for As(V) and Cu(II), respectively, indicating NMT has good adsorption capability for As(V), but not for Cu(II). Relatively low As(V) removal by CCM as compared to NMT alone simply reflects lower mass of NMT contained in CCM, but considering complex association of NMT with other moieties on CCM surface, it is likely that As(V) adsorption was inhibited to some extent. For example, phosphate groups in TPP can limit the access of As(V) to NMT by engaging electrostatic repulsion or directly competing for adsorption sites on NMT. Also, chitosan molecules associated with NMT can serve as a physical barrier to NMT surface. In contrast to As(V) adsorption, increasing mass ratio of NMT resulted in a gradual decrease of Cu(II) adsorption to give

8.5 mg g1 at the highest NMT ratio, which corresponded to 35.8% Cu(II) removal efficiency. Such a decrease of Cu(II) adsorption appeared to be related to the decrease of chitosan content in unit mass of CCM since it is assumed chitosan is mainly responsible for Cu(II) removal. In selection of optimal mass ratio of CCM composite, there was little difference in As(V) removal capacity of the CCM synthesized under NMT mass ratio 2 and 4, while there was 8.6% more removal by the former, therefore the mass ratio of 1:1:2 for HE:chitosan:NMT was determined for further adsorption characterization study. 3.3. Adsorption kinetics and isotherm Fig. 5 illustrates the adsorption kinetics for initial Cu(II) and As(V) concentrations of 45.1 and 39.5 mg L1, respectively, by 5 g L1 of 1:1:2 CCM composite. More than 86% of the Cu(II) was removed by CCM within the first 20 min and the removal efficiency of As(V) increased gradually to 27% within 120 min. The adsorption capacity for Cu(II) and As(V) was approximated to 8.4 and 2.2 mg g1 at sorption equilibrium, respectively. The change of pH during the kinetics experiments is shown in Fig. S2 in SM. The solution pH remained relatively constant for both Cu(II) and As(V) solutions until the end of reaction despite the little fluctuation in very early reaction times. The different adsorption kinetics of Cu(II) and As(V) suggests adsorption of those ions involves different adsorption sites on the adsorbent. As noted above, adsorption of Cu(II) occurs both on the surface/pores of HE or on the amine and hydroxyl groups of cross-linked chitosan. These adsorption sites were readily accessible from outer interface, and thus resulted in rapid adsorption process [43]. On the other hand, As(V) adsorption primarily occurs on NMT particles entrapped on the surface and likely to involve diffusion of As(V) to reach the adsorption sites. To quantitatively assess such difference of adsorption kinetics of Cu(II) and As(V), the experimental data were analyzed according to the linearized form of pseudo-first-order and the pseudo-second-order kinetics models. A linear form of pseudo-first-order kinetics can be expressed as:

logðqe  qt Þ ¼ log qe 

k1 t 2:303

ð2Þ

where qt (mg g1) is the amount adsorbed at time t (min), and k1 is the rate constant of pseudo first-order adsorption (min1). The values of log (qe  qt) were calculated from the kinetic data (Fig. 6).

15

Cu(II) As(V)

qe (mg g-1)

15 qe (mg g-1)

100

Adsorbed amount_Cu(II) Adsorbed amount_As(V) RE (%)_Cu(II) RE (%)_As(V)

Removal Efficiency (%)

20

10

5

0 0

200

400

600

Time (min) Fig. 5. Adsorption kinetics of Cu(II) and As(V) by CCM. (Initial conc. of Cu(II) and As(V): 45.1 and 39.5 mg L1; adsorbent dose: 5 g L1; initial pH = 5).

660

D.-W. Cho et al. / Chemical Engineering Journal 200–202 (2012) 654–662

1 1 1 ¼ þ qe bC e qm qm

20

Cu(II) As(V)

where Ce is the equilibrium concentration (mg L1) and the Langmuir constants qm (mg g1) represent the maximum monolayer adsorption capacity and b (L mg1) is the Langmuir constant related to energy of adsorption. The values of monolayer capacity (qm), Langmuir constant (b) were obtained from the intercept and slope of these plots, respectively. In addition, a dimensionless constant, separation factor or equilibrium parameter, RL, was calculated to characterize the isotherms [49]:

qe (mg g-1)

15

10

5

RL ¼ 0 0

ð4Þ

200

400

600

Ce (mg L-1) Fig. 6. Adsorption isotherm of Cu(II) and As(V) by CCM. (Conc. range of Cu(II) and As(V): 16–656 and 17–336 mg L1; adsorbent dose: 5 g L1, contact time = 24 h).

The pseudo-second-order model based on the adsorption equilibrium capacity can be expressed as:

t 1 1 ¼ þ qt k2 q2e qe

ð3Þ

where k2 (g mg1 min1) is the rate constant of pseudo-second order adsorption. The kinetics parameters obtained by fitting the kinetics data of Cu(II) and As(V) adsorption are given in Table 2. Better correlation coefficients (R2) were obtained with the pseudo-second-order kinetic model (0.9999, 0.9910) compared to the pseudo-first-order kinetic model (0.8023, 0.9841). Also, the calculated equilibrium adsorption capacity qe(cal) for the pseudo-second-order kinetic model (8.41, 2.29 mg g1) was well approximated to the experimental value qe(exp) (8.39, 2.20 mg g1), indicating better suitability of pseudo-second-order kinetic model to describe adsorption kinetics of Cu(II) and As(V) onto the adsorbents. The chemisorptions are usually regarded as the rate-determining step for adsorption processes following pseudo-second-order kinetics [11,44,48], and this is in agreement with the suggested modes of Cu(II) and As(V) adsorption that involved specific interaction with surface functional groups. The distribution of the adsorbate on adsorbent surface at equilibrium can be fundamentally explained by adsorption isotherm [15]. Langmuir and Freundlich isotherm models were used to determine adsorption capacities of the adsorbent for Cu(II) and As(V) (Fig. 6). The Langmuir model assumes adsorption occurs at specific homogeneous sites with equal adsorption energy and expressed as follows:

1 1 þ bC n

ð5Þ

where C0 (mg L1) is the initial Cu(II) and As(V) concentrations. The RL values indicate whether the adsorption is unfavorable (RL > 1), linear (RL = 1), favorable (0 < RL < 1), or irreversible (RL = 0). For all the tested Cu(II) and As(V) concentrations, RL values are in the range of 0 < RL < 1, which indicates that the adsorption of Cu(II) and As(V) on CCM is a favorable process. The Freundlich model has also been widely applied and can be written as:

qe ¼ kf C e1=n

ð6Þ

where kf is a constant taken as an indicator of adsorption capacity (L g1) and n is a constant indicative of the adsorption intensity. The Langmuir and Freundlich constants for the adsorption of Cu(II) and As(V) on CCM are given in Table 3. Also, for a comparison purpose, adsorption capacities of various chitosan-based adsorbents reported in the literature are presented in Table 4. The Freundlich constants (n) for Cu(II) and As(V) are within the range of 1 and 10, suggesting the adsorption of Cu(II) and As(V) on the adsorbents is favorable [50]. The correlation coefficient (R2) of Freundlich model for Cu(II) (0.9781) is higher than that of Langmuir (0.7798), indicating that CCM has high degrees of heterogeneous distribution of active sites on the surface, presumably due to the uneven surface Table 4 Adsorption capacities for the adsorption of Cu(II) and As(V) onto various chitosan adsorbents. Adsorbent

Adsorption capacity (mg g1)

Chitosan/Montmorillonite Crab shell chitosan TiO2-impregnated chitosan bead Protonated chitosan beads Chitosan immobilized on bentonite Magnetic chitosan nanoparticles Chitosan/Clay/Magnetite

Cu(II)

As(V)

– – – 52.0 14.9 35.5 14.3

9.0 0.15 4.9 – – – 6.5

Reference

[36] [45] [46] [47] [25] [20] This work

Table 2 Pseudo-first-order and pseudo-second-order calculated qe(cal) and experimental qe(exp) values for CCM. qe (exp)mg g1

Pseudo first-order kinetic model k1 (l min

Cu(II) As(V)

8.39 ± 0.2 2.20 ± 0.1

1

)

0.0193 0.0182

qe(cal) (mg g

Pseudo second-order kinetics model 1

)

1.54 1.66

2

R

k2 (g (mg min)

0.8023 0.9841

0.0711 0.0269

1

)

qe(cal) (mg g1)

R2

8.41 2.29

0.9999 0.9910

Table 3 Langmuir and Freundlich constants for the adsorption of Cu(II) and As(V) on CCM.

Cu(II) As(V)

qm(mg g1)

b(L mg1)

RL

R2

kf

n

R2

14.3 6.5

0.1759 0.0154

0.09–0.79 0.02–0.23

0.8006 0.9876

5.8277 0.2816

6.08 1.79

0.9781 0.9687

661

D.-W. Cho et al. / Chemical Engineering Journal 200–202 (2012) 654–662 Table 5 Thermodynamic parameters for the adsorption of Cu(II) and As(V) on CCM. qm(mg g1) Cu(II)

Temperature(°C)

15.8 18.5 20.5 7.5 9.3 10.4

As(V)

K 5

25 35 45 25 35 45

1.52  10 1.63  105 1.85  105 3.34  104 5.16  104 7.82  104

distribution cross-linked chitosan [37]. But the data fitting to the Freundlich and Langmuir isotherms for As(V) (0.9862, 0.9687) suggests that both isotherms show a good fit to the experimental data with well-matching correlation coefficients. Similar observations were reported in the previous study that investigated both monolayer and multilayer sorption of As(V) onto Fe3O4 impregnated on surface of support [48]. 3.4. Thermodynamics of adsorption and pH effect To study the thermodynamics of adsorption of Cu(II) and As(V) on CCM, the thermodynamic parameters, free energy change (DG°), enthalpy change (DH°) and entropy change (DS°), were calculated by using the following equations:

DG ¼ RT ln K ln

K 2 DH ¼ K1 R

ð7Þ

 

1 1  T1 T2

 ð8Þ

DG ¼ DH  T DS

ð9Þ

where R is the universal gas constant (8.314 J (mol K)1), T is the temperature in Kelvin and K is the equilibrium constant calculated by the following equation [49]:

K ¼ b  55:5

DG° (kJ mol1)

DS° (J mol1K1)

DH°(kJ mol1)

29.6 30.7 32.1 25.8 27.8 29.8

117.9 117.9 118.3 198.5 198.5 198.6

5.6

33.4

cating adsorption processes were endothermic. Negative DG° values (Cu(II): 29.6, 30.7, and 32.1; As(V): 25.8, 27.8, and 29.8 kJ mol1) indicate the reactions were thermodynamically favorable and spontaneous. Further, positive values of DS° (Cu(II): 117.9, 117.9, and 118.3; As(V): 198.5, 198.5, and 198.6 J (mol K)1) indicate the increased randomness at the solid-solution interface during the fixation of adsorbates on the active sites of the adsorbent. The effect of pH on the adsorption of Cu(II) and As(V) by CCM in the pH range between 3 and 9 is shown in Fig. 7. The adsorption capacity of CCM for Cu(II) increased as the solution pH increased to give 99.9% removal efficiency at pH 9. The lower adsorption of Cu(II) at lower pH values can be attributed to the development of positive surface charge on clay that created unfavorable electrostatic condition for cation adsorption. Also, the protonation of amine and hydroxyl groups in chitosan at low pH can inhibit the approach of Cu(II) due to the electrostatic repulsion [51]. On the other hand, substantially less adsorption of As(V) was observed at higher pH values, probably because As(V) speciation shifted to doubly charged HAsO2 4 [52], and also surface charge of the adsorbent became negative [53]. The results indicate that both two adsorptions are dependent on solution pH and show the conflicting adsorption behavior with respect to pH variation. 4. Conclusions

ð10Þ

The values of these parameters are given in Table 5. It was observed that the adsorption amount of Cu(II) and As(V) increased from 15.8 and 7.5 to 20.5 and 10.4 mg g1, respectively, with the temperature increase from 25 to 45 °C. As a result, the calculated DH° (Cu(II): 5.6; As(V): 33.4 kJ mol1) were positive values, indi-

120

Adsorbed amount_Cu(II) Adsorbed amount_As(V) RE(%)_Cu(II) RE(%)_As(V)

90

qe (mg g-1)

20

60

10 30

Removal Efficiency (%)

30

0

0 3

4

5

6

7

8

A magnetic composites was synthesized with chitosan, nanomagnetite, and heulandite to remove Cu(II) and As(V) from aqueous solution. The optimized mass ratio among materials appeared to be 1:1:2 for removal of both ions. The adsorption process for Cu(II) was faster than that for As(V), and the sorption equilibrium was achieved in 70 and 120 min for Cu(II) and As(V), respectively. Pseudo-second order kinetics model was the better model to describe adsorption behavior for both ions. The adsorption capacities for Cu(II) and As(V) of CCM at initial concentrations of 656 and 336 mg L1 were 17.2 and 5.9 mg g1, respectively. The adsorption data for Cu(II) was well described by Freundlich isotherm model and As(V) data was fitted well by both isotherm models. Adsorption capacities of CCM for Cu(II) and As(V) increased with an increase in temperature from 25 to 45 °C, indicating the adsorption process is endothermic and thermodynamically favorable. The adsorption capacity of CCM for Cu(II) increased as the solution pH increased from 3 to 9 and As(V) adsorption showed an opposite trend. Acknowledgment This work was supported by Global Research Laboratory Project (Ministry of Education, Science and Technology, B030305).

9

pH

Appendix A. Supplementary material 1

Fig. 7. Effect of pH on adsorbed amount (mg g ) and removal efficiencies (%) of Cu(II) and As(V) by CCM (Initial conc. of Cu(II) and As(V): 68.7 and 66.9 mg L1; adsorbent dose: 5 g L1, contact time = 24 h).

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cej.2012.06.126.

662

D.-W. Cho et al. / Chemical Engineering Journal 200–202 (2012) 654–662

References [1] M.L.P. Dalida, A.F.V. Mariano, C.M. Futalan, C.-C. Kan, W.-C. Tsai, M.-W. Wan, Adsorptive removal of Cu(II) from aqueous solutions using non-crosslinked and crosslinked chitosan-coated bentonite beads, Desalination 275 (2011) 154–159. [2] V. Lenoble, O. Bouras, V. Deluchat, B. Serpaud, J.-C. Bollinger, Arsenic adsorption onto pillared clays and iron oxides, J. Colloid Interf. Sci. 255 (2002) 52–58. [3] M. Biterna, A. Arditsoglou, E. Tsikouras, D. Voutsa, Arsenate removal by zero valent iron: batch and column tests, J. Hazard. Mater. 149 (2007) 548–552. [4] J.-C. Lin, A.K. Sengupta, Hybrid anion exchange fibers with dual binding sites: simultaneous and reversible sorption of perchlorate and arsenate, Environ. Eng. Sci. 26 (2009) 1673–1683. [5] C. Gutiérrez, H.K. Hansen, P. Nuñez, P.E. Jensen, L.M. Ottosen, Electrochemical peroxidation as a tool to remove arsenic and copper from smelter wastewater, J. Appl. Electrochem. 40 (2010) 1031–1038. [6] B. Al-Rashdi, C. Somerfield, N. Hilal, Heavy metals removal using adsorption and nanofiltration techniques, Sep. Purif. Rev. 40 (2011) 209–259. [7] O. Arar, U. Yuksel, N. Kabay, M. Yuksel, Removal of Cu2+ ions by a micro-flow electrodeionization (EDI) system, Desalination 277 (2011) 296–300. [8] H. Modin, K.M. Persson, A. Andersson, M. van Praagh, Removal of metals from landfill leachate by sorption to activated carbon, bone meal and iron fines, J. Hazard. Mater. 189 (2011) 749–754. [9] V.M. Boddu, K. Abburi, J.L. Talbott, E.D. Smith, R. Haasch, Removal of arsenic (III) and arsenic (V) from aqueous medium using chitosan-coated biosorbent, Water Res. 42 (2008) 633–642. [10] S.R. Popuri, Y. Vijaya, V.M. Boddu, K. Abburi, Adsorptive removal of copper and nickel ions from water using chitosan coated PVC beads, Bioresource Technol. 100 (2009) 194–199. [11] M.V. Dinu, E.S. Dragan, Evaluation of Cu2+, Co2+ and Ni2+ ions removal from aqueous solution using a novel chitosan/clinoptilolite composite: kinetics and isotherms, Chem. Eng. J. 160 (2010) 157–163. [12] F.-C. Wu, R.-L. Tseng, R.-S. Juang, Adsorption of dyes and humic acid from water using chitosan-encapsulated activated carbon, J. Chem. Technol. Biot. 77 (2002) 1269–1279. [13] M.-W. Wan, C.-C. Kan, B.D. Rogel, M.L.P. Dalida, Adsorption of copper (II) and lead (II) ions from aqueous solution on chitosan-coated sand, Carbohyd. Polym. 80 (2010) 891–899. [14] A.-H. Chen, S.-C. Liu, C.-Y. Chen, C.-Y. Chen, Comparative adsorption of Cu(II), Zn(II), and Pb(II) ions in aqueous solution on the crosslinked chitosan with epichlorohydrin, J. Hazard. Mater. 154 (2008) 184–191. [15] M.K. Sureshkumar, D. Das, M.B. Mallia, P.C. Gupta, Adsorption of uranium from aqueous solution using chitosan-tripolyphosphate (CTPP) beads, J. Hazard. Mater. 184 (2010) 65–72. [16] H.-Y. Zhu, R. Jiang, L. Xiao, Adsorption of an anionic azo dye by chitosan/kaolin/ c-Fe2O3 composites, Appl. Clay Sci. 48 (2010) 522–526. [17] L. Fan, C. Luo, Z. Lv, F. Lu, H. Qiu, Preparation of magnetic modified chitosan and adsorption of Zn2+ from aqueous solutions, Colloid. Surface. B 88 (2011) 574–581. [18] L. Fan, C. Luo, Z. Lv, F. Lu, H. Qiu, Removal of Ag+ from water environment using a novel magnetic thiourea-chitosan imprinted Ag+, J. Hazard. Mater. 194 (2011) 193–201. [19] M.T. Klepka, N. Nedelko, J.-M. Greneche, K. Lawniczak-Jablonska, I.N. Demchenko, A. Slawska-Waniewska, C.A. Rodrigues, A. Debrassi, C. Bordini, Local atomic structure and magnetic ordering of iron in Fe–chitosan complexes, Biomacromolecules 9 (2008) 1586–1594. [20] C. Yuwei, W. Jianlong, Preparation and characterization of magnetic chitosan nanoparticles and its application for Cu(II) removal, Chem. Eng. J. 168 (2011) 286–292. [21] S. Zhang, X.-Y. Li, J.P. Chen, Preparation and evaluation of a magnetite-doped activated carbon fiber for enhanced arsenic removal, Carbon 48 (2010) 60– 67. [22] J. Pan, H. Yao, X. Li, B. Wang, P. Huo, W. Xu, H. Ou, Y. Yan, Synthesis of chitosan/ c-Fe2O3/fly-ash-cenospheres composites for the fast removal of bisphenol A and 2,4,6-trichlorophenol from aqueous solutions, J. Hazard. Mater. 190 (2011) 276–284. [23] B. Calvo, L. Canoira, F. Morante, J.M. Martínez-Bedia, C. Vinagre, J.-E. GarcíaGonzález, J. Elsen, R. Alcantara, Continuous elimination of Pb2+, Cu2+, Zn2+, H+ and NHþ 4 from acidic waters by ionic exchange on natural zeolites, J. Hazard. Mater. 166 (2009) 619–627. [24] T.J. Daou, S. Begin-Colin, J.M. Grenèche, F. Thomas, A. Derory, P. Bernhardt, P. Legaré, G. Pourroy, Phosphate adsorption properties of magnetite-based nanoparticles, Chem. Mater. 19 (2007) 4494–4505. [25] C.M. Futalan, C.-C. Kan, M.L. Dalida, K.-J. Hsien, C. Pascua, M.-W. Wan, Comparative and competitive adsorption of copper, lead, and nickel using chitosan immobilized on bentonite, Carbohyd. Polym. 83 (2011) 528–536.

[26] Yu.N. Vodyanitskii, E.G. Morgun, L.A. Obydenova, K.A. Rumyantseva, N.V. Chapygina, Geochemistry of magnetite and maghemite in soils in European Russia, Geochem. Int. 47 (2009) 297–310. [27] W.M. Meier, D.H. Olson, Atlas of Zeolite Structure Types, third ed., Butterworth-Heinemann, London, 1992. [28] R.L. Hay, Mineralogy and geology of natural zeolites, in: F.A. Mumpton, (Ed.), Geology of zeolites in sedimentary rocks Mineralogical Society of America, Washington, DC, 1981, pp. 53–64. [29] D.S. Coombs, A. Alberti, T. Armbruster, G. Artioli, C. Colela, E. Galli, J.D. Grice, F. Liebau, J.A. Mandarino, H. Minato, E.H. Nickel, E. Passaglia, D.R. Peacor, S. Quartieri, R. Rinaldi, M. Ross, R.A. Sheppard, E. Tillmanns, G. Vezzalini, Recommended nomenclature for zeolite minerals: report of the subcommittee on zeolites of the international mineralogical association, commission on new minerals and mineral names, Can. Mineral. 35 (1997) 1571–1606. [30] P. Yang, J. Stolz, T. Armbruster, M.E. Gunter, Na, K, Rb, and Cs exchange in heulandite single crystals: diffusion kinetics, Am. Mineral. 82 (1997) 517–525. [31] N. Kantiranis, K. Sikalidis, A. Godelitsas, C. Squires, G. Papastergios, A. Filippidis, Extra-framework cation release from heulandite-type rich tuffs on exchange with NHþ 4 , J. Environ. Manage. 92 (2011) 1569–1576. [32] C. Muzzarelli, O. Francescanheli, G. Tosi, R.A.A. Muzzarelli, Susceptibility of dibutyryl chitin and regenerated chitin fibres to deacylation and depolymerization by lipases, Carbohyd. Polym. 56 (2004) 137–146. [33] J. Zawadzki, H. Kaczmarek, Thermal treatment of chitosan in various conditions, Carbohyd. Polym. 80 (2010) 394–400. [34] Th. Perraki, A. Orfanoudaki, Mineralogical study of zeolites from Pentalofos area, Thrace, Greece. Appl. Clay Sci. 25 (2004) 9–16. [35] H.C. Riese, An infrared absorption analysis of magnetite, Am. Mineral. 52 (1967) 1198–1205. [36] J.-H. An, S. Dultz, Adsorption of Cr(VI) and As(V) on chitosan-montmorillonite: selectivity and pH dependence, Clay. Clay Miner. 56 (2008) 549–557. [37] C. Gérente, Y. Andrès, G. McKay, P. Le Cloirec, Removal of arsenic(V) onto chitosan: from sorption mechanism explanation to dynamic water treatment process, Chem. Eng. J. 158 (2010) 593–598. [38] K. Ohe, T. Oshima, Y. Baba, Adsorption of arsenic using high surface area magnetites, Environ. Geochem. Health 32 (2010) 283–286. [39] H.J. Shipley, S. Yean, A.T. Kan, M.B. Tomson, A sorption kinetics model for arsenic adsorption to magnetite nanoparticles, Environ. Sci. Pollut. R. 17 (2010) 1053–1062. [40] R. Cortés-Martínez, V. Martínez-Miranda, M. Solache-Ríos, I. García-Sosa, Evaluation of natural and surfactant-modified zeolites in the removal of cadmium from aqaeous solutions, Separ. Sci. Technol. 39 (2004) 2711–2730. [41] V.W.D. Chui, K.W. Mok, C.Y. Ng, B.P. Luong, K.K. Ma, Removal and recovery of copper(II), chromium(III), and nickel(II) from solutions using crude shrimp chitin packed in small columns, Environ. Int. 22 (1996) 463–468. [42] W.S.W. Ngah, S. Fatinathan, Adsorption of Cu(II) ions in aqueous solution using chitosan beads, chitosan-GLA beads and chitosan-alginate beads, Chem. Eng. J. 143 (2008) 62–72. [43] M.R. Gandhi, G.N. Kousalya, S. Meenakshi, Selective sorption of Fe(III) using modified forms of chitosan beads, J. Appl. Polym. Sci. 124 (2011) 1858–1865. [44] K.G. Bhattacharyya, S.S. Gupta, Removal of Cu(II) by natural and acid-activated clays: an insight of adsorption isotherm, kinetic and thermodynamics, Desalination 272 (2011) 66–75. [45] M.S. Rana, M.A. Halim, S.A.M.W. Hoque, K. Hasan, M.K. Hossain, Bioadsorbtion of arsenic by prepared and commercial crab shell chitosan, Biotechnology 8 (2009) 160–165. [46] S.M. Miller, J.B. Zimmerman, Novel, bio-based, photoactive arsenic sorbent: TiO2-impregnated chitosan bead, Water Res. 44 (2010) 5722–5729. [47] M. Rajiv Gandhi, G.N. Kousalya, N. Viswanathan, S. Meenakshi, Sorption behaviour of copper on chemically modified chitosan beads from aqueous solution, Carbohyd. Polym. 83 (2011) 1082–1087. [48] R. Chen, C. Zhi, H. Yang, Y. Bando, Z. Zhang, N. Sugiur, D. Golberg, Arsenic (V) adsorption on Fe3O4 nanoparticle-coated boron nitride nanotubes, J. Colloid Interf. Sci. 359 (2011) 261–268. [49] D.W. Cho, C.M. Chon, Y. Kim, B.H. Jeon, F.W. Schwartz, E.S. Lee, H. Song, Adsorption of nitrate and Cr(VI) by cationic polymer-modified granular activated carbon, Chem. Eng. J. 175 (2011) 298–305. [50] D. Das, M.K. Sureshkumar, K. Radhakrishnan, J. Nuwar, C.G.S. Pillai, Adsorptive removal of Cr(III) from aqueous solution using tripolyphosphate cross-linked chitosan beads, J. Radioanal. Nucl. Ch. 289 (2011) 275–285. [51] L. Jin, R. Bai, Mechanisms of lead adsorption on chitosan/PVA hydrogel beads, Langmuir 18 (2002) 9765–9770. [52] J. Giménez, M. Martínez, J. de Pablo, M. Rovira, L. Duro, Arsenic sorption onto natural hematite, magnetite, and goethite, J. Hazard. Mater. 141 (2007) 575–580. [53] K. Simeonidis, T. Gkinis, S. Tresintsi, C. Martinez-Boubeta, G. Vourlias, I. Tsiaoussis, G. Stavropoulos, M. Mitrakas, M. Angelakeris, Magnetic separation of hematite-coated Fe3O4 particles used as arsenic adsorbents, Chem. Eng. J. 168 (2011) 1008–1015.