Equilibrium modeling of As(III,V) sorption in the absence/presence of some groundwater occurring ions by iron(III)–cerium(IV) oxide nanoparticle agglomerates: A mechanistic approach

Chemical Engineering Journal 228 (2013) 665–678

Contents lists available at SciVerse ScienceDirect

Chemical Engineering Journal

journal homepage: www.elsevier .com/locate /cej

Equilibrium modeling of As(III,V) sorption in the absence/presenceof some groundwater occurring ions by iron(III)–cerium(IV) oxidenanoparticle agglomerates: A mechanistic approach of surfaceinteraction

1385-8947/$ - see front matter � 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.cej.2013.05.037

⇑ Corresponding author. Tel.: +91 33 2241 3893.E-mail address: [email protected] (U.C. Ghosh).

Tina Basu a, Debabrata Nandi a, Pintu Sen b, Uday Chand Ghosh a,⇑a Department of Chemistry, Presidency University (Erstwhile Presidency College), 86/1 College Street, Kolkata 700073, Indiab Department of Physics, VECC, 1/AF Bidhannagar, Kolkata 700064, India

h i g h l i g h t s

� Eco-friendly, green synthetic routewas exploited for the synthesis ofsorbent.� Characterized as crystalline

agglomerated nanoparticles withmicroporosity.� Equilibrium data described Langmuir

and D-R isotherm equation fairlywell.� Rate of desorption established the

degree of binding affinity oversorbent surface.� Oxidation of surface sorbed As(III) in

a thermodynamically controlledreaction.

g r a p h i c a l a b s t r a c t

A pictorial representation of As(III)-sorption mechanism taking place under sufficient time lag is attrib-uted to the oxidizing nature of Cerium (IV) in acidic medium. Cerium (IV) by virtue of its high standardreduction potential value (Ce4+/Ce3+ = 1.72) in acidic medium oxidizes the surface sorbed As(III) to As(V)(As3+/As5+ = 0.56) itself getting reduced to Ce3+ according to the underlying red ox reaction.As3þ þ 2Ce4þ ¼ As5þ þ 2Ce3þ.

a r t i c l e i n f o

Article history:Received 25 February 2013Received in revised form 6 May 2013Accepted 11 May 2013Available online 22 May 2013

Keywords:ArsenicCerium–iron oxideSorptionGroundwater ionsMechanismSpontaneity

a b s t r a c t

Here, we aim to develop an efficient material by eco-friendly green synthetic route that was further char-acterized to be crystalline ranging in nano-dimension for filtering high arsenic content groundwater. Thethermal stability of iron(III)–cerium(IV) mixed oxide nanoparticle agglomerates (NICMO) was well estab-lished from the consistent particle size at different temperature and also from differential thermal anal-ysis. The bimetal mixed oxide contained agglomerated crystalline nano-particles of dimension 10–20 nmheld together by crystal packing forces, and its corresponding empirical composition was FeCe1.1O7.6.Appearance of a weak band at 534 cm�1 in the spectrum of nano-structured iron(III)–cerium(IV) mixedoxide (NICMO) is presumed for the presence of hetero-metal bonding via oxygen linkage (i.e., Fe–O–Ce). Equilibrium sorption data described Langmuir and D-R isotherm equations fairly well particularlyfor As(III) with relatively high monolayer sorption capacity [55.513 mg g�1 for As(V), 86.293 mg g�1 forAs(III)] in the absence of any foreign ions. Chemo-sorption is the actual nature of the reaction taking placein As(III) with the sorption process getting more favorable with the increase of temperature in contrast toAs(V) in which the degree of interaction suggested physiosorption type reactions. Splitting of band inFTIR spectrum of As(V) suggested the dominance of mono protonated monodentate complex [S-OAsO2

666 T. Basu et al. / Chemical Engineering Journal 228 (2013) 665–678

(OH)] on the oxide surface. As(III) sorption mechanism taking place over NICMO surface under sufficienttime lag confirmed oxidation of surface adsorbed As(III) to As(V) in a thermodynamically controlled sorp-tion reaction.

� 2013 Elsevier B.V. All rights reserved.

1. Introduction

Remediation of metal ions including the heavy metal varietyfrom the contaminated areas has become a significant topic of re-search due to the severe risks they pose to human health and theenvironment [1]. Thus, the development of novel systems forremediation is required because unlike organic pollutants toxicmetal ions can be a re-circulating contaminant [2,3] and cannotbe metabolized or decomposed. Consequently, various healths re-lated problems have been found to increase at an alarming rate.For example, millions of the population in various parts of theworld had been suffering from skin lesions, cancers and other re-lated diseases owing to the consumption of high arsenic(10 lg L�1) ground water since last couple of decades [4–11]. Inorder to reduce the arsenic exposure through consumption ofdrinking water, the United States Environment Protection Agency(USEPA), the World Health Organisation (WHO) and the Ministryof Health of the People’s Republic of China had lowered their reg-ulatory or guideline values from 50 to 10 lg L�1 for Astotal concen-tration in drinking water [12], which encouraged the researchersto develop treatment methods for arsenic-rich surface/groundwater.

In consequence, a number of treatment methods were investi-gated for the arsenic removal from contaminated water, includingflotation [13], precipitation with sulphide [14], coagulation [15],ion-exchange [16], desalting techniques such as reverse osmosisor electro-dialysis [17] and surface adsorption [16,18]. Thesemethods excepting surface adsorption are practically not accept-able for the point of use applications in rural areas owing to therequirement of (a) continuous attention during the treatment pro-cess, (b) large installation space, and (c) large space for the sludgedisposal.

Abundant Fe(III) oxide occurring in various mineralogical formsin the earth crust is a well-known natural water purifier [19–22],which has good arsenic adsorption capacity in a wide pH range.Hence, a number of surface modified Fe(III) oxide had been inves-tigated for the purpose of arsenic adsorption [23–25]. Amongthem, mixed oxides of Fe(III) with some other metal ions [26–30]showed higher arsenic removal efficiency than the single metaloxides [19,21,31–33], because incorporation of other metal withFe(III) oxide changes some physical properties such as surface area,surface charge and crystallinity [32,34].

Based on which different arsenic removal filters (ARFs) hadbeen developed and marketed by different companies in theircommercial trade name. Evaluation and the level of performanceof those ARFs were assessed to be poor in at least 82% of the in-stalled ARFs at Bengal delta basin of ‘The Ganga’ river (Technolog-ical Park, Baruipore, S-24 Parganas, West Bengal, India) [35,36].Reasons for the poor performance of ARFs include impropermaintenance, sand gushing problems, lack of user-friendlinessand absence of community participation. Ineffectiveness and poorreliability of the above mentioned ARFs are presumably due tothe adverse influence of the dissolved ions that includes sulfate,phosphate, chloride, bicarbonate, silicate, etc., on arsenic sorptionreaction usually co-occurring in groundwater. Therefore, accurateprediction of arsenic sorption behavior by the synthetic solid sor-bents is required for the different field water environment in or-der to properly design and operate the arsenic removal treatmentsystem.

In last few years, many new upcoming strategies have beenundertaken for the removal of arsenic using nanostructured ormesoporous materials [37–40].

Recently, ceria nano-particles [41] had been considered to be arepresentative member of an industrially important class of metaloxide nanoparticles [42,43]. They could be used as automotive cat-alytic converters [44], UV-blocking agents [45], and single, nano-wire-based gas sensors [46]. Owing to the presence of highaffinity surface hydroxyl groups, hydrous cerium oxide (HCO)nano-particles [41] showed encouraging arsenic adsorptioncapacity.

Hence, it had been aimed that the equilibrium arsenic removalbehavior shall be investigated from thermodynamic perspective byiron(III)–cerium(IV) mixed oxide nanoparticle agglomerates (NIC-MO) in the presence of major groundwater occurring ions for judg-ing its practical performance in treatment of high arsenic groundwater sample.

Thus, this manuscript reports the results of As(V, III) sorptionequilibriums with mechanism on synthetic NICMO in the ab-sence/presence of some ions commonly occurring in ground water.

2. Materials and methods

2.1. Chemicals

Disodium hydrogen arsenate heptahydrate (Na2HAsO4�7H2O, E.Merck, Germany) and arsenic trioxide (As2O3) (Analar Reagent,BDH, England) were used for the preparation of stock solutionsof As(V) and As(III), respectively. Ferric chloride hexahydrate(FeCl3�6H2O, Merck, India), sodium hydroxide (NaOH, SD FineChemicals, India) and ammonium cerium(IV) nitrate (ACN),[(NH4)2Ce(NO3)6, Merck, India] used for NICMO synthesis were lab-oratory grade reagents. Silver diethyl dithio carbamate (SDDC) andsodium borohydride used for the arsenic analysis were procuredfrom E. Merck (Germany).

Other chemicals such as sodium sulphate (Na2SO4), sodiumchloride (NaCl), sodium phosphate (Na3PO4), sodium hydrogencarbonate (NaHCO3) and sodium silicate (Na2SiO3) used at back-ground as a potential source of sulphate (SO2�

4 ), chloride (Cl�),phosphate (PO3�

4 ), bicarbonate (HCO�3 ) and silicate (SiO2�3 ) ions

were also procured, respectively, from Merck, Mumbai (India).

2.2. Synthesis of NICMO

Solutions of ferric chloride (0.1 M) and ACN (0.1 M) were pre-pared separately by dissolving their appropriate amounts in0.1 M hydrochloric acid. Then, ACN solution was added slowly toferric chloride solution with vigorous stirring (v/v = 1:1). To it5(M) sodium hydroxide solution was added drop wise into theabove mixture with continuous stirring to increase pH 9.0–9.5.

The gel like precipitate with mother liquid was allowed to standfor 48 h before filtration without disturbing. The filtered precipi-tate was washed three times with distilled water and dried at100 �C in an air oven. The dried product was ground in a mortarand pestle and sieved to separate the agglomerates having dimen-sion ranged in 140–290 lm. The sieved agglomerated NICMO par-ticles (140–290 lm) was homogenized at pH 7.0 (±0.2) by stirringin water adjusted at pH 7.0 to make the solid surface near to pHzpc

(7.13) before its use in experiments. This homogenized sorbent

T. Basu et al. / Chemical Engineering Journal 228 (2013) 665–678 667

(NICMO) was used in physicochemical experiments like calcina-tions temperature, pH effect, kinetic, equilibrium isotherm, etc.to study its sorption behavior.

2.3. Preparation of stock arsenic solution

A stock of standard As(V) solution (1000 mg L�1) was preparedby dissolving 4.160 g of sodium arsenate (Na2HAsO4�7H2O) in1000 mL of arsenic-free distilled water. The analyzed arsenic con-centration was 1300 mg L�1. A measured volume of this stock solu-tion was diluted with arsenic-free distilled water to attain adesired concentration level in solution prior to use in experiments.

A stock of standard As(III) solution (1000 mg L�1) was preparedby dissolving 0.132 g of As2O3 in 10 mL of 4% (w/v) NaOH solution,acidified with 2.0 mL of concentrated HCl and diluted to 100 mLwith arsenic-free distilled water. The analyzed concentration ofAs(III) was 980 mg L�1. The stock was prepared afresh every after3 day and maintained at low temperature to prevent oxidation. Ameasured volume of this stock was diluted with distilled waterfor a desired level of As(III) concentration in solution prior to usein experiments.

2.4. Analytical instruments

X-ray diffraction (XRD) analysis of NICMO was conducted usinga X-ray powder diffractometer (Philips Analytical PW-1710)equipped with Cu Ka radiation at a scanning speed 2� per min be-tween 2h values 10� and 90� operated at voltage 40 kV and appliedpotential current 30 mA.

Thermogravimetric (TG) and differential thermal (DT) spectra ofNICMO were recorded using a Setaram analyzer in argon atmo-sphere at a heating rate of 20 �C min�1 over a temperature range50–800 �C to detect the degree of thermal stability and analyzethe possibility of any phase transformation of the oxide withincreasing temperature of heat treatment.

Föurier transform infra red (FTIR) spectra of NICMO and ar-senic(V, III)-sorbed NICMO were recorded by infrared spectropho-tometer (Jasco 680 plus) with a resolution of 2 cm�1 making thinpellet of the samples prepared by mixing spec pure KBr in a ratio3:100.

Zero-point surface charge (pHZPC) was analyzed by pH metrictitration using a pH meter (Elico, model: Li-127).

Scanning electron microscopic (SEM) images along with itsEDAX technology (Tescan Vega, UK; model LSU+) of NICMO andAs(III)-sorbed NICMO was recorded by sample spraying over thecarbon tape to detect the surface morphology and composition ofthe oxide before and after sorption reaction.

Transmission electron micrograph (TEM) image of NICMO alongwith its As(III) sorbed variety was recorded by FEI high-resolutionelectron microscope (model: Tecnai STWIN) to estimate the actualparticle size and also to visualize the nature of agglomeration ofthe synthetic oxides. Here, the samples acquiring TEM image weredispersed in isopropanol by sonication, which was dropped to castonto 200 mesh copper grids coated with a holey carbon film.

The surface area and pore width distributions were analyzed byQuatachrome Autosorb-1C surface analyzer.

Raman scattering measurements were performed in the backscattering configuration using micro-Raman Jobin Yvon T64000system to establish the bond formation between different species.

Table 1Concentration of different ions added to aqueous solution of As(V) (Co = 6.5 and 4.5 mg L�

Different ions Sulphate (SO2�4 ) Chloride (Cl�)

Concentration (mg L�1) 200.0 400.0

Concentrations of arsenic in samples were estimated using UV–Vis spectrophotometer (Hitachi, model-3210U) and atomic absorp-tion spectrophotometer (AAS) (Perkin–Elmer, AAnalyst-200).

Electrochemical behavior of the samples through cyclic voltam-metry measurement was investigated using AUTOLAB-30 poten-tiostat/galvanostat. A platinum electrode and a saturated Ag/AgClelectrode were used as counter and the reference electrodes,respectively. Cyclic voltammograms (CVs) of samples were takenbetween �0.5 and 0.5 V with respect to the reference electrodein 1(M) KCl solution at a scan rate of 5 mV s�1.

2.5. Experimental procedure

2.5.1. Sorption experiments2.5.1.1. General batch experiment. Batch sorption studies were con-ducted by mechanical agitation using a thermostat shaker at 300(±10) rpm and temperature, T = 30 (±1) �C. In 250 mL polythene(PET) bottles, 50 mL arsenic (AsV or AsIII) solution (pH = 7.0 ± 0.2)at a desired level of initial concentration (Co, mg L�1) was mixedwith 0.1 g NICMO and agitated. After agitation for a desired timeperiod, the reaction mixture was filtered using 0.45 lm membranefilter. The filtrates were analyzed for the residual arsenic concen-tration (Ce, mg L�1) using UV–VIS spectrophotometer and AASspectrophotometer based on hydride generator method. The sorp-tion amount (qe, mg g�1) at equilibrium was calculated using thefollowing equation,

qe ¼ðCo � CeÞV

mð1Þ

where Co and Ce have their usual significance given earlier; m is themass of the sorbent (g) and V is volume (L) of the test solutions.

2.5.1.2. Effect of calcined temperature on NICMO. For the effect ofcalcined temperature on NICMO, the material was calcined(1.0 h) separately at 100, 150, 200, 250, 300, 350, 400 and 500 �Crespectively before being used in arsenic sorption experiments de-scribed above.

2.5.1.3. Influence of pH. Arsenic sorption experiment for the influ-ence of pH on As(V,III) sorption by NICMO (200 �C), sorption exper-iment was conducted for 2.0 h separately taking arsenic solutionadjusted at initial pH (pHi) = 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0 and10.0 respectively.

2.5.1.4. Isotherm experiment for equilibrium. Isotherm experimentsfor equilibrium on arsenic (AsIII,V) sorption reactions with NICMOin the absence/presence of some anions occurring as major ingroundwater (Table 1) were conducted at T = 30 (±1.0) �C andpHi = 7.0 (±0.1) taking initial arsenic solution (Co) in the range10.0 to 250.0 mg L�1. The set-up and other experimental condi-tions remained unchanged as described in first paragraph.

The sorption capacity (qe) (mg g�1) values were calculated usingthe mass balance Eq. (1) given above.

2.5.1.5. Batch experiment for thermodynamic parameters. For ther-modynamic parameters, the equilibrium experiment describedabove was conducted at pHi = 7.0 (±0.1) separately and T (±1)K = 283, 293, 303, 313 and 323 respectively placing 100 mg NICMO

1)/As(III) (C0 = 4.8 mg L�1) at T = 30 (±1.0) �C and pHi 7.0 (±0.1).

Phosphate (PO3�4 ) Bicarbonate (HCO�3 ) Silicate (SiO2�

3 )

50.0 300.0 50.0

668 T. Basu et al. / Chemical Engineering Journal 228 (2013) 665–678

with As(V, III) solution in the absence/presence of ions at back-ground (Table 1) for determining the equilibrium constants.

2.5.2. Desorption experimentBefore the desorption experiment, 2.0 g NICMO (200 �C) was

mixed with 50 mL of 1000 mg L�1 As(V)/As(III) solution separatelytaking inside the PET bottles and agitated for 2.0 h. After adsorp-tion, arsenic content in solid phase (mg g�1) was 16.15 and 24.95for As(V) and As(III) respectively. The filtered solid mass from thereaction mixture was washed three times with distilled water,and then, the solid arsenic sorbed NICMO was dried in an air ovenat 100 �C. 0.1 g portions of As-enriched NICMO were mixed insidethe PET bottles with 50 mL alkali solution and agitated for 2 hemploying molar strength such as 0.01, 0.05, 0.1, 0.25, 0.5, 1.0,1.0, 1.5, and 2.0. After filtering the solid particles from each reac-tion mixture, the filtrate was analyzed for arsenic and accordingly,the percentage of desorbed arsenic was calculated.

2.5.3. Source of ground water sample and experimental set upGround water sample was collected from a tube well (depth:

40–45 m) at College Street near Presidency University Campus,Kolkata which was analyzed for arsenic and some water qualityparameters. The concentration of estimated parameters (mg L�1,except pH) is given in parenthesis against each: Astotal

(1.3 � 10�3), pH (7.22), Fe2+ (0.15), F� (0.345), HCO�3 (198), Hard-ness (252), Ca2+ (88), Cl� (276), TDS (528).

This ground water sample was then spiked with As(III) solutionuntil the initial total arsenic concentration (C0) reached0.19 mg L�1 which was used for the fixed-bed column experimentto access the filtration efficiency of NICMO.

Fixed-bed column experiment was conducted using uniformlypacked NICMO over glass wool sheet up to a bed height of6.0 cm in glass tubes of internal diameter (i.d.) 1.4 cm and length25.0 cm taking care of all possible precautions. The said groundwa-ter sample spiked with As(III) up to a level 0.19 mg L�1 was passedthrough the NICMO packed fixed bed column for filtration (flowrate: 1.0 mL min�1). The filtered water was collected in 100 mLfractions and analyzed for the arsenic concentration.

3. Results and discussion

3.1. Characterization of NICMO

The physicochemical characterization of NICMO has been givenin Supporting information (SI) section, and summarized in Table 2.Analysis of the XRD patterns of NICMO samples prepared by incin-erating at temperature (C) 100�, 200�, 300�, 400� and 500� sug-gested that the sample material did neither segregate noraggregate with increasing temperature indicating reasonable ther-mal stability. The average particle size calculated by inserting theXRD data into the Scherrer’s equation had been obtained to beabout 3.56 nm. FTIR analysis revealed (SI section) that a weak bandat wave number 534 cm�1 in the spectrum of NICMO probably ap-

Table 2Summarized physical characteristic features of NICMO sample.

Physical characteristic parameters of NICMO Physical characteristics

Average particle size (XRD analysis (nm) 3.56Band assignment in FTIR spectra (cm�1) Hetero-metal bridging at

534pHZPC 7.13 (±0.1)Empirical formula derived from SEM/EDX

analysisFeCe1.1 O7.6

Particle size (TEM) image analysis (nm) 10–20BET surface area (m2 g�1) 104

peared due to the presence of Fe–O–Ce unit. Agglomerated surfacemorphology of NICMO with irregular shape was evident from itsSEM image (SI section) analysis that looked like uneven dispersionof NICMO clusters over a base matrix of oxide surface. Microcrys-talline nature with crystallite size 8–10 nm packed within NICMOagglomerates under crystal packing forces is estimated from TEMimage analysis which is about 2.5 times greater compared toXRD data. The surface area (m2 g�1) of the material is 104 as ob-tained from the BET isotherm analysis, which is lesser than pureCe(IV) oxide (122) and higher than Fe(III) oxide (98) [27]. Higherpore volume (0.132 cm3 g�1) estimated from N2(V) adsorption–desorption isotherm predicts good adsorbent possibility of thematerial. The sharp peaks observed in the wave number range2000–2300 cm�1 in Raman spectrum (SI section) confirmed thepresence of hetero-metal bonding via oxygen linkage (i.e., Fe–O–Ce).

Effects of some parameters on arsenic sorption by NICMO suchas calcined temperature, pH, solubility loss of sorbent and contacttime are presented in the SI section (S-2 to S-5), respectively. It hadbeen seen that NICMO calcined at 200 �C showed highest arsenicsorption capacity at optimized pH 7.0 (±0.1) with 2.0 h of equilib-rium time and negligible solubility loss of the material.

3.2. Equilibrium modeling

Sorption capacity (qe, mg g�1) obtained in the absence/presenceof some groundwater occurring ions against equilibrium concen-tration (Ce, mg L�1) of As(V) and As(III) have been presented sepa-rately as points in Figs. 1 and 2, respectively. To understand thesorption mechanism, the equilibrium data were analyzed by Lang-muir (Eq. (2)) and Freundlich (Eq. (3)) isotherm models [47].

Langmuir model isotherm which had developed based on thehomogeneous sorption sites is best described by equation.

qe ¼qm KL Ce

1þ KLCeð2Þ

where Ce is the aqueous equilibrium solute concentration (mg L�1),qe is the amount (mg g�1) of solute sorbed per unit gram of sorbentat equilibrium, qm and KL are constants related to the monolayersorption capacity (mg g�1) and the energy of sorption (L mg�1),respectively. qm represents the practical limiting sorption capacitywhen the surface is fully covered with sorbed species. Langmuirconstants qm and KL can be determined from non-linear as well aslinear analyzes of experimental data using Eq. (2)

Freundlich model isotherm which was developed based on themultilayer solute sorption on heterogeneous surface is best de-scribed by equation.

qe ¼ KF C1ne ð3Þ

where KF is the Freundlich sorption capacity and 1/n is an arbitraryconstant related to the sorption intensity. KF and n are empiricalconstants dependent on several environmental factors. Other termshave their usual significance given elsewhere.

The equilibrium data shown as points for As(V) and As(III) inFigs. 1 and 2 were analyzed by both non-linear and linear fit meth-od using Langmuir (Eq. (2)) and Freundlich (Eq. (3)) models,respectively. Non-linear fits of equilibrium data with Langmuirand Freundlich models have also been presented in Figs. 1 and 2,respectively for As(V) and As(III). Linear model fits of equilibriumdata with the said isotherm Eqs. (2) and (3) for both As(V) andAs(III) were also shown in a and b of Figs. S12 and S13, respectively(SI section). The different model parameters estimated from thenon-linear fits for the said isotherms are given separately inTables 3a and 3b for As(V) and As(III), respectively, includingregression coefficient (R2) and statistical error chi-square (v2),

0 20 40 60 80 100 120 140 160 1800

10

20

30

40

50

60Ad

sorp

tion

Cap

acity

(qe)

Equilibrium concentration (Ce)0 20 40 60 80 100 120 140 160 180

10

20

30

40

50

60

Adso

rptio

n C

apac

ity (q

e)

Equilibrium concentration (Ce)(a) (b)

Fig. 1. Plots of the equilibrium adsorption data on As(V) removal by NICMO at pH = 7.0 (±0.1) and T = 30 (±1) �C in the presence of some groundwater occurring ions,including non-linear model fittings with (a) the Langmuir and (b) the Freundlich isotherms. Ion added with As(V) solution at background: j no ion, d SO2�

4 , N Cl�, . HCO�3 , �PO3�

4 , J SiO23�.

0 20 40 60 80 100 120

10

20

30

40

50

60

70

80

Satu

rate

d So

rptio

n C

apac

ity (q

e, mg.

g-1)

Equilibrium Concentration (Ce, mg. L-1)0 20 40 60 80 100 120

10

20

30

40

50

60

70

80

90

100

Satu

rate

d So

rptio

n C

apac

ity (q

e, mg.

g-1

)

Equilibrium Concentration (Ce, mg. L-1)

(a) (b)Fig. 2. Plots of the equilibrium sorption data on As(III) removal by NICMO at pH = 7.0 (±0.1) and T = 30 (±1) �C in the presence of some groundwater occurring ions, includingnon-linear model fittings with (a) the Langmuir and (b) the Freundlich isotherms. Ion added with As(III) solution at background: j no ion, d SO2�

4 , N Cl�, . HCO�3 ,� PO3�4 ,

J SiO2�3 .

T. Basu et al. / Chemical Engineering Journal 228 (2013) 665–678 669

and those estimated from the linear analysis are given in TablesS4a and S4b, respectively, for As(V) and As(III).

Based on the values of v2 or R2, it can be said that the goodnessof equilibrium data fits for As(V) and also As(III) sorption by NIC-MO with the Langmuir model (R2 P 0.98) was comparatively bet-ter than the Freundlich model (R2 = 0.92–0.97). These indicatethat the sorption sites of NICMO surface are homogeneous andaccessible equally by tested arsenic species in the absence/pres-ence of foreign ions. Values of the dimensionless Freundlich con-stant (n) laid between 1.9 and 2.9 for As(V) while those between2.4 and 3.3 for As(III) have indicated that NICMO has greater affin-ity for the As(III) sorption compared to the As(V) from the aqueousphases in the presence of foreign ions, establishing higher degreeof strong chemical interaction of As(III) to the surface of NICMO.

Values of the monolayer sorption capacity (qm, mg g�1) for As(V)and As(III) obtained in the absence of any foreign ions are 55.513(±2.235) and 86.293 (±6.884) (Tables 3a and 3b). It has been seenthat the qm values for As(V) sorption in presence of foreign ionswere about 15–35% lower than that value obtained in absence ofbackground ions. However, the qm values for As(III) sorption inpresence of foreign ions were about 4–25% lower than in the valuein absence of background ions. Decrease of the qm values for botharsenic species is notable in the presence of PO3�

4 ; SO2�4 and

HCO�3 , indicating strong adverse influence of the said ground wateroccurring ions on the sorption of either arsenic species by NICMO.Decrease of removal efficiency of As(V) in the presence of SO2�

4 hadalready been reported by Wijnja and Schulthess [48]. The authorshad also confirmed the formation of both inner and outer-sphere

Table 3aThe isotherm parameters estimated by the non-linear model fit method of equilibrium As(V) sorption data on NICMO at pHi = 7.0 (±0.2) and T = 30 (±1) �C.

Ion + As(V) Freundlich isotherm parameters Ion + As(V) Langmuir isotherm parameters

R2 v2 KF (mg g�1)�(L mg�1)(1/n) n R2 v2 qm (mg g�1) KL (�10�2)�(L g�1)

PO3�4

0.955 4.236 2.826 ± 0.679 2.276 PO3�4

0.988 1.133 29.937 ± 1.336 3.081 ± 0.004

HCO�3 0.974 3.283 2.947‘ ± 0.569 2.159 HCO�3 0.996 0.536 35.687 ± 1.065 2.725 ± 0.002

SO2�4

0.936 14.543 4.992 ± 1.333 2.371 SO2�4

0.995 1.09 44.281 ± 1.234 4.165 ± 0.004

SiO2�3

0.924 22.518 3.773 ± 1.329 1.977 SiO2�3

0.982 5.386 54.777 ± 3.977 2.623 ± 0.005

Cl� 0925 26.749 6.028 ± 1.768 2.289 Cl 0.984 5.816 55.014 ± 2.984 4.455 ± 0.008No ion 0.946 23.455 10.295 ± 2.055 2.84 No ion 0.986 6.064 55.513 ± 2.235 9.147 ± 0.014

Table 3bIsotherm parameters estimated by the non-linear model fit method of equilibrium As(III) sorption data on NICMO at pHi = 7.0 (±0.2) and T = 30 (±1) �C.

Ion + As(III) Freundlich isotherm parameters Ion + As(III) Langmuir isotherm parameters

R2 v2 KF (mg g�1)(L mg�1)(1/n) n R2 v2 qm (mg g�1) KL (�10�2) (L g�1)

PO3�4

0.943 29.157 10.15 ± 0.846 2.469 PO3�4

0.993 3.381 64.851 ± 8.419 7.219 ± 0.032

HCO�3 0.951 35.273 17.293 ± 2.722 3.322 HCO�3 0.991 6.547 67.303 ± 2.018 17.668 ± 0.023

SO2�4

0.971 20.583 12.339 ± 1.89 2.663 SO2�4

0.98 14.541 71.347 ± 4.097 8.76 ± 0.02

SiO2�3

0.975 15.913 9.129 ± 1.559 2.373 SiO2�3

0.984 10.032 75.053 ± 4.71 4.497 ± 0.009

Cl� 0923 66.451 15.778 ± 3.481 2.445 Cl� 0.99 8.291 83.061 ± 0.967 14.141 ± 0.02No ion 0.963 28.884 9.47 ± 1.987 2.237 No ion 0.980 16.017 86.293 ± 6.884 4.268 ± 0.010

670 T. Basu et al. / Chemical Engineering Journal 228 (2013) 665–678

surface complexes of SO2�4 on the oxide surface in the presence of

As(V) using Raman and ATR-FTIR spectroscopy. The decrease in theremoval efficiency of arsenic in the presence of PO3�

4 indicated thatPO3�

4 competes well with AsO3�4 to combine with the active sites on

NICMO surface owing to their close chemical and structural simi-larity [49,50]. Despite that As(V) competitively sorbs over theoxide surface as the binding affinity constant of AsO3�

4 is only aboutseven times greater than that of PO3�

4 [49]. So the interaction ofPO3�

4 with the oxide surface is generally of the physisorption/ion-exchange type, but definitely hinders the As(V) sorption causingdecrease in removal efficiency. The strong interaction of HCO�3with the oxide surface occurs at lower pH which is related to thepH dependency of the carbonate sorption [51]. Analysis of thesorption data obtained for As(V) in the presence of bicarbonate re-vealed the formation of a bidentate surface complex as confirmedfrom ATR-FTIR spectroscopy.

Negligible decrease of the sorption capacity of either arsenicspecies in the presence of chloride ion compared to that in the ab-sence of any ion is due to the fact that the complexes of chloridewith the bimetal mixed oxide is much weaker than those betweenarsenate or arsenite and the respective sorbent [49].

Although the formation of SFe–CeH2SiO4–Na+ is quite evidentthat reduces the surface potential thereby, increasing electrostaticrepulsion between As(V) and the negatively charged surface sites[52] causing decrease in sorption capacity. But this observationprevails only when the final pH of the system was in the range9.0. Since, the experimental pH was set around 7.0 in the presentstudy, so, it can be safely concluded that As(V) removal had notbeen noticeably affected by silicate addition from external source.It has been observed that the present results were similar to otherresearchers [34,53–55]. The decrease of As(III) removal efficiencyof NICMO in the presence of PO3�

4 ; SO2�4 and HCO�3 despite poor

chemical similarity with the reduced arsenic species may be theconsequence of surface oxidation of As(III) to chemically similarAs(V) at least partly in contact of NICMO.

It is clear that arsenic removal capacity of Fe(III) oxide could beremarkably enhanced by incorporating Ce(IV), owing to enhance-ment of specific surface area (SSA) of NICMO. Besides increasingthe surface area, incorporation of Ce(IV) also increases the numberof sorption active sites on Fe(III) oxide, which contributes to theenhancement of As(III) sorption performance of NICMO. Increased

hydroxyl groups on the metal oxide surface are the active sites forincreased sorption of As(III) which agrees well with the surfacecomplex model concept of solute sorption [56].

3.3. Mean adsorption energy

Mean sorption energy (EDR, kJ mol�1) which helps to predict thenature of the reaction is defined as the free energy of transfer ofone mole solute from infinity (in solution) onto the solid surface.The mean sorption energy (EDR) is related to the Dubinin–Rad-ushkevich (D–R) constant (KDR, mol2 kJ�2) as expressed by therelation,

EDR ¼1p

2KDRð4Þ

EDR could be computed from the analysis of equilibrium data of asorption reaction using D–R isotherm Eq. (5) [57] and Polanyi po-tential (e)

ln Q e ¼ ln Q m � KDRe2 ð5Þ

where Qe and Qm are the equilibrium and saturated sorption capac-ities (mol kg�1), respectively. Polanyi potential (e) that appears inEq. (5) is actually expressed by Eq. (6).

e ¼ RT ln 1þ 1Ce

� �ð6Þ

where R is the gas constant (kJ mol�1 K�1) and T, the absolute tem-perature (K). Linearity of ln Qe against e2 plots (straight line) con-firms the applicability of D–R model isotherm. Values of Qm andKDR can be obtained from the intercept and slope of the linear lnQe versus e2 plots.

The mean free energy (EDR, kJ mol�1) of sorption can be calcu-lated by computing the value of KDR in Eq. (4). If the EDR value islesser than 8.0 kJ mol�1 or ranged between 8.0 and 16.0 kJ mol�1,then the nature of sorption reaction will be physical or ion-ex-change [57,58]. Further, if the EDR value is greater than16.0 kJ mol�1, the nature of sorption reaction will be chemisorp-tion. The plots of ln Qe versus e2 for the present equilibrium dataare presented in Fig. 3. The values of KD–R and Qm estimated forthe present sorption reactions from the slope and intercept of lin-

300 400 500 600 700 800 900 1000

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0ln

Qe

ε2200 300 400 500 600 700 800 900 1000

1.5

2.0

2.5

3.0

3.5

4.0

4.5

ln Q

e

ε2

(a) (b)Fig. 3. Dubinin–Radushkevich (D–R) isotherm plots of the (a) As(V) and (b) As (III), equilibrium sorption data on NICMO at pH 7.0 (±0.1) and at 30 (±1) �C in the presence ofsome groundwater occurring ions at background: j no ion, d SO2�

4 , N Cl�, . HCO�3 ,� PO3�4 , J SiO2�

3 .

T. Basu et al. / Chemical Engineering Journal 228 (2013) 665–678 671

ear plots of ln Qe against e2, respectively, and the corresponding EDR

values computed using Eq. (4) were summarized inTables 4a and 4b.

Good fits (R2 = 0.978–0.996) of the present experimental dataagreed well with the D–R isotherm [24]. Values of the EDR forAs(V) sorption reactions with NICMO in the absence/presence offoreign ions at background lied in the range 8.18–11.99 kJ mol�1

(Table 4a), indicating As(V) sorption reaction occurring via physicalto ion-exchange mechanism [54,59]. However, the values of EDR forAs(III) sorption by NICMO ranged in 10.30–12.35 kJ mol�1 in theabsence/presence of background ions (Table 4b), establishingAs(III) sorption occurring via ion exchange/chemisorptionphenomena.

Depending upon the above calculation, it can be concluded thatAs(III) is more favorably sorbed than As(V) by NICMO in the pres-ence of groundwater occurring ions as the estimated EDR values ofAs(III) is higher than that of As(V), respectively.

Thus, it can be concluded that As(III) uptake by NICMO is lesshindered in presence of ground water occurring ions comparedto As(V) and so, the degree of interference of those externallyadded ions is observed to be minimal.

3.4. Thermodynamic parameters

Thermodynamic parameters for As(V) and As(III) removalreactions with NICMO in the presence of the ions occurring ingroundwater have been estimated using the literature availablestandard relations (Eqs. (7)–(10)) [60] based on the assumptionthat the activity coefficient of the solutes added in solution wasunity.

DG� ¼ DH� � TDS� ð7ÞDG� ¼ �2:303RT log Kc ð8Þ

logKc ¼DS�

2:303R� DH�

2:303RTð9Þ

Kc ¼qe

Ceð10Þ

The equilibrium constant (Kc) values for the present reactionshave been calculated using the relation (Eq. (10)), where qe/Ce iscalled the sorption affinity (L g�1); qe and Ce have their usual

significance given elsewhere. Combining the Eqs. (9) and (10), itcan be written as Eq. (11)

logqe

Ce¼ DS�

2:303R� DH�

2:303RTð11Þ

According to Eq. (11), the plot of log10 (qe/Ce) versus (1/T) would bea straight line with positive intercept.

The thermodynamic constant (Kc) values have been calculatedusing Eq. (10) taking the equilibrium data (qe and Ce) from eachset of experiment for the initial As(V) and As(III) concentra-tion = 4.5 and 4.8 mg L�1 at T (±1.0) K = 283, 293, 303, 313 and323 and pHi = 7.0 (±0.1), respectively. The values of standard en-tropy change (DS�) and enthalpy change (DH�) have been calcu-lated from the slope and intercept of each straight line plotsobtained from the log (qe/Ce) versus (1/T) (shown in a and b ofFig. 4) of the linear relation (Eq. (11)). The values for DS� andDH� obtained are shown in Tables 5a and 5b. Taking the valuesof DS� and DH�, the standard Gibb’s free energy (DG�) for As(V)and As(III) removal reactions with NICMO have been calculatedin the absence/presence of the foreign ions at background(Tables 5a and 5b).

It is noticed that the arsenic removal reactions in the absence orin the presence of foreign ions at background are always endother-mic (DH� = positive). In the absence of any background ion, DH� ismuch higher for As(III) compared to that for As(V), indicatingAs(III) removal reactions with NICMO are more endothermic. Therange of DH� values evaluated for the case of As(V) (10.92–15.9) re-moval indicates the physical sorption type reaction, while that ofAs(III) (21.059–35.435) denotes more inclination towards chemi-sorption type mechanism with NICMO in the presence of the for-eign ions at background. The positive entropy change (DS�)values of either arsenic species removal reactions with NICMO inthe systems investigated conclude that the reactions are entropydriven, occurring with increase of randomness at solid–liquidinterface owing to the release of water molecules when hydratedarsenic species binds onto the solid surface. Again, the more posi-tive DS� for binding As(III) than As(V) suggests that the sorptionprocess of the former species is more entropy driven comparedto the later one. Values of the Gibb’s free energy (DG�) (Table 5)have been found to be progressively negative with increasing

Table 5aThermodynamic parameters estimated for As(V) sorption by NICMO at different reaction temperatures (K) at pHi 7.0 (±0.2) in presence of some co-occurring ions [C0 = 4.5 mg.As(V) L�1].

As(V) + Ion added +DS� (kJ mol�1 K�1) +DH� (kJ mol�1) (�DG�) (kJ mol�1)

283 K 293 K 303 K 313 K 323 K

PO3�4

0.084 10.923 12.850 13.690 14.530 15.370 16.210

HCO�3 0.101 15.067 13.613 14.626 15.640 16.653 17.667

SO2�4

0.104 15.138 14.190 15.226 16.262 17.299 18.335

SiO2�3

0.100 13.623 14.627 15.625 16.624 17.622 18.620

No ion 0.098 12.800 15.045 16.029 17.013 17.997 18.981Cl� 0.107 15.900 14.428 15.500 16.571 17.643 18.715

3.1 3.2 3.3 3.4 3.5

5.55.65.75.85.96.06.16.26.36.46.56.66.76.86.97.07.1

ln {(

q e×103 )/C

e}

1/T (*10-3)3.1 3.2 3.3 3.4 3.5

8.8

9.0

9.2

9.4

9.6

9.8

10.0

10.2

10.4

10.6

10.8

11.0

11.2

11.4

11.6

ln {(

q e×103 )/C

e}

1/T (*10-3)

(a) (b)Fig. 4. Plots of ln (qe/Ce) versus (1/T, K�1) of (a) As(V) (C0 = 4.5 mg L�1) and (b) As(III) (C0 = 4.8 mg L�1) sorption by NICMO at pH 7.0 (±0.1) and T = 30 (±1) �C in the presence ofsome groundwater occurring ions: j no ion, d SO2�

4 , N Cl�, . HCO�3 ,� PO3�4 , J SiO2�

3 .

Table 4aDubinin–Radushkevich (D–R) isotherm parameters evaluated for As(V) sorption by NICMO at pH = 7.0 (±0.2) and at T = 30 (±1) �C.

Ion + As(V) in solution D–R isotherm parameters

R2 SD EDR (kJ mol�1) �KD–R (�10�3) (mol2 kJ�2) Qm (mol kg�1)

SiO2�3

0.995 0.077 8.176 7.48 4.788

Cl� 0.995 0.070 9.221 5.88 3.406HCO�3 0.978 0.136 9.341 5.73 1.736

SO2�4

0.994 0.071 9.535 5.51 2.354

PO3�4

0.993 0.070 9.543 5.49 1.346

No ion 0.996 0.056 11.987 3.48 1.781

Table 4bDubinin–Radushkevich (D–R) isotherm parameters evaluated for As(III) sorption by NICMO at pH = 7.0 (±0.2) and at T = 30 (±1) �C.

Ion with As(III) in solution D–R isotherm parameters

R2 SD EDR (kJ mol�1) �KD–R (�10�3) (mol2 kJ�2) Qm (mol kg�1)

No ion 0.964 0.209 10.325 4.69 3.695Cl� 0.955 0.246 10.588 4.46 4.928

SiO2�3

0.988 0.111 10.403 4.62 3.143

HCO�3 0.972 0.142 12.347 3.28 2.413

PO3�4

0.968 0.185 12.004 3.47 2.113

SO2�4

0.967 0.194 11.070 4.08 3.041

672 T. Basu et al. / Chemical Engineering Journal 228 (2013) 665–678

Table 5bThermodynamic parameters estimated for As(III) sorption by NICMO at different reaction temperatures (K) at pHi 7.0 (±0.2) in presence of some co-occurring ions [C0 = 4.8 mg.As(III) L�1].

As(III) + Ion added +DS� (kJ mol�1 K�1) +DH� (kJ mol�1) (�DG�) (kJ mol�1)

283 K 293 K 303 K 313 K 323 K

PO3�4

0.148 21.059 20.826 22.306 23.786 25.266 26.746

Cl� 0.170 25.501 22.694 24.397 26.100 27.803 29.506

SiO2�3

0.169 26.114 21.810 23.504 25.197 26.891 28.584

SO2�4

0.177 28.493 21.485 23.251 25.017 26.783 28.549

HCO�3 0.175 28.771 20.725 22.474 24.223 25.972 27.721No ion 0.204 35.435 22.354 24.396 26.438 28.480 30.522

T. Basu et al. / Chemical Engineering Journal 228 (2013) 665–678 673

temperature for all the sorption reactions with NICMO in the ab-sence or presence of ground water occurring ions indicating anoverall increase of spontaneity of the reactions. However, the val-ues of DG� for As(III) species are more negative than that for As(V),predicting As(III) to be more favorably sorbed by the present mate-rial. The results reported herein have been found to be similar tothose single systems (in the absence of foreign ions) establishedearlier by various authors [24,54,61,62].

3.5. Arsenic desorption

Respective arsenic desorption percentage from arsenic-richNICMO versus employed alkali concentrations have been depictedin a and b of Fig. 5. It has been found that the efficiency of KOH washigher compared to NaOH in desorbing arsenic from either As(V) orAs(III) saturated NICMO. Again, the desorption percentage of ar-senic increased with increasing alkali concentration employed forthis purpose.

It has been observed that the 2.0 M KOH solution could desorbgreater amount of As(V) (82.69%) compared to As(III) (63.41%)from the solid surfaces. Further, increase of employed alkali con-centration increases the arsenic desorption percentage but simul-taneously dissolves out the iron component of the bimetal mixedoxide. Higher percentage of As(V) desorption in comparison toAs(III) implied that the oxidized arsenic is bound by somewhatweak forces onto the NICMO surface compared to the reducedone. This agrees well with the lower binding energy (EDR) of

0.0 0.5 1.0 1.5 2.0

51015202530354045505560657075808590

NaOH solution KOH solution

% o

f des

orpt

ion

Concentration of Alkali (M)

(a)Fig. 5. Plots of desorption percentages of (a) As(V) and (b) As(III), respectivel

As(V)-sorption than As(III) by NICMO (Tables 4a and 4b) in the ab-sence/presence of ions at background. Low desorption percentageof As(III) from AsIII-NICMO supports the strong binding affinity ofthis reduced species with solid surface.

3.6. Experimental evidences of As(V) and As(III)-sorption over sorbentsurface

3.6.1. X-ray diffraction (XRD) pattern analysis of synthetic arsenic-sorbed NICMO

XRD pattern of As(V) rich NICMO (AsV-NICMO) is shown in a ofFig. 6, whose pattern is observed to be almost similar to NICMO(pattern-c in Fig. S1A). However, the degree of crystallinitydecreases with the reduction in sharpness of the peaks in AsV-NIC-MO presumably due to the NICMO surface weakly interacting withAs(V) species.

XRD pattern of As(III) rich NICMO (AsIII-NICMO) shown in b ofFig. 6 is observed to be similar to a of Fig. 6 owing to the stronginteraction of NICMO surface with As(III) species.

After As(V) and As(III) sorption, the peak positions got shiftedso also, the crystal structure of the material causing the peakintensity to reduce to half. However, the peaks observed inAs(V) sorbed sample (AsV-NICMO, a of Fig. 6) were at 28.59�,32.88�, 48.42�, 57.63� respectively. The peak data when comparedwith standard JCPDS file, suggested that the peaks at 32.85�,48.42� and 57.63� in AsV-NICMO were due to As(V)–O phase withorthorhombic structure corresponding to reflection from (310),

0.0 0.5 1.0 1.5 2.0

10

15

20

25

30

35

40

45

50

55

60

65

% o

f des

orpt

ion

Concentration of Alkali (M)

Desorption with NaOH Desorption with KOH

(b)y, from the arsenic NICMO surface with increasing alkali concentrations.

10 20 30 40 50 60 70 800

2

4

6

8

Inte

nsity

(cps

)

2 theta10 20 30 40 50 60 70 80

0

2

4

6

8

10

Inte

nsity

(cps

)

2 THETA

(a) (b)Fig. 6. X-ray diffraction patterns of synthetic (a) As(V) (AsV-NICMO) and (b) As(III) (AsIII-NICMO), respectively, sorbed NICMO.

674 T. Basu et al. / Chemical Engineering Journal 228 (2013) 665–678

(141) and (341) planes, respectively. The peak with doubleintensity at 28.59 was due to the overlap of two peaks at exactlysame 2h angle: one arising from CeO2 phase with cubic structureand the other peak was that from As2O5 phase with orthorhombicstructure corresponding to reflection from (111) and (141)planes respectively.

Similarly, the peaks observed in As(III) sorbed sample (AsIII-NIC-MO, b of Fig. 6) were at 29.4�, 33.14�, 34.37�, 48.42�, 57.63� and78.9� respectively, which were compared with the standard dif-fraction data base of JCPDS files. Surprisingly, some unexpectedand interesting results were obtained while assigning and rational-izing the peaks. The peak at 33.14� was retained in the XRD patternof AsIII-NICMO sample which was due to a-Fe2O3 phase of Hema-tite variety having rhombohedra structures corresponding toreflection from (104) planes. The peak at 34.37� developed dueto As2O3 phase of monoclinic variety having Claudetite II structurescorresponding to reflection from (103) plane.

Contrary to our expectation, the peaks at 48.42� and57.63�developed due to As2O5 phase with orthorhombic structurecorresponding to reflection from (141) and (341) planes respec-tively. The peaks at 29.4� and 78.9� were due to the presence ofCe2O3 phase with hexagonal structure corresponding to reflectionfrom (002) and (005), respectively.

3.6.2. Spectral pattern of NICMO and As(V)/(III)-sorbed NICMO3.6.2.1. Interpretation of FTIR spectra of NICMO and AsV-NICMO: anoverview of probable mechanism. Comparing the FTIR spectra ofNICMO (spectrum-a) with AsV-NICMO (spectrum-b) (Fig. S14),the facts given hereafter can be enlightened. Broad stretching bandof O–H in spectrum-a of NICMO became exceedingly weak in spec-trum-b at wave numbers (m) 3670–3300 cm�1 of AsV-NICMO. Areaoccupied by the stretching band of O–H group in NICMO (spec-trum-a) was significantly reduced indicating disappearance ofsome O–H group from the solid surface on As(V) sorption in AsV-NICMO (spectrum-b). Two additional bands at wave number1060 and 1385 cm�1 that was due to the symmetrical and asym-metrical bending of M–OH group in spectrum-a of NICMO almostdisappeared in As(V) sorbed material (spectrum-b). Band at wavenumber 534 cm�1 of spectrum-a indicated hetero-metal bridgingof MOM’ bond that almost disappeared in spectrum-b suggestingconsiderable reduction of metal–oxygen character at the substrate

surface after As(V) sorption. This concluded that As(V) species wasweakly interacting with the surface of the bimetal mixed oxide.Appearance of band at wave number �820 cm�1 in spectrum-bcorresponds to AsV–O stretching vibration mode in AsO3�

4 whichsupported the incorporation of AsV–O bond onto the material sur-face tabulated in Table S3.

Absence of any notable difference of spectrum-a with spectra-(b to d) shown in (Fig. S15) indicated that the ground water occur-ring tested ions had no influence on As(V) sorption mechanismover NICMO. Additionally, solution pH after As(V) sorption reactionwith NICMO almost remained unchanged indicating the release ofaqua molecules onto the system.

Slight shifting and well splitting of band in FTIR spectrum sug-gested the dominance of mono protonated monodentate complex[S-OAsO2(OH)] (S stands for NICMO surface) on the oxide surface.

Based on the experimental results, the mechanism for As(V)sorption could be suggested as

SFe�Ce—OHþH2AsO�4 ! SFe�Ce—OAsðOHÞO2 þH2O

3.6.2.2. Interpretation of FTIR spectra of NICMO and AsIII-NICMO: anoverview of probable mechanism. Comparing the FTIR spectra ofNICMO (spectrum-a) with AsIII-NICMO (spectrum-b) (Fig. S16),the facts given hereafter are highlighted. Broad stretchingadsorption bands of O–H at wave number 3650–3250 cm�1 ofNICMO (spectrum-a) became somewhat weak in AsIII-NICMO(spectrum-b). Area occupied by the stretching and bendingabsorption bands of O–H group in spectrum-a of NICMO slightlyreduced (Fig. S16), indicating proportionate disappearance ofsome O–H group from the solid surface on As(III) sorption asseen from spectrum-b of AsIII-NICMO. Two additional well-de-fined sharp band at m � 1375 and 1060 cm�1 presumably dueto the symmetrical and asymmetrical bending of MOH groupin spectrum-a of NICMO that disappeared completely in As(III)-sorbed material (spectrum-b). Surprisingly, the appearance ofdistinct and well-defined band at m � 821 cm�1 in spectrum-b(Table S3) does not correspond to As(III)–O stretching vibrationmode in AsO3�

3 , which is contrary to our expectation. This anom-alous spectral behavior was probably due to the incorporation ofAs(III)–O bond in the adsorbed sample that had been partiallyoxidized to As(V)–O.

Spectrum In stats.

O Fe As Ce Total

Spectrum 3 Yes 22.20 18.33 5.91 53.55 100.00

Spectrum 4 Yes 20.96 16.00 7.40 55.64 100.00

Mean 21.58 17.17 6.66 54.60 100.00

Std. deviation 0.88 1.65 1.05 1.48

Max. 22.20 18.33 7.40 55.64

Min. 20.96 16.00 5.91 53.55

All results in weight%

(a) (b)

Fig. 7. Scanning electron microscopic (SEM) image of (i) As(III) sorbed NICMO (AsIII-NICMO) and (ii) As(III) sorbed NICMO (AsIII –NICMO) with EDAX data.

Fig. 8. Transmission electron microscopic image of As(III) sorbed NICMO (AsIII-NICMO).

T. Basu et al. / Chemical Engineering Journal 228 (2013) 665–678 675

3.7. Additional experimental evidences of As(III)-sorption over NICMOsurface

3.7.1. Scanning and transmission electron microscopic images ofAsIII-NICMO

A slight morphological change visible in the images atmicroscopic level indicated uneven distribution of large sized

agglomerates over the entire surface partially suppressing thesmall scale agglomeration (a of Fig. 7).

Demonstration of SEM image with EDAX analyzed data (b ofFig. 7) showed an irregular surface morphology of AsIII-NICMO.EDAX analyzed data showed that the mean percentages of oxygen,iron, cerium and arsenic at the marked sites of AsIII-NICMO surfacewere 21.58, 17.17, 54.6 and 6.66, respectively, indicating theempirical composition to be Fe3.4Ce4.3As1.0O15. The changes ofcomposition and morphology of As(III) sorbed NICMO had beenrevealed by comparing with that of NICMO (FeCe1.1O7.6) (a ofFig. S5), which was due to the incorporation of As(III) as (HO)3Asat pH � 7.0.

Agglomeration of the crystallite particles in NICMO was notdistinctly visible in the TEM image of As(III) sorbed NICMO(AsIII-NICMO) sample (Fig. 8). In fact, the concept of particleagglomeration was not evident here but an overall smearing of ar-senic over the sorbent surface almost covering up the pictoriallydark adsorbent sites.

3.7.2. Cyclic voltammetryCyclic voltammetry is a versatile electrochemical technique for

a quick search of redox couples present in a system indicated inFig. 9. During the forward scan starting from negative to positive(�0.5 V to +0.5 V), if the potential of the working electrode is morepositive than that of any redox couple present in the solution, then,the corresponding species in the couple is likely to be oxidized (i.e.,electrons flowing away from the electrode) and under such cir-cumstances the potential is referred as Eox. Similarly, if the reversescan of the potential (+0.5 V to �0.5 V) indicates working electrode

Fig. 9. Current (A) against potential (Volt) of (a) pure NICMO, (b) 10 min reaction ofNICMO with As(III) and (c) 2 h equilibration of NICMO with As(III) respectively.

0 500 1000 1500 20000.00

0.05

0.10

0.15

0.20

As(II

I) co

ncen

tratio

n in

the

field

sam

ple

(Ce, m

g.L-1

)

Ve

Cb

Co

Cx

Breakpoint

Exhaustion point

VbVx

Saturation point

Fig. 10. Breakthrough curve for the filtration of groundwater sample by NICMOpacked fixed-bed column.

676 T. Basu et al. / Chemical Engineering Journal 228 (2013) 665–678

potential to be more negative compared to the reduction potentialof the redox couple, then, reduction (i.e., electrons flowing awayfrom the electrode) is likely to take place and the correspondingpotential is referred as Ered.

It is quite interesting to note that the potential of pure NICMO(Eox/Ered = 0.158 V/�0.261 V) shifted to lower potential (Eox/Ered = 0.038 V/�0.156 V) after 10 min of sorption reaction withAs(III) solution. Finally, equilibration with As(III) solution after atime interval of 2 h cause further depression of redox couple po-tential to a lower value (Eox/Ered = 0.023 V/�0.141 V) creating a

Table 6Fixed-bed NICMO column filtration of high arsenic groundwater.

Flow rate (mL min�1) Co (mg L�1) Cx (mg L�1) Cb (mg L�1) Vx (�103) (g cm�2)

Observed parameters1.0 0.190 0.169 0.010 1.883

Flow rate (mL min�1) tx (�103) (min) tD (�102) (min)

Calculated parameters1.0 2.9 8.999

small potential difference of DE = 0.164 V between Eox and Ered

compared to the relative high potential difference of pure NICMOstate (DE = 0.419 V) indicating rapid, spontaneous electron transferkinetics. Thus, the more positive DE value of pure NICMO surfacerelative to the redox couple in the system accounts for the oxida-tion of surface As(III), being transferred from the aqueous phase.Higher positive fractional value of transferred electron (n) in theredox reaction estimated from the relation (DE = 59/n mV)undoubtedly added weightage to the prediction of oxidation ofAs(III) to As(V) by NICMO surface. Thus, from the above observa-tions, it can be concluded that Ce(IV) itself being a good oxidizingagent (Ce4+/Ce3+ = 1.72) in acidic medium oxidizes the surfacesorbed As(III) to As(V) (As3+/As5+ = 0.56) itself getting reduced toCe3+ according to the underlying redox reaction.

As3þ þ 2Ce4þ ¼ As5þ þ 2Ce3þ

3.8. Establishment of probable mechanism of As(III)-sorption afteranalysis of experimental evidence

The comparative less disappearance of OH group from NICMOon As(III) sorption supports two facts. The expected disappearanceof some O–H group from the solid surface on As(III) sorption withthe further introduction of O–H group into the system at the laterstages of the reaction This agreed well with the increase of pHf overpHi during sorption of As(III) over the sorbent surface.

Thus, As (III)-sorption mechanism taking place under sufficienttime lag is explained by the increase of pH, peaks in XRD, absorp-tion bands in FTIR and cyclic voltammetry analysis as described be-low, indicating S for the solid surface.

SFe�Ce—OHþ AsðOHÞ3 ! SFe�Ce—O—AsðOHÞ2 þH2OSFe�CeðCeO2Þ þ AsðIIIÞ ! SFe�CeðCeO2Þ—AsðIIIÞSFe�CeðCeO2Þ—AsðIIIÞ þH2O! SFe�CeðCeOOHÞ þ AsðVÞ þ OH�

SFe�Ce þ AsðVÞ ! SFe�Ce—AsðVÞ

3.9. Arsenic removal by NICMO fixed-bed column from groundwater

Fig. 10 demonstrates the result obtained by filtering the As(III)spiked groundwater sample (C0 = 0.19 mg L�1). To estimate therate of attainment of equilibrium between mobile and stationaryphases, the breakthrough curve obtained from the treatment offield sample was analyzed. The curve was idealized by the assump-tion that removal of the solute was complete over the initial stagesof operation. The break-point was so chosen where arsenic concen-tration (Cb) crossed over 0.01 mg L�1 in the effluent water. At anarbitrarily selected effluent concentration, Cx, closely approachingCo, the sorbent was considered to be essentially exhausted.

Based on the above theories and description [63], the experi-mental and calculated parameters for the stationary fixed-bed(depth, D = 6.0 cm, id: 1.4 cm) arsenic removal by NICMO bed werecalculated and summarized along with its observed values inTable 6. It was found that the stationary fixed-bed column of NIC-MO filtered about 2.2 L water with outlet water volume (Vb) atbreakthrough point (Cb 6 0.01 mg L�1) to be 1.3 L. The column

Vb (�103) (g cm�2) (Vx–Vb) (g cm�2) Fm (�10�1) (g min�1 cm�2) D (cm)

1.298 5.844 6.494 6.0

f (�10�1) tf (�102) (min) D (cm) % Saturation

9.372 2.789 1.899 70.336

T. Basu et al. / Chemical Engineering Journal 228 (2013) 665–678 677

saturation at exhaustion point estimated to be 70.34%. The filteredwater quality parameters analyzed (in mg L�1, except pH) at breakpoint were pH (6.9 ± 0.1), Fe2+ (0.02 ± 0.01), F� (0.07 ± 0.01), HCO�3(76 ± 5), Hardness (111.7 ± 0.3), Ca2+ (40.1 ± 0.8), Cl� (42.0 ± 1.5),TDS (218 ± 8).

3.10. Conclusion

In summary, the equilibrium data for the sorption of As(V,III) byNICMO in the absence/presence of background ions indicated thesorption sites on solid surface to be more homogeneous andequally accessible to either of the arsenic species. Range of the val-ues of standard enthalpy (DH�) indicated that the nature of As(V)sorption is more akin to physio-sorption while the nature of As(III)is more akin to chemisorptions. However, the negative values ofGibb’s free energy (DG�) confirm spontaneous nature of the reac-tions and more favorable with the increasing temperature on thesorption reactions in the presence of co-existing ions. Betterdesorption of As(V) than As(III) from arsenic rich solid surfaceestablished weak interaction of As(V) over the given oxide surface.Average desorption of adsorbed As(III) from the oxide surface andthe values of EDR obtained in the absence of background ions estab-lished the degree of strong binding affinity of As(III) to the givensorbent surface. Microscopic images prove to be quite handy inquantifying the degree of change of morphology and compositionof the sorbent surface after As(III) binding. Interpretation of exper-imental evidences of As(III) sorption over NICMO surface con-firmed the oxidation of surface adsorbed As(III) to As(V) in athermodynamically controlled sorption reaction. This study dem-onstrated that NICMO could be an effective media for the arsenicremoval from contaminated groundwater in a fixed stationarybed column.

Acknowledgements

Authors are thankful to the Head, Department of Chemistry andVice-Chancellor, Presidency University (formerly, Presidency Col-lege), Kolkata, India for providing laboratory facilities; and SaheliDe, Department of Geology for scanning microscopic analysis ofthe sample. Authors are also thankful to the Head, Department ofPhysics, Presidency University, for extensive x-ray analysis andHead, Department of Chemistry, Calcutta University for TG/DTanalysis of the sample, Prof. Nihar Ranjan Ray (Retired Professor,Saha Institute of Nuclear Physics) for AFM image. Authors wouldalso like to acknowledge Miss Saheli De for performing SEM instru-mentation. One of the authors (Tina Basu) is also grateful to theCSIR (New Delhi) for awarding Senior Research Fellowship.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.cej.2013.05.037.

References

[1] L. Jarup, Hazards of heavy metal contamination, Br. Med. Bull. 68 (2003) 167–182.

[2] M.P. Ireland, Biological Monitoring of Heavy Metals, John Wiley & Sons, NewYork, 1991.

[3] J.P. Vernet, Impact of Heavy Metals on the Environment, Elsevier, New York,1992.

[4] M.M. Karim, Arsenic in groundwater and health problems in Bangladesh,Water Res. 34 (2000) 304–310.

[5] C.H. Tseng, Y.K. Huang, Y.L. Hhung, C.J. Chung, M.H. Yang, C.J. Chen, Y.M. Hsueh,Arsenic exposure, urinary arsenic speciation, and peripheral vascular diseasein blackfoot disease hyperendemic villages in Taiwan, Toxicol. Appl.Pharmacol. 206 (2005) 299–308.

[6] R. Nickson, C. Sengupta, P. Mitra, S.N. Dave, A.K. Banerjee, A. Bhattacharya, S.Basu, N. Kakoti, N.S. Moorthy, M. Wasuja, M. Kumar, D.S. Mishra, A. Ghosh, D.P.

Vaish, A.K. Srivastava, R.M. Tripathi, S.N. Singh, R. Prasad, S. Bhattacharya, P.Deverill, Current knowledge on the distribution of arsenic in groundwater infive states of India, J. Environ. Sci. Health., Part A 42 (12) (2007) 1707–1718.

[7] B.K. Mandal, K.T. Suzuki, K. Anzai, Impact of arsenic in foodstuffs on the peopleliving in the arsenic-affected areas of West Bengal-India, J. Environ. Sci.Health., Part A 42 (12) (2007) 1741–1752.

[8] J.X. Guo, L. Hu, K. Tanabe, M. Miyatalre, C.Y. Yao, Chronic arsenic poisoning indrinking water in inner Mongolia and its associated health effects, J. Environ.Sci. Health., Part A 42 (12) (2007) 1853–1858.

[9] A. Hussam, A.K.M. Munir, A simple and effective arsenic filter based oncomposite iron matrix: development and deployment studies for groundwaterof Bangladesh, J. Environ. Sci. Health., Part A 42 (12) (2007) 1869–1878.

[10] G.C.D. Adamson, D.A. Polya, Critical pathway analysis to determine keyuncertainties in net impacts on disease burden in Bangladesh of arsenicmitigation involving the substitution of surface water for groundwaterdrinking water supplies, J. Environ. Sci. Health., Part A 42 (12) (2007) 1909–1917.

[11] N. Shibasaki, P. Lei, A. Kamata, Evaluation of deep groundwater developmentfor arsenic mitigation in Western Bangladesh, J. Environ. Sci. Health., Part A 42(12) (2007) 1919–1932.

[12] X.H. Guan, T. Su, J. Wang, Quantifying effects of pH and surface loading onarsenic adsorption on NanoActive alumina using a speciation-based model, J.Hazard. Mater. 166 (2009) 39–45.

[13] Y. Zhao, A.I. Zouboulis, K.A. Matis, Removal of molybdate and arsenate fromaqueous solutions by flotation, Sep. Sci. Technol. 31 (1996) 769–785.

[14] D. Bhattacharya, R.B. Jumawan Jr., Grieves, Separation of toxic heavy metals bysulfide precipitation, Sep. Sci. Technol. 14 (1979) 441–452.

[15] J.G. Hering, P.Y. Chen, J.A. Wilkie, M. Elimelech, Arsenic removal from drinkingwater during coagulation, J. Environ. Eng. 123 (1998) 800–807.

[16] D.A. Clifford, Ion exchange and inorganic adsorption, in: F.W. Pontius (Ed.),Water Quality and Treatment, A Handbook of Community Water Supplies,McGraw-Hill, New York, 1990, pp. 561–639.

[17] M. Bissen, F.H. Frimmel, Arsenic—a review. Part II: oxidation of arsenic and itsremoval in water treatment, Acta Hydroch. Hydrob. 31 (2003) 97–107.

[18] D. Mohan, C.U. Pittman Jr., Arsenic removal from water/wastewater usingadsorbents – a critical review, J. Hazard. Mater. 142 (2007) 1–53.

[19] J.A. Wilkie, J.G. Hering, Adsorption of arsenic onto hydrous ferric oxide: effectsof adsorbate/adsorbent ratios and co-occurring solutes, Colloid Surf. A:Physicochem. Eng. Aspects. 107 (1996) 97–110.

[20] L.C. Roberts, S.J. Hug, T. Ruettimann, A.W. Khan, M.T. Rahman, Arsenic removalwith iron(II) and iron(III) in waters with high silicate and phosphateconcentrations, Environ. Sci. Technol. 38 (2004) 307–315.

[21] B. Saha, R. Bains, F. Greenwood. Physicochemical characterization of granularferric hydroxide (GFH) for As(V) sorption from water, Sep. Sci. Technol. 40(2005) 2909–2920.

[22] K.P. Raven, A. Jain, R.H. Loeppert, Arsenite and arsenate adsorption onferrihydrite: kinetics, equilibrium, and adsorption envelopes, Environ. Sci.Technol. 32 (1998) 344–349.

[23] R.C. Vaishya, S.K. Gupta, Modelling of arsenic(III) sorption from water sulfatemodified iron oxide coated sand (SMIOCS), J. Chem. Technol. Biotechnol. 78(2002) 73–80.

[24] S. Kundu, G.P. Gupta, Investigation on the adsorption efficiency of iron oxidecoated cement (IOCC) towards As(V)- kinetics, equilibrium andthermodynamics studies, Colloids Surf. A.,: Physicochem. Eng. Aspects 273(2006) 121–128.

[25] J. Halavy, K. Polyak, Determination of surface properties of iron hydroxide-coated alumina adsorbent for removal of arsenic from drinking water, J.Colloid Interface Sci. 284 (2005) 71–77.

[26] Y. Zhang, M. Yang, X. Huang, Arsenic(V) removal with a Ce(IV)-doped ironoxide adsorbent, Chemosphere 51 (2003) 945–952.

[27] Y. Zhang, M. Yang, X.M. Don, Arsenate adsorption on an Fe-Ce bimetal oxideadsorbent: role of surface properties, Environ. Sci. Technol. 39 (2005) 7246–7253.

[28] G. Zhang, J.H. Qu, R.R. Liu, G.T. Li, Removal mechanism of As(III) by a novel Fe-Mn binary oxide adsorbent: oxidation and sorption, Environ. Sci. Technol. 41(2007) 4613–4619.

[29] E. Deschamps, V.S.T. Ciminelli, W.H. Holl, Removal of As(III) and As(V) fromwater using a natural Fe-Mn enriched sample, Water Res. 39 (2005) 5212–5220.

[30] H. Hong, W. Farooq, J. Yang, J. Yang, Preparation and evaluation of Fe-Al binaryoxide for arsenic removal: comparative study with single metal oxides, Sep.Sci. Technol. 45 (2010) 1975–1981.

[31] S. Bang, M. Patel, L. Lee, X.G. Meng, Removal of arsenic from ground water bygranular titanium dioxide adsorbent, Chemosphere 60 (2005) 389–397.

[32] M.E. Pena, G.P. Korfiatis, M. Patel, L. Lippincott, X. Meng, Adsorption of As(V)and As(III) by nanocrystalline titanium dioxide, Water Res. 39 (2005) 2327–2337.

[33] K. Banerjee, G.L. Amy, M. Prevost, S. Nour, M. Jekel, P.M. Gallagher, C.D.Blumenschein, Kinetic and thermodynamic aspects of adsorption of arseniconto granular ferric hydroxide (GFH), Water Res. 42 (2008) 3371–3378.

[34] E.A. Deliyanni, D.N. Bakoyannakis, A.I. Zouboulis, K.A. Matis, Sorption of As(V)ion by akaganite type nanocrystals, Chemosphere 50 (2003) 155–163.

[35] M.A. Hossain, M.K. Sengupta, S. Ahamad, M.M. Rahman, D. Mondal, D. Lodh, B.Das, B. Nayak, B.K. Roy, A. Mukherjee, D. Chakraborti, Ineffectivenesss and poorreliability of arsenic removal plants in West Bengal, India, Environ. Sci.Technol. 39 (11) (2005) 4300–4306.

678 T. Basu et al. / Chemical Engineering Journal 228 (2013) 665–678

[36] M.A. Hossain, A. Mukherjee, M.K. Sengupta, S. Ahamad, B. Das, B. Nayak, A. Pal,M.M. Rahman, D. Chakraborti, Million dollar arsenic removal plants in WestBengal, India, Water Qual. Res. J. Can. 41 (2006) 216–225.

[37] T. Basu, K. Gupta, U.C. Ghosh, Performances of As(V) adsorption of calcined(250 �C) synthetic iron(III)–aluminum(III) mixed oxide in the presence of somegroundwater occurring ions, Chem. Eng. J. 183 (2012) 303–314.

[38] S. Lin, D. Lu, Z. Liu, Removal of arsenic contaminants with magnetic c-Fe2O3nanoparticles, Chem. Eng. J. 211–212 (2012) 46–52.

[39] C. Han, H. Li, H. Pu, H. Yu, L. Deng, S. Huang, Y. Luo, Synthesis andcharacterization of mesoporous alumina and their performances forremoving arsenic(V), Chem. Eng. J. 217 (2013) 1–9.

[40] T. Zhang, D.D. Sun, Removal of arsenic from water using multifunctional micronano-structured MnO2 spheres and microfiltration, Chem. Eng. J. 225 (2013)271–279.

[41] Q. Feng, Z. Zhang, Y. Ma, X. He, Y. Zhao, Z. Chai, Adsorption and desorptioncharacteristics of arsenic onto ceria nanoparticles, Nanoscale Res. Lett. 7(2012) 84–92.

[42] K. Birbaum, R. Brogioli, M. Schellenberg, E. Martinoia, W.J. Stark, D. Günther,L.K. Limbach, No evidence for cerium dioxide nanoparticle translocation inmaize plants, Environ. Sci. Technol. 44 (2010) 8718–8723.

[43] K.L. Ludwig, B. Robert, M. Elisabeth, K. Rolf, G. René, J.S. Wendelin, Removal ofoxide nano-particles in a model wastewater treatment plant: influence ofagglomeration and surfactants on clearing efficiency, Environ. Sci. Technol. 42(2008) 5828–5833.

[44] J.P. Chen, S. Patil, S. Seal, J.F. Mcginnis, Rare earth nanoparticles prevent retinaldegeneration induced by intracellular peroxides, Nat. Nanotechnol. 1 (2006)142–150.

[45] N.M. Zholobak, V.K. Ivanov, A.B. Shcherbakov, A.S. Shaporev, O.S. Polezhaeva,A.Y. Baranchikov, N.Y. Spivak, Y.D. Tretyakov, UV-shielding property,photocatalytic activity and photocytotoxicity of ceria colloid solutions, J.Photochem. Photobiol., B 102 (2010) 32–38.

[46] Q. Yuan, H.H. Duan, L.L. Li, L.D. Sun, Y.W. Zhang, C.H. Yan, Controlled synthesisand assembly of ceria-based nanomaterials, J. Colloid Interface Sci. 335 (2009)151–167.

[47] S.D. Faust, O.M. Aly, Adsorption Processes for Water Treatment, ButterworthPublishers, USA, 1987.

[48] H. Wijnja, P.C. Schulthess, Vibrational spectroscopy study of selenate andSulphate adsorption mechanisms on Fe and Al (Hydr)oxide surfaces, J. ColloidInterface Sci. 229 (1) (2000) 286–297.

[49] Y. Jeong, F. Maohong, J.V. Leeuwen, J.F. Belczyk, Effect of competing solutes onarsenic(V) adsorption using iron and aluminium oxides, J. Environ. Sci. 19(2007) 910–919.

[50] E.A.El. Khatib, A.M. Balba, Arsenate chemistry in soils of arid ecosystems,Alexandria Sci. Exchange J. 25 (1) (2004) 229–261.

[51] M. Stachowicz, T. Hiemstra, W.H.V. Riemsdijk, Arsenic-Bicarbonate Interactionon Goethite Particles, Environ. Sci. Technol. 41 (2007) 5620–5625.

[52] X. Meng, S. Bang, G.P. Korfiatis, Effects of silicate, sulfate, and carbonate onarsenic removal by ferric chloride, Water Res. 34 (4) (2000) 1255–1261.

[53] P. Laxmipathiraj, B.R.V. Narshimhan, S. Prabhakar, G. Bhaskarraju, Adsorptionstudies of arsenic on manganese substituted iron oxyhydroxide, J. ColloidInterface Sci. 304 (2006) 317–322.

[54] K. Gupta, T. Basu, U.C. Ghosh, Sorption characteristics of As(V) for removalfrom water using agglomerated nanostructure iron(III)- zirconium(IV) bimetalmixed oxide, J. Chem. Eng. Data 54 (2009) 2222–2228.

[55] T.S. Anirudhan, M.R. Unnithan, As(V) removal from aqueous solutions using ananion exchanger derived from coconut coir pith and its recovery,Chemosphere 66 (2007) 60–66.

[56] R. Li, Q. Li, S. Gao, J.K. Shang, Enhanced arsenite adsorption onto Litchi-like Al-doped iron oxides, J. Am. Ceram. Soc. 94 (2011) 584–591.

[57] M.M. Dubinin, L.V. Radushkevich, The equation of the characteristic curve ofactivated charcoal, Proc. Acad. Sci. USSR Phys. Chem. Sect. 55 (1947) 331.

[58] W.J. Thomas, B. Crittenden, Adsorption Technology and Design, Butterworth-Heinemann, Oxford, 1998.

[59] W. Reiman, H. Walton, Ion-Exchange in Analytical Chemistry, InternationalSeries of Monographs in Analytical Chemistry, Pergamon Press, Oxford, 1970.38.

[60] P. Atkins, De J Paula, Atkin’s Physical Chemistry, 8th ed., Oxford UniversityPress, U.K., 2006.

[61] K. Gupta, U.C. Ghosh, Arsenic removal using hydrous nano-structure iron(III)-titanium(IV) binary mixed oxide from aqueous solution, J. Hazard. Mater. 161(2009) 884–892.

[62] K. Gupta, K. Biswas, U.C. Ghosh, Nanostructure iron(III)-zirconium(IV) binarymixed oxide: synthesis, characterization, and physicochemical aspects ofarsenic(III) sorption from the aqueous solution, Ind. Eng. Chem. Res. 47 (2008)9903–9912.

[63] J. Waber Jr., Physicochemical Processes for Water Quality Control, John Wiley& Sons, Inc., New York, 1972.