Study of factors affecting Ni2+ immobilization efficiency by temperature activated red mud

Page 1

Sr

Sa

b

c

a

ARR2A

KRNSpCD

1

osimmaqoc(wsfPTm

1d

Chemical Engineering Journal 168 (2011) 610–619

Contents lists available at ScienceDirect

Chemical Engineering Journal

journa l homepage: www.e lsev ier .com/ locate /ce j

tudy of factors affecting Ni2+ immobilization efficiency by temperature activateded mud

. Smiljanic a,∗, I. Smiciklasb, A. Peric-Grujic c, M. Sljivic b, B. Ðukic a, B. Loncarc

University of East Sarajevo, Faculty of Technology, Karakaj bb, 75400 Zvornik, Republic of Srpska, Bosnia and HerzegovinaUniversity of Belgrade, Institute of Nuclear Sciences “Vinca”, P.O.B. 522, 11000 Belgrade, SerbiaUniversity of Belgrade, Faculty of Technology and Metallurgy, Karnegijeva 4, 11000 Belgrade, Serbia

r t i c l e i n f o

rticle history:eceived 28 October 2010eceived in revised form8 December 2010ccepted 5 January 2011

eywords:ed mudi2+

orption

a b s t r a c t

The waste red mud, remaining in remarkable quantities after the digestion of bauxite ores following theBayer process, contains number of voluble minerals with excellent sorption properties towards aque-ous heavy metals. Heating at 600 ◦C was found to be a favorable treatment for revalorization of rinsedred mud into an efficient Ni2+ sorbent (RBRM600). As potential practical application of RBRM600 forNi2+ accumulation greatly depends on the solution composition and pH, the influence of these vital pro-cess variables was considered in this study. The initial pH rise from 2 to 3.5 caused the most evidentincrease in the amounts of Ni2+ removed; furthermore, the effect was more obvious for lower initial sor-bate concentrations. Conversely, changes of the solution pH between 3.5 and 8 did not have a significantinfluence on the sorption. The increase of initial cation concentration caused the increase of the sorbed

Hompeting ionsesorption

amount, following Langmuir isotherm model. The calculated maximum sorption capacity of 27.54 mg/gdemonstrated capacity increase of approximately 20% in respect to inactivated rinsed red mud. Coexist-ing cations inhibited Ni2+ removal in the following order: Cu2+ > Pb2+ ≥ Zn2+ > Cd2+ � Ca2+, whereas thepresence of Na+ and K+ did not affect the process. The investigated anions caused decrease of Ni2+ removalefficiency in the order: EDTA > chromate > acetate > sulphate, however, fluoride slightly improved sorp-

i2+ inispos

tion. Low desorption of NRBRM600 surface, thus d

. Introduction

Red mud is a waste remaining after the digestion of bauxiteres with sodium hydroxide at elevated temperature and pres-ure, following a refining method known as the Bayer process,n the production of pure alumina [1]. In its composition, red

ud can contain over 20 different minerals. The most importantineral fractions resulting from processing ore of boehmite type

re: hematite (Fe2O3), goethite (�-FeOOH), boehmite (�-AlOOH),uartz (SiO2), titanium oxide (TiO2), sodalite (Na4Al3Si3O12Cl)r cancrinite (Na6Ca1.5Al6O24(CO3)1.6), calcium aluminum sili-ate (Ca2Al2(SiO4)(OH)8), calcium silicate (CaSiO3) and gypsumCaSO4H2O). In addition to these main components, calcite (CaCO3),hewellite (CaC2O4H2O), gibbsite (Al(OH)3), perovskite (CaTiO4),

iderite (FeCO3) muscovite (K2O3Al2O36SiO22H2O), can also be

ound, whereas as trace elements V, Ga, Cr, P, Mn, Cu, Cd, Ni, Zn,b, Mg, Zr, Hf, Nb, U, Th, K, Ba, Sr, rare earth, etc., may occur [1,2].hough the chemical composition and physical properties of redud from all over the world are different and depend mainly on

∗ Corresponding author. Tel.: +387 (0)56 260 190; fax: +387 (0)56 260 190.E-mail address: [email protected] (S. Smiljanic).

385-8947/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.cej.2011.01.034

various media showed that the heavy metal cations are tightly bound toal or reuse of spent sorbent should be considered.

© 2011 Elsevier B.V. All rights reserved.

characteristics of refined ore, common characteristics of all sam-ples are: fine particles having an average particle size <10 �m, andhigh alkalinity pH (10–13). The amount of red mud per tonne of pro-duced alumina varies depending on the type of bauxite used andranges from 0.3 tonnes of high-quality bauxite, up to 2.5 tonnes forbauxite of poor quality. Every year over 66 million tonnes of thiswaste is produced around the world [3].

The heterogeneous chemical and mineral composition of redmud suggest that it represents a potential sorbent material [1,2,4].In the last decade intensive research of red mud application as sor-bent and coagulant for wastewater treatment was conducted, andfavorable results were obtained in terms of its use for the removal ofheavy metal ions [4,9], radionuclides [10], metalloids arsenic ions[11], inorganic anions as phosphate [12,13] and fluoride [14,15],and organic matter such as dyes [16,17], phenol [18] and bacteria[19]. It is important to emphasize that according to US Environmen-tal Protection Agency (EPA), red mud is not classified as hazardouswaste [20,21], and that toxicology tests show that red mud does

not show high toxicity to the environment before or after re-use[2,22].

Although it is known that Ni2+ is essential trace element forhumans and animals in small amounts, many medical studies showthat in excess amounts it can be very toxic, even cancerogenic [23].

Page 2

ineeri

Neppp

toetf

gtapbb[

iota(bbp[tttpTo

2

2

mhmdthfamaec

wtapaRyotwa

S. Smiljanic et al. / Chemical Eng

ickel is a typical representative of heavy metals released into thenvironment through various processes such as ore processing,roduction or use of nickel, nickel alloys or its compounds, finalrocessing of metals, production of fertilizers, paper, paints andrinting industry, waste incineration, and electroplating.

European Union regulations include Ni2+ among the substanceshat have harmful effects on aquatic systems [24]. US EPA rec-mmends that the level of nickel in drinking water should notxceed 0.7 mg/L [25], whereas the European Union (Council Direc-ive 98/83/EC) and the World Health Organization (WHO) standardor the level of Ni2+ in drinking water is 0.02 mg/L [26,27].

A series of unconventional sorbents of natural and waste ori-in have been studied as Ni2+ immobilization agents: differentypes of clay and zeolite [28,29], algae and biomass [30,31], humiccids [32], silica [33], and various waste products such as orangeeel [34], grape stalk [35], tea factory waste [36], activated car-on derived from almond shell [37], olive stone waste [38], animalones [39], fly ash [40], waste paper sludge [41] and red mud4,9,42].

Despite the fact that the actual wastewaters, in addition to Ni2+

ons, often contain other cations and anions, only a small numberf researches examined the impact of these species on Ni2+ sorp-ion efficiency. In our previous study, the applicability of rinsednd thermally activated red mud from the “Birac” Alumina FactoryBosnia and Herzegovina) for Ni2+ removal was confirmed [42]. Theest sorption efficiency was achieved using the sample activatedy heating at 600 ◦C. Water exclusion from gibbsite and bayeritehases leads to improvement of sorbent surface area and porosity12], as well as its solubility [43]. The aim of this paper is to examinehe influence of process parameters on the sorption properties ofemperature activated red mud sample. Nickel removal was inves-igated as a factor of the sorbate and sorbent concentration, solutionH, presence and concentration of competing cations and anions.he stability of sorbed Ni2+ was estimated using leaching solutionsf different composition and pH.

. Materials and methods

.1. Preparation and characterization of red mud sorbent

The original red mud sample, supplied from the “Birac” Alu-ina Factory (Zvornik, Bosnia and Herzegovina), was in the form of

ighly alkaline suspension (pH > 12), with the concentration of a dryatter between 250 and 350 g/L. After drying at 105 ◦C, the sample,

enoted as BRM, was grounded by a mortar and pestle. In ordero get a better insight into the processes that occur during BRMeating, thermal analysis was performed in the temperature range

rom 20 ◦C to 1000 ◦C. Thermo gravimetric and differential thermalnalyses (TGA/DTA) were carried out on a SDT Q600 (TA instru-ents) apparatus, under following conditions: flowing nitrogen

tmosphere, sample mass about 8 mg and alumina as the refer-nce material. A heating rate of 10 ◦C/min was employed both foralibration and measurement of the red mud sample.

Neutralization and activation of the original red mud sampleas performed by repeated washing with distilled water in order

o remove excess NaOH from the technological process, then dryingnd annealing the powder 3 h at 600 ◦C, according to the procedurereviously described [42]. The obtained rinsed and temperaturectivated sample was denoted as RBRM600. Chemical analysis ofBRM600 was performed using standard procedures for the anal-

sis of various oxide content. The point of the zero charge (pHPZC)f the RBRM600 sorbent was determined using batch equilibriumechnique [44]. The experiments were conducted using solutionsith different initial pH (2–11) obtained by adding minimum

mounts of variously concentrated HCl and NaOH to either distilled

ng Journal 168 (2011) 610–619 611

water or 10−3 mol/L and 10−2 mol/L NaCl. The time of equilibrationwas 24 h, and the solid/liquid ratio was 5 g/L.

2.2. Sorption experiments

The Ni2+ ions sorption onto RBRM600 was studied using batchexperiments. All experiments were carried out at room temper-ature (20 ± 1 ◦C), by agitating the suspensions on a horizontallaboratory shaker (LT2, Made in Czechoslovakia) at a constant speedof 120 rpm. As our preliminary experiments for various initial sor-bate concentrations showed that 24 h of contact was sufficientfor attaining sorption equilibrium, reaction time was kept con-stant at this level. Suspensions were equilibrated in PVC bottles,at solid/liquid ratio of 5 g/L, except in the experiments where theinfluence of sorbent mass was examined. The initial solution pHvalues were adjusted by adding minimum volumes of variouslyconcentrated HCl and NaOH solutions. After equilibration, solutionswere separated from the solid phase by centrifugation (JanetzkiK23) for 10 min at 3500 rpm and filtering through blue band fil-ter paper. Subsequently, the final pH values were measured. Tomeasure the concentration of Ni2+ in the solution, flame AtomicAbsorption Spectrometer (Philips Pye Unicam SP9), at 232.0 nmwas used. The detection limit for Ni2+ ions was 0.059 mg/L. The fol-lowing equations are used to calculate the amount of metal sorbed(mg/g) and the percentage of nickel ions removed from solution:

qe = (Ci − Ce)V

m(1)

Sorption (%) = Ci − Ce

Ci× 100 (2)

where V (L) – is solution volume, m (g) – sorbent mass, Ci (mol/L)– initial Ni2+ concentration and Ce (mol/L) – equilibrium Ni2+ con-centration in the solution.

The limited number of experiments performed at duplicate forreproducibility check, showed that at various experimental condi-tions differences between samples were less than 5%.

2.2.1. Effect of sorbent dozeThe impact of sorbent mass was studied using 2 × 10−3 mol/L

Ni(NO3)2 solution and initial pH 5. In the fixed volume of 20 mLdifferent amounts of sorbent were added in order to achievesolid/liquid ratio of 0.5, 1, 5, 10, 15 and 20 g/L.

2.2.2. Effect of pHFor the determination of the influence of pH on Ni2+ sorption,

variously concentrated solutions prepared from Ni(NO3)2 wereused: 1 × 10−4, 2 × 10−3, 5 × 10−3 and 8 × 10−3 mol/L. Initial pHvalues were adjusted in the range from ∼2 to ∼9.

2.2.3. Effect of initial cation concentrationEffect of initial Ni2+ concentration was analyzed at fixed initial

pH 5, using NiCl2 solutions of the concentrations in the range from1 × 10−4 mol/L to 8 × 10−3 mol/L.

2.2.4. The effect of competing cations on Ni2+ sorptionThe influence of coexisting cations on Ni2+ sorption efficiency

was studied depending on both the cation type and its concentra-tion. In all experiments fixed Ni2+ concentration of 2 × 10−3 mol/Lwas applied. The influence of alkali, alkaline earth and heavy met-

als was analyzed, namely: Na+, K+, Ca2+, Pb2+, Cu2+, Cd2+ and Zn2+.Solutions were prepared using nitrate salts of analytical purity.Competing cation concentrations in the working solutions haveranged from 1 × 10−4 mol/L to 1 × 10−2 mol/L, whereas all initialpH values were set to 5.

Page 3

612 S. Smiljanic et al. / Chemical Engineering Journal 168 (2011) 610–619

2

r(wm

2

2N

ftuwaTset

2

b8stweS

•

••

3

3

fe

Table 1Chemical composition of RBRM600 sample.

Component Mass fraction (%)

SiO2 13.59Fe2O3 45.86Al2O3 18.64TiO2 4.90CaO 3.54Na2O 8.40

0

2

4

6

8

10

12

Fin

al p

H

Fig. 1. TG/DTA curves of the Bosnian red mud sample.

.2.5. The effect of inorganic and organic anions on Ni2+ sorptionThe influence of the organic and inorganic anions, namely: fluo-

ide, sulphate, chromate, acetate and ethylenediaminetetraacetateEDTA) on the efficiency of nickel removal was considered. All saltsere in the Na+ form. The solutions were prepared in analogousanner as described in Section 2.2.4.

.3. Desorption experiments

.3.1. Desorption as a function of previously sorbed amount ofi2+

Solid phases obtained after Ni2+ sorption from solutions of dif-erent initial concentrations (Section 2.2.3) were dried at roomemperature and used for testing the stability of obtained prod-cts. As a leaching solution, acetic acid solution (pH 2.88 ± 0.05)as used, and it was prepared according to standard Toxicity Char-

cteristics Leaching Procedure (TCLP) [45]. Solution, denoted asCLP2, was added to achieve solid/liquid ratio of 5 g/L, and suspen-ions were then shaken under the conditions applied in sorptionxperiments. Next, the concentrations of Ni2+ ions released intohe solution were measured, as well as the final pH values.

.3.2. Desorption as a function of leaching solution compositionFirstly, RBRM600 powder was equilibrated with Ni2+ ions

y shaking the batches containing 0.1 g of solid and 20 mL of× 10−3 mol/L Ni2+ solution. This particular concentration was

elected based on the results from Section 2.2.3, as an initial concen-ration of sorbate at which the maximum saturation of the sorbentas attained. Dried solid phases were then exposed to the influ-

nce of different leaching solutions, under conditions described inection 2.3.1.

For leaching, the following solutions were used:

distilled water and acidic and alkaline solutions, with initial pH inthe range 2–11, adjusted with minimum amounts HCl and NaOH;TCLP2 solution;Ca(NO3)2 and Na2EDTA solutions at concentrations of 10−3 mol/L,5 × 10−3 mol/L and 10−2 mol/L. Initial pH values were set at 5.

. Results and discussion

.1. Sorbent characterization

The results of BRM thermal analysis are given in Fig. 1. Startingrom the room temperature, a mass loss of 1.91% up to 229 ◦C wasstimated from TG plot. This loss can be attributed to the physi-

Loss of ignition 3.28

cal water content of the red mud. The mass loss was very rapidbetween 229 ◦C and 286 ◦C (2.63%), whereas additional 4.16% werelost with the temperature increase to 907 ◦C. The total mass lossin the investigated temperature range was 8.7%. Considering datafrom DTA analysis, an endothermic peak at 40.47 ◦C can be associ-ated with water evaporation, whereas another endothermic peakat 280.57 ◦C can be related to the loss of chemically bonded watermolecules. These data can be correlated with the XRD analysis ofannealed rinsed red mud samples [42]. The increase of temperatureto 200 ◦C caused the decrease of XR-diffraction peaks intensities,characteristic for gibbsite and bayerite, while the complete loss ofthese peaks was observed in the specimens heated at ≥400 ◦C. Themass loss which occurred between 350 ◦C and 900 ◦C, can be relatedto sodalite and carbonate decomposition [42].

Chemical analysis showed that the main component of the acti-vated sample was Fe2O3, followed by Al2O3 and SiO2 (Table 1).When compared to the composition of rinsed Bosnian red mud [42],the biggest change induced by annealing at 600 ◦C, was reflectedthrough decreased loss of ignition (7.03% vs. 3.28%).

The pH values, obtained after equilibrating RBRM600 withsolutions of different initial pH, with and without backgroundelectrolyte (NaCl), are presented in Fig. 2. The pHPZC value ofapproximately 9.5 can be estimated from the plateau of the graphs.This value is higher in respect to pHPZC of rinsed red mud [42]. Thesame trend was observed considering pH values of investigatedmaterials, since due to the loss of the major quantity of releasableconstituents and higher specific surface area, higher solubility ofthe sample RBRM600 can be expected.

121086420

Initial pH

Fig. 2. The relationships between the initial and the pH values after equilibrationof RBRM600 with distilled water (�), 10−3 mol/L NaCl (�) and 10−2 mol/L NaCl (�).

Page 4

S. Smiljanic et al. / Chemical Engineering Journal 168 (2011) 610–619 613

201510500

20

40

60

80

100

Rem

oval

effi

cien

cy (

%)

5

6

7

8

9

10

Final pH

Fo

3

dl

cic1dRnpwmiasstrtaibi

tw

3

rirstt(tifi

10987654321

0

20

40

60

80

100

Rem

oval

effi

cien

cy (

%)

Initial pH

a

10987654321

4

5

6

7

8

9

Fin

al p

H

Initial pH

b

therefore precipitation of insoluble Ni(OH) can be expected as the

Sorbent doze (g/L)

ig. 3. Nickel removal efficiency (�) and final solution pH values (�) as a functionf the RBRM600 doze. Initial Ni2+ concentration 2 × 10−3 mol/L.

.2. Effect of sorbent dose

For determining the influence of sorbent dose on Ni2+

econtamination process, different masses of RBRM600 were equi-ibrated with the solution of constant sorbate concentration (Fig. 3).

An increase of the sorbent dose in the range 0.5 g/L to 10 g/Laused a great influence on the sorption process, and for the appliednitial nickel concentration (2 × 10−3 mol/L) the percentages ofation removed from the solution increased rapidly from 16.6% to00%. Further increase of sorbent mass up to 20 g/L was irrelevant,ue to nickel removal below the detection limit. The increase ofBRM600 amount added into the solution increased not only theumber of binding sites available for sorption, but also the solutionH (Fig. 3), both of which increased the process efficiency. However,ith sorbent dose increase, the amounts of nickel sorbed per unitass of RBRM600 decreased from 34.8 mg/g to 10.48 mg/g, show-

ng that enhanced utilization of a sorbent capacity was actuallychieved at smaller doses. As the RBRM600 represent a mixture ofeveral oxides and other minerals, it exhibits highly heterogeneousurface in terms of binding sites energies. At low sorbent doses, allypes of sites are entirely exposed for sorption, and thus higheremoval capacities are observed [46]. In addition, particle aggrega-ion at higher doses may cause a decrease in the total surface area ofsorbent and lower sorption capacity per unit mass. Consequently,

n each case of potential RBRM600 application, the appropriate sor-ent amount has to be estimated based on initial Ni2+ concentration

n contaminated effluent and its required reduction.In order to assure detectable concentrations of Ni2+ cations in

he solution after the contact with RBRM600, a sorbent dose of 5 g/Las applied in further experiments.

.3. Effect of solution pH

The influence of initial solution pH on the efficiency of nickelemoval is presented in Fig. 4a. Regardless of the solution pH, withncreasing initial sorbate concentration, the percentage of nickelemoved declined. The most evident increase in the amount of Ni2+

orbed was observed with the initial pH rise from 2 to 3.5. Fur-hermore, the lower the initial sorbate concentration, the higherhe influence of pH: for the lowest investigated Ni2+ concentration

10−4 mol/L) the removal efficiency increased radically from 11%o 100%, whereas for the highest concentration (8 × 10−3 mol/L)ncrease from 0.5% to 22% was detected. In the mentioned pH range,nal pH values were <8, and given that Ni2+ cations are the most

Fig. 4. Nickel removal efficiency (a) and final pH values (b) as a function of initialsolution pH. Initial Ni2+ concentrations (mol/L): (�) 10−4, (�) 2 × 10−3, (�) 5 × 10−3

and (�) 8 × 10−3.

dominant ionic species in the solution [42], sorption of Ni2+ onRBRM600 active sites was the most dominant mechanism of nickelremoval.

The increase of the initial pH values between 3.5 and 8, didnot cause significant changes in the amount of Ni2+ removed fromsolutions of 10−4 mol/L, 2 × 10−3 mol/L and 5 × 10−3 mol/L, and for8 × 10−3 mol/L, a slight increase from 22% to 30% was observed asthe pH increased from 3.5 to 5. These observations are of practi-cal importance, signifying that due to RBRM600 buffering capacitythere was a wide initial pH range where Ni2+ removal was nota function of initial pH. On the other hand, removal efficiencychanges were in the good correlation with final pH changes (Fig. 4b).Changes of the final pH in the initial pH 2–3.5, are the most promi-nent for the lowest Ni2+ concentration applied, in accordance withNi2+ removal profile. The pH at the plateau of the pHinitial vs.pHfinal plots decreased from ∼8.5 to ∼5 with the increase of sor-bates concentration, meaning that in this buffering pH range, themechanism of nickel removal was a function of its initial concen-tration. For 10−4 mol/L, final pH values at the plateau were >8,

2preferred mode of nickel immobilization. Decrease of final pH incomparison to pHPZC is an indicator of H+ ions release, i.e. the sur-face complexation mechanism. At pH < pHPZC, surface active groupsare protonated, neutral or positively charged, therefore possible

Page 5

614 S. Smiljanic et al. / Chemical Engineering Journal 168 (2011) 610–619

0.0080.0060.0040.0020.0004

5

6

7

8

9

10

Initial cation concentration (mol/dm3)

Fin

al p

H

a

30

40

50

60

70

80

90

100

110

Rem

oval efficiency (%)

300250200150100500

0

5

10

15

20

25

30

300250200150100500

0

2

4

6

8

10

Ce/

Qe

(g/L

)

Ce (mg/L)

Qe

(mg/

g)

mg/L

b

F ial sorr l).

rl

≡2

≡≡Ffim

3

ec

Ce (

ig. 5. (a) Ni2+ removal efficiency (�) and final pH values (�) as a function of initepresents linear fitting of isotherm sorption data using Langmuir theoretical mode

eactions on various oxide surfaces (≡S) can be described as fol-ows:

S–OH + Ni2+ ⇔ ≡S–ONi+ + H+ (3)

(≡S–OH) + Ni2+ ⇔ (≡S–O)2Ni + 2H+ (4)

S–OH + NiOH+ ⇔ ≡S–ONiOH + H+ (5)

SOH2+ + Ni2+ ⇔ ≡S–ONi+ + 2H+ (6)

inally, enhanced removal of nickel observed at initial pH > 8or all investigated sorbate concentrations, can be attributed toncreased role of Ni(OH)2 precipitation in the overall sorption

echanism.

.4. Effect of initial cation concentration

The influence of the initial sorbate concentration on processfficiency and final pH values is presented in Fig. 5a. In the con-entration range 10−4 mol/L to 10−3 mol/L, Ni2+ immobilization by

)

bate concentration; (b) the isotherm of Ni2+ sorption on RBRM600 (small picture

RBRM600 was complete. With the further concentration increaseup to 8 × 10−3 mol/L, the percentages of sorbed cation decreasedgradually to 34%. Furthermore, it can be observed that equilibriumsolution pH values decreased with the rise of initial concentration,from 8.78 to 5.08.

Fig. 5b represents Ni2+ sorption isotherm on activated red mud,i.e. relationship between equilibrium concentrations of Ni2+ in thesolution and in the solid phase. Maximum sorption capacity wascalculated by fitting the experimental data using linear form ofLangmuir isotherm model [47]:

Ce

Qe= 1

XmKL+ Ce

Xm(7)

where Ce (mg/L) and Qe (mg/g) represent the equilibrium Ni2+ con-

centrations in solution and in solid phase, respectively; Xm (mg/g)is the maximum sorption capacity of RBRM600, whereas KL (L/mg)represents the Langmuir constant, associated to sorption affinity.The results presented in Fig. 5b, show a good correlation betweenthe model and experimental points, with the high regression coef-

Page 6

ineering Journal 168 (2011) 610–619 615

fic

r[i

Noaco

tg([v3C(wz7

3

oiTnpfGbtabtv

mCwocwTw(cRoNdrbwehaetT

0.0100.0080.0060.0040.0020.0000

10

20

30

40

50

60

70

80

90

Rem

oval

effi

cien

cy (

%)

Competing cation concentration (mol/L)

Na+

K+

Ca2+

Pb2+

Cu2+

Cd2+

Zn2+

a

0.0100.0080.0060.0040.0020.000

4.5

5.0

5.5

6.0

6.5

7.0

Fin

al p

H

Competing cation concentration (mol/L)

Na+

K+

Ca2+

Pb2+

Cu2+

Cd2+

Zn2+

b

S. Smiljanic et al. / Chemical Eng

cient (R2 = 0.990). Calculated value of maximum Ni2+ sorptionapacity was Xm = 27.54 mg/g and KL = 0.116 L/mg.

Under the same set of experimental conditions, the capacity ofinsed Bosnian red mud (RBRM) was determined to be 21.8 mg/g42], thus this study shows that annealing of the powder at 600 ◦Cncreased sorption capacity for approximately 20%.

It is important to notice (Fig. 5a), that the final pH values at loweri2+ concentrations indicate possible precipitation of Ni(OH)2. Thebserved cation sorption at higher initial concentrations took placet pH < 8, and also at pH < pHPZC of sorbent, confirming that the spe-ific sorption by the oxide surfaces was the dominant mechanismf sorptive uptake.

Sorbents derived from red mud have higher sorption capacitiesowards Ni2+ cations in comparison to a range of other investi-ated materials: kaolinite (10.4 mg/g) and acid-activated kaolinite11.9 mg/g) [28]; hydrous silica (0.09 mg/g or 1.68 × 10−6 mol/g)33]; natural clinoptilolite (3.8 mg/g) [48]; fly ash (5.9 mg/g), acti-ated carbon (5.4 mg/g) [40]; olive stone waste (2.13 mg/g or.63 × 10−5 mol/g) [38]; tea factory waste (15.26 mg/g) [36], etc.omparable capacities have been reported for montmorillonite28.4 mg/g) and acid-activated montmorillonite (29.5 mg/g) [28],hereas higher capacities for composite of activated carbon and

eolite 70.43 mg/g (1.20 mmol/g) [49] and paper sludge fired at00 ◦C (230.65 mg/g or 3.93 mmol/g) [41].

.5. Effect of competing cations on Ni2+ removal efficiency

To assess the potential applicability of RBRM600 for the removalf Ni2+ from aqueous solutions, sorption efficiency was also stud-ed as a function of competing ion presence and concentration.he choice of cations has been made based on their presence inatural waters, or industrial wastewaters. To facilitate the com-arison of the results, the data were obtained using nitrate saltsor cation evaluation and all initial pH values were adjusted to 5.enerally, heavy metal cations can be removed from the solutiony red mud via different sorption mechanisms including precipita-ion, specific sorption (chemisorption mainly on Fe- and Al-oxides)nd ion-exchange [1,5]. The results of our study can be interpretedy the competition of different positive ions with Ni2+ for interac-ion with the RBRM600 surface. The influences of Na+, K+, Ca2+ andarious divalent heavy metal cations are presented in Fig. 6a.

Considering all investigated competing cation concentrations,onovalent metals exhibited negligible effect on Ni2+ sorption.

onversely, a certain competition with Ca2+ was observed, whichas more obvious at its higher initial concentrations. Low ratio

f electric charge to the radius of an ion (ionic potential), that isharacteristic for alkaline metal cations, generally results in theireak electrostatic attraction to the surface ions of opposite charge.

hus, these elements most likely form outer-sphere complexesith RBRM600 or remain in the solution. The higher valence ion

Ca2+) was somewhat easily and strongly bonded. The additionalonfirmation of the existence of stronger bond between Ca2+ andBRM600 was a certain final pH drop (Fig. 6b) suggesting formationf inner-sphere complexes. Contrary to that, final pH values afteri2+ sorption from Na+ and K+ containing solutions were indepen-ent of both cation type and its concentration in the investigatedange. Similar observations were published for Ni2+ sorption onentonite, from the solutions containing Na+, K+ and Ca2+ [50],here removal was found to be the highest in Na+ and the low-

st in Ca2+ containing solutions. Sorption of some other divalent

eavy metal cations was also more influenced by the presence oflkaline earth than alkali metal cations [51]. Furthermore, the pres-nce of Ca2+ ions in the solution was found to significantly enhancehe sorption capacity of red mud in respect to arsenate anions [11].he positive influence of Ca2+ was explained by linking the sor-

Fig. 6. The influence of competing cations on Ni2+ sorption efficiency (a) and finalpH values (b). Initial Ni2+ concentration 2 × 10−3 mol/L.

bent particles with arsenate, forming a metal–arsenate complex ora metal–H2O–arsenate complex, which is another indication of redmud–Ca2+ bonds.

The competing effect of divalent heavy metal cations on Ni2+

removal by RBRM600 was more obvious in comparison with alkaliand alkaline earth metals (Fig. 6a). Regardless of the cation type, theNi2+ sorption efficiency decreased with the increase of competingcation concentration. At lower initial cation concentrations (up tothe value equal to initial Ni2+ concentration) investigated cationswere sorbed by RBRM600 following decreasing order of affin-ity: Cu2+ > Zn2+ > Pb2+ ≥ Cd2+, whereas order of affinity at higherinitial concentrations was: Cu2+ > Pb2+ ≥ Zn2+ > Cd2+. Ni2+ removaldecreased to 18%, 11%, 9% and 7% in the most concentrated(10−2 mol/L) solutions of Cd2+, Zn2+, Pb2+ and Cu2+, respectively.As coexisting cation, Cu2+ produced the most pronounced nega-tive effect on Ni2+ sorption, moreover this effect was obvious evenwhen initial Cu2+ concentrations were significantly below initialNi2+ concentration (<2 × 10−3 mol/L).

The studies of sorption affinity of different red mud sam-ples towards divalent heavy metals have shown that selectivitysequences fluctuate, for example: Cu2+ > Zn2+ > Cd2+ ≥ Ni2+ [4],Cu2+ > Cd2+ > Pb2+ [5], Zn2+ ≥ Pb2+ > Cd2+ [6]. In general, Cu2+ ions

were preferably sorbed from single metal solutions. When mixedsolution containing Cu2+, Zn2+ and Cd2+ were used, Cu2+ was foundto be the most competitive as well [52]. This is in agreement withthe results of our study. Taking into consideration heavy metal

Page 7

6 ineering Journal 168 (2011) 610–619

ctsoafese

idtRhfptdtgtrpaCptsoic

3s

ratatpTp

≡

Tadoinc

≡

≡

csc

et

0.0100.0080.0060.0040.0020.0000

20

40

60

80

100

Rem

oval

effi

cien

cy (

%)

Competing anion concentration (mol/L)

Fluoride Sulphate Acetate Chromate EDTA

a

0.0100.0080.0060.0040.0020.000

5.0

5.5

6.0

6.5

7.0

7.5

Fin

al p

H

Competing anion concentration (mol/L)

Fluoride Sulphate Acetate Chromate EDTA

b

16 S. Smiljanic et al. / Chemical Eng

ations other than Cu2+, observed differences in published selec-ivity orders are probably the results of different compositions oftarting sorbent materials. As red mud represents a mixture of vari-us minerals, sorption sequence obtained for a particular sample iscombination of separate sorption sequences characteristic for dif-

erent constituents. In addition, different experimental conditions,specially in terms of cation concentration and pH, may affect theorption mechanism of investigated cations and thus their removalfficiency.

Although the nickel sorption efficiency decreased with thencreasing concentration of heavy metal cations, the final pH valuesecreased (Fig. 6b). This indicates the specific sorption of competi-ive cations. The solution pH values were lower then the calculatedBRM600 point of zero charge, thus the competitive sorption ofeavy metals occurred on positively charged surfaces through the

ormation of inner-sphere complexes. The decrease of equilibriumH values followed the order: Pb2+ > Cu2+ > Cd2+ > Zn2+. The quan-ity of specifically sorbed ions on the surface of solids [53] is stronglyependent on the characteristics such as electric charge, radii ofhe hydrated ions and metal electronegativity. Since all the investi-ated heavy metals were of the same valence, it can be assumedhat Pb2+ form inner sphere complexes on oxide surfaces moreeadily due to the smaller hydrated radius of Pb2+ (0.401 nm) com-ared to Cu2+ (0.419 nm), Cd2+ (0.426 nm) and Zn2+ (0.430 nm);nd the higher electronegativity of Pb2+ (2.33) than Cu2+ (1.90),d2+ (1.69) and Zn2+ (1.65). However, different sequences of finalH decrease and competing cation influence on Ni2+ immobiliza-ion, suggest that the specific sorption was not the only operatingorption mechanism for all investigated cations. The contributionf surface precipitation may have attributed to the strong compet-ng effect of Cu2+ ions for RBRM600 surface, especially at low initialoncentrations, where final pH values are >6 (Fig. 6b).

.6. Effect of inorganic and organic anions (ligands) on Ni2+

orption

The effect of organic and inorganic anions (ligands) on the Ni2+

emoval efficiency by RBRM600 was considered in the same oper-tive conditions. Sodium salts were used, since Na+ did not disturbhe sorption of Ni2+ by investigated sorbent. Anions in solution canffect the cation sorption in several ways [54,55]. They can reducehe cation uptake by formation of stable aqueous metal-anion com-lexes, or by competition with cations for available surface sites.he ligands (Ln−) can be sorbed by non-specific outer sphere com-lex formation, according to the following general equation:

S–OH2+ + Ln− ⇔ ≡S–OH2–L(n−1)− (8)

he creation of such ion pairs is dominated by the electrostaticttractive forces between surface and the sorbate, and it largelyepends on the ionic strength of the surrounding solution. On thether hand, specific sorption involves direct coordination of sorb-ng species to the surface metal ions. These bonds are stronger inature, thus specific sorption may take place both on positivelyharged and the neutral surface:

S–OH2+ + Ln− ⇔ ≡SL(n−1)− + H2O (9)

S–OH + Ln− ⇔ ≡SL(n−1)− + OH− (10)

Alternatively, the co-adsorption of anions may enhance theation sorption on mineral surfaces through enhanced electro-

tatic interactions, formation of a ternary cation-anion-surfaceomplexes or surface precipitates.

Important factors in determining whether metal sorption isnhanced or inhibited include cation/ligand ratio, nature of bothhe ligand and the cation and surface properties of the sorbent [56].

Fig. 7. The influence of organic and inorganic anions on Ni2+ sorption efficiency (a)and final pH values (b). Initial Ni2+ concentration 2 × 10−3 mol/L.

At investigated pH 5, RBRM600 surface was positively charged;investigated cation was in the divalent form, while anions were pre-dominantly in the forms of F−, SO4

2−, CH3COO−, HCrO4−/Cr2O7

2−

and H2EDTA2−. The influence of an anion on the Ni2+ sorptiondepends on the anions relative affinity for the solid surface or theNi2+ cations, as well as on the relative concentrations of the anionstudied.

As it can be observed from Fig. 7a, the effect of anions decreasedin the following way: EDTA > chromate > acetate > sulphate >fluoride. The presence of fluoride did not change considerably theNi2+ removal efficiency, in fact a slight increase in sorption wasobserved.

Previously, it was confirmed that fluoride anions can be effi-ciently removed from aqueous solutions by red mud [15], byspecific sorption on metal oxides according to the Eq. (9). The max-imum F− removal was obtained at pH 4.7, whereas capacity of8.921 mg/g was estimated from the sorption isotherm. Our resultsshow that Ni2+ sorption was not inhibited by F−, which means thatfluoride neither form sufficiently stable complex with aqueous Ni2+,nor exhibit sorption affinity for RBRM600 surface higher than Ni2+.Reduction of the positive RBRM600 surface charge after sorption ofnegative anions such as F− may result in a more attractive surfacefor Ni2+ sorption. The similar weak effect was observed with coex-isting sulphate anions. This is in agreement with the recent study onheavy metal and sulphate removal from synthetic acid mine water,using red mud [57].

High affinity of red mud for heavy metals was demonstratedthrough almost complete removal of Fe3+ (99%) and Cu2+ (97%)and a significant removal of Zn2+ (84%), whereas only about 27% ofsulphate reduction was attained. Furthermore, sulphate ions were

Page 8

ineering Journal 168 (2011) 610–619 617

fn

iWaimisths1tw[

bblacwut

N

TR

t

aitiono

3

wbNis

2toradtdoirsb

so

302826242220181614121086420-20

2

4

6

8

10

12

14

16

18

2+

Des

orbe

d am

ount

of N

i2+ (

mg/

L)

5

10

15

20

25

30

35

Desorption efficiency (%

)

after EDTA treatment (6.2–6.6), Ni2+ concentrations in the solu-tion increased significantly from nearly 2% to approximately 25%with the increasing EDTA concentration, confirming the stability ofaqueous Ni–EDTA complex.

Table 2Ni2+ desorption efficiency and final pH values as a function of leaching solutioncomposition.

Leaching solution Desorbed amount (mg/L) Desorption efficiency (%)

Water solution, pH 2 10.75 7.96Water solution, pH 3 0.3 0.22Water solution, pH 4 0.07 0.05Water solution, pH 5 0 0Water solution, pH 6 0 0Water solution, pH 7 0 0Water solution, pH 8 0 0Water solution, pH 9 0 0Water solution, pH 10 0 0Water solution, pH 11 0 0TCLP2 17.5 12.96Ca2+, 10−3 mol/L 0.30 0.22

S. Smiljanic et al. / Chemical Eng

ound to be sorbed essentially as other sphere complexes on mag-etite [58].

The strong competing effect of Cr(VI) can be explained byts relatively high affinity towards the red mud surface [7,8].

hile early works considered chromate sorption on Fe oxidess an other sphere complexation process, later studies, involv-ng different spectroscopic techniques, as well as electrophoresis

easurements, support the inner sphere mechanism [55]. It isnteresting to compare the competing effect of Cr(VI), which attudied pH exists in the form of HCrO4

− and Cr2O72− anions, with

he effect of other investigated heavy metal cations (Fig. 6a). At theighest studied concentration (10−2 mol/L), Cr(VI) decreased Ni2+

orption efficiency to 57%, whereas a decrease to 9%, 7%, 18% and1% was observed in Pb2+, Cu2+, Cd2+ and Zn2+ solutions, respec-ively. The similar was reported for Pb2+ sorption by red mud, whichas also more reduced in the presence of Cd2+ and Zn2+, than Cr(VI)

7].Sorption of organic ligands on Fe oxides is usually dominated

y electrostatic effects, although ligand exchange and hydrogenonding may also be involved [55]. Among investigated organic

igands, as a monodentate ligand and a mild complexing agent,cetate exhibited weaker effect on Ni2+ sorption and reducedation removal efficiency to 42% at its highest concentration. EDTAas found to considerably decrease metal sorption, reaching totalptake inhibition at equimolar and higher Ni2+/ligand concentra-ions. At pH 5, EDTA forms stable, negatively charged complex:

i2+ + H2EDTA2− ⇔ [NiEDTA]2− + 2H+ (11)

he obtained complex has a lower sorption affinity for theBRM600 surface, and thus suppresses Ni2+ removal.

The similar effect of EDTA was reported for divalent metal sorp-ion onto other inorganic sorbent materials [59].

The changes of solution pH values as a function of competingnion type and concentration are presented in Fig. 7b. Compar-ng to pH changes, observed in the cases of cations co-sorption,he final pH values either increased (in respect to initial pH 5) withncreased anion concentration, or remained relatively constant. Thebserved complex final pH profiles represent the result of a combi-ation of processes involving both RBRM600 surface and formationf various complex ionic species in solution.

.7. Desorption studies

Reversibility of heavy metal sorption is an important factorhich determines the future management of the spent sor-

ent material. The chemical stability of the derived, stabilizedi/RBRM600 products was studied under different conditions and

t was found that Ni2+ ions release depends on both the previouslyorbed amount and the composition of the surrounding solution.

The acidic TCLP2 solution prepared from acetic acid (pH.93 ± 0.05) is generally used in TCLP method as it simulates condi-ions that exist at municipal waste landfills (low pH and presencef mild complexing agents). This solution was applied to solidesidues containing various amounts of sorbed Ni2+, and the resultsre presented in Fig. 8. Although absolute amounts of metalsesorbed into the extracting solutions increased from 1.5 mg/Lo 16.5 mg/L, with increasing sorption, their relative amountsecreased from approximately 35 to 10%. Decrease in metal des-rption with increasing sorbate amount may be connected withncreased proportion of Ni2+ in solid fractions from which it is lesseadily desorbed. In addition, final pH values were relatively con-

tant (∼4) and by 1 pH unit higher than initial pH, due to sorbentsuffering properties.

From Table 2, it is evident that Ni2+ desorption from RBRM600urface strongly depend on the use of extracting solutions, if previ-usly sorbed amounts of Ni2+ were constant. Seventeen extracting

Sorbed amount of Ni (mg/g)

Fig. 8. The amounts of Ni2+ released into TCLP2 solution (�) and desorption effi-ciency (�), as a function of previously sorbed amount of Ni2+.

solutions were applied to estimate metal leachability as a func-tion of pH, presence of competing cations and a complexing agent.The desorbed amounts decreased continuously with increasing pHin the range 2–4, whereas no detectable amounts were desorbedabove pH 5 (final pH > 6.6). It is interesting to notice that activatedred mud, fully loaded with Ni2+ cations, buffers water solution ofthe initial pH in the range from ∼4 to ∼10 to 6.6 ± 0.2 (Fig. 9). Thespecific sorption of Ni2+ shifted the pHPZC by nearly 3 pH units incomparison to the RBRM600 value before addition of nickel (Fig. 2).

The amount of heavy metal cations extracted by TCLP2 solutionwas several times higher than that leached in HCl solution hav-ing the same pH. This is in agreement with the inhibiting effect ofacetate anion observed in Section 3.6. In addition, the final solutionpH was lower after TCLP treatment (3.95) then after treatment withHCl solution of initial pH 3 (4.87), which also explains enhanceddesorption.

In accordance with the weak effect on Ni2+ sorption, Ca2+ cationscaused maximum Ni2+ desorption of only 1%, when applied at high-est studied concentration (10−2 mol/L). The final pH values werein the range 5.2–5.4. Although the final pH values were higher

Ca2+, 5 × 10−3 mol/L 0.97 0.72Ca2+, 10−2 mol/L 1.36 1.01EDTA, 10−3 mol/L 2.66 1.97EDTA, 5 × 10−3 mol/L 9.50 7.04EDTA, 10−2 mol/L 34.16 25.30

Page 9

618 S. Smiljanic et al. / Chemical Engineeri

121086423

4

5

6

7

8

9

10

Fin

al p

H

Fa

wua

mnFePp

4

ccatdoao2acsuwrtriRoiesps12c

[

[

[

[

[

[

[

[

[

[

[

[[

[

[

[

[

Initial pH

ig. 9. The relationships between initial pH values of water solutions and pH valuesfter equilibration with fully Ni2+-loaded RBRM600.

The obtained results of Ni2+ desorption are in a good correlationith the data published for Pb2+, Cd2+ and Zn2+ desorption fromntreated and acid activated red mud samples by H2O, Ca(NO3)2nd EDTA treatments [6].

Desorption studies in various media thus show that the heavyetal cations are tightly bound to red mud surface and would

ot be expected to be released readily under natural conditions.rom the technological point of view this also means that regen-ration of red mud is, if not impossible, at least not economical.roper disposal or reuse of spent red mud sorbent should beroposed.

. Conclusions

The influence of process variables (sorbent and sorbate con-entration, initial pH as well as presence and concentration ofoexisting cations and anions) on Ni2+ sorption by temperaturectivated red mud particles was studied and discussed. Using ini-ial cation concentration of 2 × 10−3 mol/L an increase of sorbentose in the range from 0.5 g/L to 10 g/L caused the rapid increasef sorbed amount from 16.6% to 100%. Regardless of the initialqueous concentration, the most obvious increase in the amountf Ni2+ sorbed was detected with the initial pH increase fromto 3.5, while initial pH rise from 3.5 to 8 did not consider-

bly influence the process. With the increase of initial sorbateoncentration the sorbed amount increased, while equilibriumolution pH decreased. The maximum sorption capacity, calculatedsing linear form of Langmuir model, was found to be 27.54 mg/g,hich is about 20% higher in respect to capacity of inactivated

insed red mud. Presence of Na+ and K+ at various concentra-ions in the studied range 10−4 to 10−2 mol/L did not affect Ni2+

emoval efficiency, Ca2+ induced a slight decrease, while coexist-ng divalent heavy metals significantly reduced Ni2+ uptake byBRM600 in the order: Cu2+ > Pb2+ ≥ Zn2+ > Cd2+. The presence ofrganic and inorganic anions decreased Ni2+ removal efficiencyn the order EDTA > chromate > acetate > sulphate, while fluoridexhibited small but detectable positive effect on Ni2+ removal. Thetudies of metal leachability have revealed dependence on both thereviously sorbed amount and the composition of the surrounding

olution. The maximum desorbed amount (25.3%) was achieved in0−2 mol/L EDTA, 12.9% in TCLP2 and 7.9% in HCl solution at pH, whereas in neutral, alkaline conditions, as well as in variouslyoncentrated Ca2+ solutions desorption was negligible. As a conse-

[

[

ng Journal 168 (2011) 610–619

quence of the high chemical stability of sorbed Ni2+, proper disposalor reuse of spent red mud sorbent should be anticipated.

Acknowledgement

This work was financially supported by the Ministry of Scienceand Technological Development of Serbia, under Project No. 43009.

References

[1] R.K. Paramguru, P.C. Rath, V.N. Misra, Trends in red mud utilization—a review,Miner. Process. Extract. Metall. Rev. 26 (2005) 1–29.

[2] S. Wang, H.M. Ang, M.O. Tadé, Novel applications of red mud as coagulant,adsorbent and catalyst for environmentally benign processes, Chemosphere72 (2008) 1621–1635.

[3] S Agatzini-Leonardou, P. Oustadakis, P.E. Tsakiridis, Ch. Markopoulos, Titaniumleaching from red mud by diluted sulfuric acid at atmospheric pressure, J.Hazard. Mater. 157 (2008) 579–586.

[4] E. López, B. Soto, M. Arias, A. Núnez, D. Rubinos, M.T. Barral, Adsorbent prop-erties of red mud and its use for wastewater treatment, Water Res. 32 (1998)1314–1322.

[5] R. Apak, E. Tutem, M. Hugul, J. Hizal, Heavy metal cation retention by uncon-ventional sorbents (red mud and fly ashes), Water Res. 32 (1998) 430–440.

[6] L. Santona, P. Castaldi, P. Melis, Evaluation of the interaction mechanismsbetween red muds and heavy metals, J. Hazard. Mater. B136 (2006) 324–329.

[7] V.K. Gupta, M. Gupta, S. Sharma, Process development for the removal of leadand chromium from aqueous solutions using red mud—an aluminium industrywaste, Water Res. 35 (2001) 1125–1134.

[8] J. Pradhan, S.N. Das, R.S. Thakur, Adsorption of hexavalent chromium fromaqueous solution by using activated red mud, J. Colloid Interface Sci. 217 (1999)137–141.

[9] A.I. Zouboulis, K.A. Kydros, Use of red mud for toxic metals removal: the caseof nickel, J. Chem. Tech. Biotechnol. 58 (1993) 95–101.

10] R. Apak, G. Atun, K. Güclü, E. Tütem, G. Keskin, Sorptive removal of cesium-137and strontium-90 from water by unconventional sorbents usage of bauxitewastes (red muds), J. Nucl. Sci. Technol. 32 (1995) 1008–1017.

11] S.W. Zhang, C.J. Liu, Z.K. Luan, X.J. Peng, Arsenate removal from aqueous solu-tions using modified red mud, J. Hazard. Mater. 152 (2008) 486–492.

12] Y. Li, C. Liu, Z. Luan, X. Peng, C. Zhu, Z. Chen, Z. Zhang, J. Fan, Z. Jia, Phosphateremoval from aqueous solutions using raw and activated red mud and fly ash,J. Hazard. Mater. B137 (2006) 374–383.

13] W. Huang, S. Wang, Z. Zhu, L.Li.X. Yao, V. Rudolph, F. Haghseresht, Phosphateremoval from wastewater using red mud, J. Hazard. Mater. 158 (2008) 35–42.

14] Y. Cengeloglu, E. Kır, M. Ersöz, Removal of fluoride from aqueous solution byusing red mud, Sep. Purif. Technol. 28 (2002) 81–86.

15] A. Tor, N. Donaoglu, G. Arslan, Y. Cengeloglu, Removal of fluoride from waterby using granular red mud: batch and column studies, J. Hazard. Mater. 164(2009) 271–278.

16] C. Namasivayam, D.J.S.E. Arasi, Removal of Congo red from wastewater byadsorption onto waste red mud, Chemosphere 34 (1997) 401–417.

17] C. Namasivayam, R.T. Yamuna, D.J.S.E. Arasi, Removal of procion orange fromwastewater by adsorption on waste red mud, Sep. Sci. Technol. 37 (2002)2421–2431.

18] S.H. Lin, R.S. Juang, Adsorption of phenol and its derivatives from water usingsynthetic resins and low-coast natural adsorbents: a review, J. Environ. Manage.90 (2009) 1336–1349.

19] G.E. Ho, R.A. Gibbs, K. Mathew, Bacteria and virus removal from secondaryeffluent in sand and red mud columns, Water Sci. Technol. 23 (1991) 261–270.

20] US EPA, Overview of Solid Waste Generation, Management and Chemical Char-acteristics in the Bauxite Refining and Primary Aluminum Industry, Office ofSolid Waste, Washington, DC, November 1984.

21] US EPA, Code of Federal Regulations, Title 40, part 261, sect. 4(b)(7)(iii), 1990.22] V. Orescanin, K. Durgo, J. Franekic-Colic, K. Nad, V. Valkovic, Physical, chemical,

and genotoxic properties of waste mud by-product of waste water treatment,J. Trace Microprobe Tech. 21 (2003) 123–132.

23] Draft toxicological profile for nickel, Agency for Toxic Substances and DiseaseRegistry (http://www.atsdr.cdc.gov), U.S. Department of Health and HumanServices, Atlanta, Georgia, 2001.

24] Directive 2006/11/EC of the European Parliament and of the Council of 15February 2006 on pollution caused by certain dangerous substances dischargedinto the aquatic environment of the Community. Off. J. Eur. Union, 4.3.2006.

25] US EPA, 2009 Edition of the Drinking Water Standards and Health Advisories,822-R-09-011, Office of Water U.S. Environmental Protection Agency, Wash-ington, DC, October 2009.

26] Council Directive 98/83/EC of 3 November 1998 on the quality of waterintended for human consumption. Off. J. Eur. Union, 5.12.98.

27] World Health Organization (WHO), Guidelines for Drinking-Water Quality,Incorporating First Addendum to 3rd ed., vol. 1 Recommendations, WorldHealth Organization, Geneva, 2006, pp. 595.

28] K.G. Bhattacharyya, S.S. Gupta, Uptake of Ni(II) ions from aqueous solution bykaolinite and montmorillonite: influence of acid activation of the clays, Sep.Sci. Technol. 43 (2008) 3221–3250.

Page 10

ineeri

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

Metall. Rev. 119 (2010) 153–162.[58] C. Mansour, G. Lefèvre, E.M. Pavageau, H. Catalette, M. Fédoroff, S. Zanna, Sorp-

tion of sulfate ions onto magnetite, J. Colloid Interface Sci. 331 (2009) 77–

S. Smiljanic et al. / Chemical Eng

29] S.M. Dal Bosco, R.S. Jimenez, W.A. Carvalho, Removal of toxic metals fromwastewater by Brazilian natural scolecite, J. Colloid Interface Sci. 281 (2005)424–431.

30] C. Gabaldón, M. Izquierdo, M. Martin, P. Marzal, Fixed-bed removal of freeand complexed Ni from synthetic and industrial aqueous solutions, Sep. Sci.Technol. 43 (2008) 1157–1173.

31] N. Akhtar, J. Iqbal, M. Iqbal, Removal and recovery of nickel(II) from aque-ous solution by loofa sponge-immobilized biomass of Chlorella sorokiniana:characterization studies, J. Hazard. Mater. B108 (2004) 85–94.

32] G. Arslan, S. Cetin, E. Pehlivan, Removal of Cu(II) and Ni(II) from aqueous solu-tion by lignite-based humic acids, Energy Sources 29 (2007) 619–630.

33] P.N. Pathak, G.R. Choppin, Nickel(II) sorption on hydrous silica: a kinetic andthermodynamic study, J. Radioanal. Nucl. Chem. 268 (2006) 467–473.

34] M. Ajmal, R.A.K. Rao, R. Ahmad, J. Ahmad, Adsorption studies on Citrus reticu-lata (fruit peel of orange): removal and recovery of Ni(II) from electroplatingwastewater, J. Hazard. Mater. B79 (2000) 117–131.

35] N. Miralles, M. Martínez, A. Florido, I. Casas, N. Fiol, I. Villaescusa, Grape stalkswaste as low cost biosorbents: an alternative for metal removal from aqueoussolutions, Solvent Extract. Ion Exchange 26 (2008) 261–270.

36] E. Malkoc, Y. Nuhoglu, Investigations of nickel(II) removal from aqueous solu-tions using tea factory waste, J. Hazard. Mater. B127 (2005) 120–128.

37] H. Hasar, Adsorption of nickel(II) from aqueous solution onto activated carbonprepared from almond husk, J. Hazard. Mater. B97 (2003) 49–57.

38] N. Fiol, I. Villaescusa, M. Martínez, N. Miralles, J. Poch, J. Serasols, Sorption ofPb(II), Ni(II), Cu(II), and Cd(II) from aqueous solution by olive stone waste, Sep.Purif. Technol. 50 (2006) 132–140.

39] S. Al-Asheh, F. Banat, F. Mohai, Sorption of copper and nickel by spent animalbones, Chemosphere 39 (1999) 2087–2096.

40] E. Pehlivan, S. Cetin, Application of fly ash and activated carbon in the removalof Cu2+ and Ni2+ ions from aqueous solutions, Energy Sources 30 (2008)1153–1165.

41] V.K. Jha, Y. Kameshima, A. Nakajima, K. Okada, K.J.D. MacKenzie, Multi func-tional uptake behaviour of materials prepared by calcining waste paper sludge,J. Environ. Sci. Health A 41 (2006) 703–719.

42] S. Smiljanic, I. Smiciklas, A. Peric-Grujic, B. Loncar, M. Mitric, Rinsed and ther-mally treated red mud sorbents for aqueous Ni2+ ions, Chem. Eng. J. 162 (2010)75–83.

43] D. Uzun, M. Gulfen, Dissolution kinetics of iron and aluminium from red mud

in sulfuric acid solution, Indian J. Chem. Technol. 14 (2007) 263–268.

44] S.K. Milonjic, A.Lj. Ruvarac, M.V. Susic, The heat of immersion of natural mag-netite in aqueous solutions, Thermochim. Acta 11 (1975) 261–266.

45] US EPA, Test Methods for Evaluating Solid Waste, Laboratory Manual Physi-cal/Chemical Methods, SW-846, 3rd ed., U.S. Gov. Print. Office, Washington,DC, 1999.

[

ng Journal 168 (2011) 610–619 619

46] J. Huang, Y. Liu, X. Wang, Selective adsorption of tannin from flavonoidsby organically modified attapulgite clay, J. Hazard. Mater. 160 (2008) 382–387.

47] I. Langmuir, The adsorption of gases on plane surfaces of glass, mica, and plat-inum, J. Am. Chem. Soc. 40 (1918) 1361–1403.

48] N. Rajic, D.J. Stojakovic, M. Jovanovic, N. Zabukovec Logar, M. Mazaj, V. Kaucic,Removal of nickel(II) ions from aqueous solutions using the natural clinoptilo-lite and preparation of nano-NiO on the exhausted clinoptilolite, Appl. SurfaceSci. 257 (2010) 1524–1532.

49] V. Kumar Jha, M. Matsuda, M. Miyake, Sorption properties of the activatedcarbon–zeolite composite prepared from coal fly ash for Ni2+, Cu2+, Cd2+ andPb2+, J. Hazard. Mater. 160 (2008) 148–153.

50] S. Yang, J. Li, Y. Lu, Y. Chen, X. Wang, Sorption of Ni(II) on GMZ bentonite: effectsof pH, ionic strength, foreign ions, humic acid and temperature, Appl. Radiat.Isotopes 67 (2009) 1600–1608.

51] S. Yang, D. Zhao, H. Zhang, S. Lu, L. Chen, X. Yu, Impact of environmental condi-tions on the sorption behavior of Pb(II) in Na–bentonite suspensions, J. Hazard.Mater. 183 (2010) 632–640.

52] Y. Ma, C. Lin, Y. Jiang, W. Lu, C. Si, Y. Liu, Competitive removal of water-bornecopper, zinc and cadmium by a CaCO3-dominated red mud, J. Hazard. Mater.172 (2009) 1288–1296.

53] M.B. McBride, Chemisorption and Precipitation of Inorganic Ions. Environmen-tal Chemistry of Soils, Oxford Univ. Press, New York, 1994, pp. 121–168.

54] C. Collins, K.V. Ragnarsdottir, D.M. Shermane, Effect of inorganic and organicligands on the mechanism of cadmium sorption to goethite, Geochim. Cos-mochim. Acta 63 (1999) 2989–3002.

55] R.M. Cornell, U. Schwertmann, The Iron Oxides, Structure, Properties, Reac-tions, Occurrences and Uses, 2nd ed., Wiley-VCH Verlag GMbH & Co. KGaA,Weinheim, 2003.

56] A. Violante, M. Ricciardella, M. Pigna, R. Capasso, Effects of organic ligands onthe adsorption of trace elements onto metal oxides and organo-mineral com-plexes, in: G.R. Gobran, P.M. Huang (Eds.), Biogeochemistry of Trace Elementsin the Rhizosphere, 2005, pp. 157–182.

57] E. Chockalingam, S. Subramanian, J.-J. Braun, Bioremediation of acid mine waterutilising red mud and Desulfotomaculum nigrificans, Miner. Process. Extract.

82.59] K. Tonsuaadu, K. Viipsi, A. Trikkel, EDTA impact on Cd2+ migration in

apatite–water system, J. Hazard. Mater. 154 (2008) 491–497.