Rapid and high capacity adsorption of heavy metals by Fe3O4/montmorillonite nanocomposite using response surface methodology: Preparation,characterization,optimization, equilibrium

ARTICLE IN PRESSJID: JTICE [m5G;December 1, 2014;7:6]

Journal of the Taiwan Institute of Chemical Engineers 000 (2014) 1–7

Contents lists available at ScienceDirect

Journal of the Taiwan Institute of Chemical Engineers

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

Rapid and high capacity adsorption of heavy metals by

Fe3O4/montmorillonite nanocomposite using response surface

methodology: Preparation, characterization, optimization, equilibrium

isotherms, and adsorption kinetics study

Katayoon Kalantari∗, Mansor B. Ahmad∗∗, Hamid Reza Fard Masoumi, Kamyar Shameli,Mahiran Basri, Roshanak Khandanlou

Department of Chemistry, Faculty of Science, Universiti Putra Malaysia, 43400, Serdang, Selangor, Malaysia

a r t i c l e i n f o

Article history:

Received 12 June 2014

Revised 19 October 2014

Accepted 25 October 2014

Available online xxx

Keywords:

Heavy metals

Fe3O4/montmorillonite nanocomposites

Adsorption

Response surface methodology (RSM)

Adsorption kinetics

a b s t r a c t

Fe3O4/montmorillonite nanocomposite (Fe3O4/MMT NC) was synthesized for removal of Pb2+, Cu2+ and Ni2+

ions from aqueous systems. The nanoadsorbent was characterized by X-ray diffraction and transmission elec-

tron microscopy and mean diameter of magnetic nanoparticles was about 8.24 nm. The experiments were

designed by response surface methodology and quadratic model was used to prediction of the variables.

The adsorption parameters of adsorbent dosage, removal time, and initial heavy metal ions concentration

were used as the independent variables and their effects were investigated on the heavy metal ions removal.

Variance analysis was utilized to judge the adequacy of the chosen models. Optimum conditions with initial

heavy metal ions concentration of 510.16, 182.94, and 111.90 mg/L, 120 s of removal time and 0.06 g/0.025 L,

0.08 g/0.025 L, and 0.08 g/0.025 L of adsorbent amount were given 89.72%, 94.89%, and 76.15% of removal effi-

ciency Pb2+, Cu2+ and Ni2+ ions, respectively. Prediction of models was in good agreement with experimental

results and Fe3O4/MMT NC was found successful in removing heavy metals from their aqueous solutions.

© 2014 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

1

w

h

c

l

h

t

n

b

r

h

k

u

t

p

m

a

l

s

h

n

t

b

n

m

p

s

i

c

h

l

m

t

t

h

1

. Introduction

Pollution of water by heavy metal ions occurs throughout the

orld [1]. Heavy metals pollution is a serious problem threatening

uman health and other organisms due to their high level toxicity,

arcinogenicity and non-biodegradability, even though they are in

ow concentration [2,3].

Cu2+, Ni2+ and Pb2+ ions are classified as toxic metals. Several

ealth problems have been associated with these toxic metals. Gas-

rointestinal diseases such as nausea and stomach ache are the most

oted problems associated with Cu ion poisoning. Anemia disease has

een resulted by low amounts of Pb while presence of high amounts

esults in serious disorder of the central nervous system. Presence of

igh levels of Ni in consumed water may cause serious harm to the

idneys, lungs, and the skin [4,5]. Numerous techniques have been

sed for metal ions removing from aqueous solutions [6]. Among

hese techniques, adsorption method is accepted as one of the most

romising and effective approaches [7]. The significant benefit of

∗ Corresponding author. Tel.: +60 142336067.∗∗ Corresponding author.

E-mail addresses: [email protected], [email protected] (K. Kalantari),

[email protected] (M.B. Ahmad).

c

o

a

f

c

ttp://dx.doi.org/10.1016/j.jtice.2014.10.025

876-1070/© 2014 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All righ

Please cite this article as: K. Kalantari et al., Rapid and high capacity ads

using response surface methodology: Preparation, characterization, op

Journal of the Taiwan Institute of Chemical Engineers (2014), http://dx.do

dsorption is that adsorbent has high tendency and great level of

oading capacity for heavy metal ions [8]. Among the available ad-

orbents, metal nanoparticles are categorized as the best ones for

eavy metal ion removal from aqueous solutions. Beside traditional

anoparticles, magnetic nanoparticles are gaining increasing atten-

ion in remediation of the environment due to the fact that they can

e simply separated from water using a magnetic field [9]. Magnetic

anoparticles have been used for the adsorption of numerous heavy

etal ions [10]. Additionally, for the recycling and regeneration pur-

ose, magnetic nanoparticles based composite adsorbents allowed

imple separation from aqueous systems. This kind of facile separat-

ng is important to enhance the function performance and reduce the

ost during treatment of water and waste water [11]. The resulting

eterogeneous nanocomposite seemed greater in recycling used with

ower cost in comparison with their homogeneous counterparts. It

ight be quite fascinating to prepare supported magnetite nanopar-

icles and to discover the influence of the supporting components on

he related activities and characteristics of the magnetic nanoparti-

les [12]. The adsorbents such as clay [13], have been used broadly in

rder to removing metal ions from aqueous solutions due to higher

dsorption capacity and low cost materials. Clays are environmental

riendly and also have alumina silicate minerals layers with negatively

harges that make excellent cationic adsorbents materials due to their

ts reserved.

orption of heavy metals by Fe3O4/montmorillonite nanocomposite

timization, equilibrium isotherms, and adsorption kinetics study,

i.org/10.1016/j.jtice.2014.10.025

http://dx.doi.org/10.1016/j.jtice.2014.10.025

http://www.ScienceDirect.com

http://www.elsevier.com/locate/jtice

mailto:[email protected]





2 K. Kalantari et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2014) 1–7


t

1

s

i

s

d

A

e

i

w

Y

w

i

(

3

3

s

w

a

a

F

t

b

p

s

fi

s

3

t

f

T

b

a

3

s

r

(

i

p

r

T

C

N

T

[

d

C

N

significant surface areas [14]. Montmorillonite has been traditionally

applied regarding the removal of metal ions and organic contami-

nants from drinking and waste water. The adsorption properties of

montmorillonite can often be enhanced either by the intercalation of

metal ions in the space between the layers. Of these modifications,

clays with metal ions interlayered have been broadly analyzed re-

cently [15]. The layers of montmorillonite have negative charges so

montmorillonite has cation exchange capacity. The negative charge

is compensated by Ca2+ and Na+ intercalated between the layers area

[16,17]. The objective of the present work is to consider the feasi-

bility of a Fe3O4/montmorillonite nanocomposite for the removal of

lead, copper and nickel ions from aqueous solution and to optimize

parameters for maximum removal efficiency by a central composite

rotatable design under response surface methodology (RSM). RSM

is an empirical statistical and mathematical technology that has at-

tracted tremendous interest for examining the effects of numerous

independent parameters [18]. The application of RSM has been em-

phasized to enhance, develop, and optimize complex procedures and

to assess the magnitude of numerous influencing variables. It has

been broadly used in chemical engineering and sorption process opti-

mization [19]. The optimization of copper, nickel and lead adsorption

were performed by varying three independent factors (heavy metal

ion concentration, removal time and adsorbent dosage) to predict

the response (removal percentage). The characterization of sorbent

was carried out by X-ray diffraction and transmission electron mi-

croscopy. The kinetics and isotherms parameters were analyzed to

determine the adsorption mechanism.

2. Materials

All chemical reagents were of analytical grade. FeCl3� 6H2O and

FeCl2� 4H2O of 96% (GPR, USA) were used as the iron precursor and

also, montmorillonite were obtained from (Kunipia-F, Japan). NaOH of

99% was obtained from Merck (Germany). CuCl2, NiCl2 and Pb(NO3)2

were supplied by Hamburg Chemical. All these aqueous solutions

were prepared with deionized water.

2.1. Fe3O4/montmorillonite nanocomposite (adsorbent) preparation

The Fe3O4 nanoparticles (Fe3O4-NPs) were prepared using chem-

ical co-precipitation technique. As first step, 120 mL of deionized

water was bubbled by N2 gas for 15 min and then 2.0 g of montmo-

rillonite and measured amount of Fe3+ and Fe2+ with the (1:2) molar

ratio were dissolved in deionized water with vigorous stirring for 2 h.

Under the protection of nitrogen gas, 25 mL of fresh NaOH (2 M) was

added drop-wise into this solution. Finally, the suspensions were cen-

trifuged, washed twice with distilled water and ethanol, and kept for

2 h in a vacuum stove at 100 °C.

2.2. Preparation of the heavy metal solutions

First, appropriate amounts of CuCl2, NiCl2 and Pb(NO3)2 were dis-

solved in deionized water and transferred into the 1-liter volumetric

flasks for preparing three stock solutions (1000 mg/L) of Cu2+, Pb2+

and Ni2+ ions. These stock solutions were diluted with deionized wa-

ter to obtain the certain concentration range of Cu2+, Pb2+ and Ni2+

standard solutions.

2.3. Batch adsorption experiments

For the removing of heavy metal ions, batch adsorption analysis

was performed in 100 mL glass flasks. For this purpose, 0.02–0.08 g

of Fe3O4/MMT NC was added into 25 mL of heavy metal ions solu-

tion with the initial ion concentration (C0) ranged between 97.73 and

602.27 (mg/L). The pH of the solution was not changed to prevent

precipitation. The mixture solution was agitated in a shaker at room




emperature and desired speed with removal time between 40 and

46 s according to RSM design. After reach to designed removal time,

orbent separated from the solution magnetically and the heavy metal

ons content in residual solutions were analyzed using an atomic ab-

orption spectroscopy (Thermo scientific, S series). The experimental

ata with different mathematical models were evaluated and their

NOVA results showed that the reaction of removals was most prop-

rly demonstrated with a “quadratic” polynomial model for all three

ons. All experiments were done in duplicate and the averaged values

ere considered.

The removal efficiency of heavy metal ions was calculated as:

=(

C0 − Ct

C0

)× 100 (1)

here Y is the metal ions adsorption percentage, C0 and Ct are the

nitial concentration of heavy metal ions at time (0 and t) in solution

mg/L).

. Results and discussion

.1. Characterization of Fe3O4/montmorillonite nanocomposite

The TEM image and distribution histogram of the Fe3O4-NPs as

hown in (Fig. 1(a and b)) illustrated that the prepared Fe3O4-NPs

ere in average diameter around 8.24 ± 1.258 nm. The identity

nd structure of the Fe3O4/MMT NC were verified by XRD (Fig. 1(c)),

nd the nano crystallite peaks with matching well with the standard

e3O4 (01-088-0315). The verification of paramagnetic properties of

he Fe3O4/MMT NC was done by the magnetization curve measured

y vibrating sample magnetometer. The saturation magnetization of

repared nanocomposite was found to be 2.9 emu/g. Therefore, the

ynthesized Fe3O4/MMT NC is expected to respond well to magnetic

elds and therefore the separation of the liquid and solid phases is

imple. Moreover, it is well known that Fe3O4-NPs with less than

0 nm diameter show super paramagnetism [20]. The surface area of

he nanocomposite was measured to be 121.279 and 210.660 m2/g

or MMT and Fe3O4/MMT NC, respectively, by the Brunauer–Emmett–

eller gas sorptometry measurements method. The BET isotherm is

ased on determining the extent of nitrogen adsorption on a surface

rea.

.2. Fitting of process models and statistical analysis

The initial heavy metal ions concentration, removal time, and ad-

orbent dosage were chosen as the independent variables. In this

esearch, twenty runs of the “Central Composite Rotatable Design”

CCRD) experimental design consisted of eight factorial points, six ax-

al points and also six center points, were conducted with three inde-

endent factors (initial metal ion concentration (97.73–602.27 mg/L),

emoval time (40–146 s) and adsorbent dosage (0.2–0.8 g/0.025 L).

he predicted values and experimental design matrix obtained by

CRD are given in Table 1.

The analysis of variance (ANOVA) for adsorption study of Cu2+,

i2+ and Pb2+ ions was applied in order to ensure a desirable model.

he ANOVA results are shown in Table 2.

Below are empirical polynomial quadratic model equations (2–4),

21,22], for Cu2+, Ni2+ and Pb2+ respectively that were used to fit the

ata:

u2+removal (%) = 49.16 − 19.86A + 1.20B + 11.05C + 0.81AB

− 2.82AC + 0.77BC + 3.04A2

− 1.71B2 − 0.49C2 (2)

i2+removal (%) = 35.89 − 5.96A − 1.44B + 7.22C + 1.59AB

− 2.53AC + 4.36A2 + 1.50B2

− 2.08C2 − 2.93A3 (3)



i.org/10.1016/j.jtice.2014.10.025


K. Kalantari et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2014) 1–7 3


Fig. 1. TEM and XRD patterns of Fe3O4/MMT nanocomposite.

Table 1

Predicted and experimental design matrix obtained by CCRD.

Run no. Initial ion concentration (mg/L) Removal time (s) Adsorbent dosage (g) Removal of Cu (II) (%) Removal of Ni(II) (%) Removal of Pb(II) (%)

Exp. Pre. Exp. Pre. Exp. Pre.

1 97.73 90.00 0.05 93.37 91.17 73.81 71.85 91.30 92.99

2 200.00 60.00 0.03 54.77 56.38 39.34 41.60 99.89 99.24

3 200.00 60.00 0.07 81.46 82.58 59.96 61.11 88.63 87.28

4 200.00 120.00 0.03 54.36 55.63 35.36 35.53 100.0 99.66

5 200.00 120.00 0.07 84.75 84.89 53.06 55.03 92.61 91.56

6 350.00 40.00 0.05 43.49 42.32 45.28 44.56 93.07 93.91

7 350.00 90.00 0.02 30.54 29.19 19.41 18.86 78.00 77.64

8 350.00 90.00 0.05 50.50 49.16 33.19 34.89 97.39 96.76

9 350.00 90.00 0.05 49.95 49.16 37.29 35.89 97.31 96.76

10 350.00 90.00 0.05 48.41 49.16 36.46 35.89 97.15 96.76

11 350.00 90.00 0.05 52.40 49.16 37.73 35.89 96.85 96.76

12 350.00 90.00 0.05 46.10 47.16 34.57 35.89 95.30 96.76

13 350.00 90.00 0.05 47.49 48.16 35.43 35.89 96.71 96.76

14 350.00 90.00 0.08 65.79 66.36 44.52 42.15 94.15 95.47

15 350.00 146.00 0.05 45.96 46.36 38.92 37.71 98.10 98.22

16 500.00 60.00 0.03 20.27 20.67 25.16 26.19 61.44 61.81

17 500.00 60.00 0.07 36.33 35.61 34.19 35.57 96.42 96.08

18 500.00 120.00 0.03 23.74 23.17 25.10 26.50 66.94 67.61

19 500.00 120.00 0.07 42.23 41.17 35.13 35.87 95.85 95.82

20 602.27 90.00 0.05 22.95 23.38 26.57 24.61 70.00 69.27

P

w

i

t

t

s

p

ering the determination coefficient (R2 = 0.9941, 0.9834 and 0.9955),

b2+removal (%) = 96.56 − 7.05A + 1.28B + 5.30C + 10.32AC

− 5.50A2 − 3.58C2 − 1.24ABC (4)

here (A), (B) and (C) are the initial concentration of heavy metal

ons, the removal time, and the adsorbent dosage respectively, as




he experimental variables. Eqs. (2–4), make a good visualization of

he influence of each parameter and their interactions on the re-

ponse. Good correlation also, between the predicted values and ex-

erimental results show that the designs are properly fitted, consid-



i.org/10.1016/j.jtice.2014.10.025




Table 2

Analysis of variance of the fitted quadratic equation and model summary statistics A: initial ion concentration (mg/L), B: removal time (s), and C: adsorbent

dosage (g).

Source Removal of Cu(II) (%) Removal of Ni(II) (%) Removal of Pb(II) (%)

Mean square F-value P-value Mean square F-value P-value Mean square F-value P-value

Model 816.01 186.34 <0.0001 324.04 194.68 <0.0001 253.05 172.96 <0.0001

A 5385.61 1229.83 <0.0001 91.91 311.54 0.0068 679.08 464.10 <0.0001

B 19.70 4.50 0.0599 28.41 3.57 0.0883 22.37 15.29 0.0036

C 1667.65 380.82 <0.0001 712.02 189.39 <0.0001 383.83 262.32 <0.0001

AB 5.27 1.20 0.2986 20.35 2.56 0.1410 0.088 0.060 0.8116

AC 63.45 14.49 0.0034 51.31 6.44 0.0295 851.61 582.01 <0.0001

BC 4.70 1.07 0.3248 274.26 – – 0.61 0.41 0.5362

A2 133.53 30.49 0.0003 32.52 34.43 0.0002 439.85 300.61 <0.0001

B2 41.94 9.58 0.0114 62.41 4.08 0.0709 0.86 0.59 0.4629

C2 3.46 0.79 0.3952 96.70 7.84 0.0188 187.45 128.11 <0.0001

ABC – – – – – – 12.35 8.44 0.0174

A3 – – – – 12.14 0.0059 – – –

Residual 4.38 – – 7.97 – – 1.46 – –

Lack of fit 3.63 0.71 0.6427 8.83 1.25 0.4079 2.54 4.26 0.0720

Pure error 5.13 – – 7.10 – – 0.60 – –

Standard deviation 2.09 2.02 1.21

PRESS 175.09 194.32 142.44

R2 0.9941 0.9834 0.9948

Adjusted R2 0.9887 0.9495 0.9891

Predicted R2 0.9763 0.9311 0.9440

Adequate precision 47.643 27.053 42.190

Fig. 2. Response surface 3D plots of interaction between heavy metal ions concentration and removal time and between heavy metal ions concentration and adsorbent dosage.

3

s

a

i

d

a

p

e

i

t

t

i

c

t

a

t

respectively for Cu2+, Ni2+ and Pb2+ removal. The high value of the

adjusted determination coefficient implicates to the significance of

the model parameters.

The low P-value (probability) (<0.0001) with F-value (186.34,

94.68 and 172.96) for Cu2+, Ni2+ and Pb2+ respectively, implied

that the model was accurate. Generally, highly significant regres-

sion model is justified by higher Fischer’s ‘F statistics’ values with

P-value as low as possible [23]. The term of lack-of-fit is actually

non-significant as it is essential. The quadratic model was valid for

our work, due to the non-significant value of lack of fit (more than

0.05). The linear effects of all the variables were significant and A (ini-

tial heavy metal concentration) and C (adsorbent dosage) were also

appreciable as F has high values and played important role on the

removal efficiency. As shown in Table 2, the prediction error sum of

squares (PRESS) provides a good residual scaling. Usually, a signifi-

cant difference between the PRESS residual and the ordinary residual

(170.71, 184.56 and 140.98) demonstrates a point where the model

fits the data well.




.2.1. Three-dimensional response surface plots

To gain the better comprehensive of Cu2+, Ni2+ and Pb2+ ions ad-

orption process, the three dimensional response surface plots were

nalyzed. In each plot, the influence of two factors on heavy metal

ons adsorption capacity was shown. The response surface plots are

emonstrated in Figs. 2 and 3. Initial heavy metal ions concentration

nd adsorbent dosage are the most important parameters that can im-

ress on the adsorption efficiency. Fig. 2(a–c) shows the simultaneous

ffect of removal time and initial ion concentration on heavy metal

ons adsorption efficiency. Also, for Cu2+ and Ni2+, in constant removal

ime, adsorption efficiency decreased with heavy metal ions concen-

ration enhancement. While for Pb2+ it can be seen an improvement

n removal efficiency until about 500 ppm and then a significant de-

reasing was occurred. This is due to lack of available active sites on

he adsorbent surface [24]. However, there are not enough spaces for

ll the ions in high concentration of metal ions [25].

Metal ions and adsorbent dosage is shown in Fig. 2(d–f). In

his figure, the removal efficiency improved with increasing the



i.org/10.1016/j.jtice.2014.10.025




Table 3

validation set.

Initial ion concentration (mg/L) Removal time(s) Adsorbent dosage (g) Cu(II) removal% Ni(II) removal% Pb(II) removal%

Exp. Pre. Exp. Pre. Exp. Pre.

300 90 0.05 55.87 56.12 34.65 35.89 97.63 98.49

400 90 0.05 41.39 42.88 33.24 34.36 92.43 93.79

350 70 0.06 51.56 52.75 41.58 40.60 96.53 97.63

200 110 0.06 63.33 64.25 41.12 42.14 99.21 100

350 150 0.04 37.21 38.32 33.27 34.88 93.89 95.06

400 120 0.04 46.11 47.96 36.21 37.62 96.42 97.98

Table 4

Optimized condition.

Metal Initial ion concentration (mg/L) Removal time (s) Adsorbent dosage (g) Removal (%)

Actual Predicted Error Std. dev.

Cu(II) 182.94 120.00 0.08 94.89 95.82 0.93 0.65

Ni(II) 111.90 120.00 0.08 76.15 77.43 1.28 0.90

Pb(II) 510.16 120.00 0.06 89.72 90.53 0.81 0.57

Fig. 3. Langmuir isotherm fitting by the magnetic/MMT nanocomposite.

a

t

i

o

N

T

P

b

o

w

m

t

3

w

p

e

t

c

T

a

3

s

p

w

[

o

Table 5

Values of Langmuir and Freundlich constant for the heavy metal ions

removal.

Isotherm Langmuir Freundlich

R2 b Qm R2 n KF

Cu(II) 0.9822 0.08 70.92 0.9210 7.69 0.62

Ni(II) 0.9789 0.04 65.78 0.9001 7.70 0.63

Pb(II) 0.9922 0.163 263.15 0.9348 6.50 0.55

o

s

i

w

e

r

s

t

a

N

p

E

l

w

a

t

i

i

m

h

3

a

t

r

dsorbent dosage. This increase in removal efficiency is due to that,

otal available binding sites for the interactions with the heavy metal

ons increase with increasing nanoadsorbent amount.

The selectivity sequence of Pb2+ > Cu2+ > Ni2+ adsorption was

bserved on Fe3O4/MMT NC. The differences in the radius of Cu2+,

i2+ and Pb2+ ions have significant influence on adsorption efficiency.

he heavy metal removal efficiency followed the decreasing order:

b2+ > Cu2+ > Ni2+. Since the radius of Pb2+ (1.32 A) is noticeably

igger than that of Cu2+ (0.73 A) and Ni2+ (0.69 A), the hydration

f Pb2+ is more difficult compared to Ni2+ and also forming a larger

ater layer on the surface. As a result, Ni2+ and Cu2+ are more

obile in solution and would have a lesser tendency to adsorb on

he nanoadsorbent [26].

.2.2. Comparison of the results

Table 3 shows six new trials combinations of experimental factors,

hich do not participate in the training data set (model). The com-

arison between experimental observed and predicted data shows

xcellent agreement for all responses.

Additionally to support the optimized data as provided by statis-

ical modeling in optimized conditions, the confirmatory tests were

arried out with the variables as proposed by the model (Table 4).

hese experiments displaying that the results were in good

greement.

.2.3. Sorption isotherms

The equilibrium between heavy metal ions adsorbed on the ad-

orbent and the one retained in the aqueous solution is generally dis-

layed in two different adsorption isotherms. These equilibriums are

idely demonstrated by using Langmuir and Freundlich relationship

27]. The Langmuir isotherm considers that adsorption in monolayer

ccurs on an adsorbent that have a homogeneous surface structural,




n which the binding sites possess the same tendency for the ad-

orption, and no interaction occurs. The linear from of the Langmuir

sotherm is shown in the following equation:

Ce

qe= 1

bQm+ Ce

Qm(5)

here qe (mg/g) and Ce (mg/L) signify the adsorption capacity and the

quilibrium concentration of the adsorbate in the aqueous solution,

espectively; Qm (mg/g) is the maximum adsorption capacity of ad-

orbents and b (L/mg) is the constant of the Langmuir model related

o the affinity of binding sites and is a measurement of the energy of

dsorption [28]. Fig. 3 shows the Langmuir isotherms fitting of Cu2+,

i2+ and Pb2+ adsorption.

Additionally, Freundlich model is an empirical expression em-

loyed to explain a heterogeneous system and is described in

q. (6):

og qe = log KF + 1

n· log Ce (6)

here KF and n represent the equilibrium constants indicative of the

dsorption capacity and adsorption intensity, respectively (if n > 1,

he adsorption is considered favorable).

The values for constant KF and n were obtained from the slope and

ntercept of the plot of log (qe)versus log (Ce) [29] and they are listed

n Table 5.

In comparison with the Freundlich isotherm model, the Longmuir

odel fits much better with the experimental values because of the

igher correlation coefficients which apply a monolayer adsorption.

.2.4. Kinetics studies

Adsorption kinetics has been examined for the purpose to obtain

n understanding how the metal ions move rapidly from solution to

he adsorbent, as well as the removal time needed to reach equilib-

ium between phases. The adsorption kinetics of Fe3O4/MMT NC is



i.org/10.1016/j.jtice.2014.10.025




Fig. 4. Kinetics models for the adsorption of heavy metals on magnetic/MMT nanocom-

posite.

t

t

i

d

t

s

b

t

d

f

I

n

s

d

c

m

T

i

T

n

s

f

c

n

c

F

s

q

a

m

e

a

t

c

[

b

F

a

4

t

t

c

f

h

d

m

s

m

w

a

presented in Fig. 4. The kinetics data were dealt with pseudo-first or-

der and pseudo-second order models. In order to estimate the kinetics

adsorption mechanism, pseudo-first-order and pseudo-second-order

models were employed to interpret the experimental data as given in

Eqs. (7) and (8):

ln (qe − qt) = ln qe − k1t (7)

t

qt= 1

k2qe2

+ 1

qet (8)

where k1(min−1) and k2 (g/mg min) is the rate constant of the pseudo-

first-order and pseudo-second order adsorption, respectively. The

term of qe is the amount of ions adsorbed at equilibrium (mg/g),

qt (mg/g) is the amount of heavy metal ions adsorbed at time (t)

[30]. The kinetics studies for the adsorption of Cu2+, Ni2+ and Pb2+ on

Fe3O4/MMT NC were performed using pseudo-first order and pseudo-

second order kinetics models and pseudo-second order kinetics plot

provided the perfect straight line for the adsorption of all metals on

adsorbent surface area (Fig. 4).

The fitting results of pseudo-first-order and pseudo-second order

adsorption are presented in Table 6.

As shown in Table 6, pseudo-second order model provided much

better correlation coefficients than pseudo-first order model, and sug-

gesting pseudo-second order model is more desirable to explain the

adsorption kinetics processes.

3.2.5. Adsorption mechanism

When montmorillonite sheets are broken, two different surfaces

are created, one resulting from the easy cleavage of the layers, known

as “faces” composed of completely compensated oxygen atoms, show

low electrical charge, and have nonpolar properties in water and hy-

drophobic, also the other arising from the break of the ionic bonds

within the layers, named “edges” made up of hydroxyl ions, siloxane

groups (–Si–O–Si–), O2−, and Mg ions that simply undergo hydrolysis,

provide a comparatively high level electrical charge and are polar in

water. Nevertheless, the hydrophilic effect was resulted from some

functional hydroxyl groups such as SiOH, and –MgOH on the edge

surfaces [31]. The faces and the edges of the clay particles can adsorb

polar pollutants, some anions and cations, from water. The pollu-

Table 6

Parameter of the kinetics models for the adsorption of heavy

metals onto magnetic/MMT nanocomposite.

Heavy metal ions First order Second order

R2 K1 R2 K2

Cu(II) 0.96 0.0203 0.9989 0.37

Ni(II) 0.933 0.0165 0.9885 0.26

Pb(II) 0.919 0.0068 0.9971 0.025

1

s

0

T

m

c

e

f

p

fi




ants adsorbed on clay surface resulting in their immobilization with

he processes of coordination, ion exchange, or interaction between

on and dipole. The contaminants can be held by H-bonding or van

er Waals interactions arising from either strong or weak interac-

ions. The strength of the interactions is determined by numerous

tructural and other features of the clay mineral [32]. According to

asic 2:1 structure, montmorillonite has high surface charge due to

he spread of isomorphous substitution in octahedral and tetrahe-

ral sheets [33]. Two main reaction mechanisms have been reported

or the adsorption of heavy metal ions process on montmorillonite.

on exchange, which is mainly of electrostatic nature on the perma-

ent negatively charged sites (the siloxane groups) by forming outer-

phere complex and other one complexation on the edge sites (hy-

roxyl groups) by forming inner-sphere complex which is mainly of

hemical nature. Due to the interfaces between aqueous solution and

ineral, surface complexation is a macroscopic adsorption behavior.

he hydrophilic negative surface sites of montmorillonite (SiO¯) are

nvestigated by adsorption of cationic ions used as molecular probes.

he electrostatic attraction between opposite charges of metals and

egative surface sites on montmorillonite leads to adsorption and

uggests a strong affinity. Metal ions can be bonded to edge sur-

aces of montmorillonite sheets using hydrogen bonding [34]. Ac-

ording to BET results, magnetic NPs provides higher surface area in

anocomposite, comparing to pure montmorillonite powder and in-

rease the ability of montmorillonite in heavy metal adsorption [35].

urthermore, the metal ions all rapidly adsorbed at less than 5 min,

o fast binding of metals to Fe(III) species at the external surface and

uickly accessed by metals can be mentioned as a subsidiary mech-

nism. The pure clay (montmorillonite) removes heavy metal ions

ainly by cation exchange, and subsidiary magnetic nanoparticles

xhibit high surface area, which results in enhanced heavy metals

dsorption [36].

This shows that most of the adsorption active sites of the nanopar-

icles can be found in the exterior of the adsorbent and are easily ac-

essible by the metal ions species, thus resulting in a rapid adsorption

37]. It is well known that electrostatic attraction and coordination,

etween metal ions and adsorbent is the mechanism of adsorption.

inally, in this research, the potential of Fe3O4 nano magnetic plays

n important role [38].

. Conclusion

Traditional methods that use one factor at a time are expensive and

ime consuming. Response surface methodology was used to optimize

he removal of heavy metals by synthesized Fe3O4/MMT NC. A central

omposite design was used to provide the experimental conditions

or removing the heavy metals. The variables include removal time,

eavy metals concentration and nano adsorbent dosage.

All the results show that synthesized Fe3O4/MMT NC stand apart

ue to its great adsorption capacity with respect to all the three heavy

etals. This shows that adsorption capacity of synthesized nanoad-

orbent is better than many other synthesized adsorbent; further-

ore it is a rapid adsorbent. A fast adsorption of Cu2+, Ni2+ and Pb2+

as found on the Fe3O4/MMT NC during less than 2 min. Moreover, in

queous systems, the high concentration of heavy metal ions (510.16,

82.94, and 111.90 mg/L) for Pb2+, Cu2+ and Ni2+ respectively were

uccessfully adsorbed in the low level amount of adsorbent (0.06 and

.08 g/0.025 L). Adsorption isotherms fitted well by Langmuir model.

he adsorption kinetics could be explained by pseudo-second-order

odel.

The synthesized nanoadsorbent showed characteristics of low

ost, environmentally friendly which made it great potential to be an

conomic and effective alternative for the removal of heavy metal ions

rom industrial effluents. The magnetic nanocomposite after removal

rocess is can simply recovered by applying an external magnetic

eld which is an environmentally clean alternative.



i.org/10.1016/j.jtice.2014.10.025




A

U

a

t

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

cknowledgments

The authors would like to acknowledge the financial support from

niversiti Putra Malaysia (UPM) (RUGS Project No. 9199840). They

re also grateful to the staff of the Department of Chemistry UPM and

he Institute of Bioscience UPM for the technical assistance.

eferences

[1] Karami H. Heavy metal removal from water by magnetite nanorods. Chem Eng J

2013;219:209–16.[2] Niu Y, Qu R, Sun C, Wang C, Chen H, Ji C, et al. Adsorption of Pb (II) from aqueous

solution by silica-gel supported hyperbranched polyamidoamine dendrimers. JHazard Mater 2013;244:276–86.

[3] Lin Z, Zhang Y, Chen Y, Qian H. Extraction and recycling utilization of metal ionsCu 2+, Co2+ and Ni2+ with magnetic polymer beads. Chem Eng J 2012;200:104–12.

[4] Bhattacharyya KG, Gupta SS. Adsorptive accumulation of Cd (II), Co (II), Cu (II),

Pb (II), and Ni (II) from water on montmorillonite: influence of acid activation. JColloid Interface Sci 2007;310:411–24.

[5] Garg UK, Kaur M, Garg V, Sud D. Removal of nickel (II) from aqueous solution byadsorption on agricultural waste biomass using a response surface methodologi-

cal approach. Bioresour Technol 2008;99:1325–31.[6] Kul AR, Koyuncu H. Adsorption of Pb (II) ions from aqueous solution by native

and activated bentonite: kinetic, equilibrium and thermodynamic study. J Hazard

Mater 2010;179:332–9.[7] Crini G. Recent developments in polysaccharide-based materials used as adsor-

bents in wastewater treatment. Prog Polym Sci 2005;30:38–70.[8] Li Y, Du Q, Wang X, Zhang P, Wang D, Wang Z, et al. Removal of lead from

aqueous solution by activated carbon prepared from Enteromorpha prolifera byzinc chloride activation. J Hazard Mater 2010;183:583–9.

[9] Mahdavian AR, Mirrahimi MA-S. Efficient separation of heavy metal cations by an-choring polyacrylic acid on superparamagnetic magnetite nanoparticles through

surface modification. Chem Eng J 2010;159:264–71.

10] Banerjee SS, Chen D-H. Fast removal of copper ions by gum arabic modified mag-netic nano-adsorbent. J Hazard Mater 2007;147:792–9.

11] Zhao X, Lv L, Pan B, Zhang W, Zhang S, Zhang Q. Polymer-supported nanocompos-ites for environmental application: a review. Chem Eng J 2011;170:381–94.

12] Arruebo M, Fernández-Pacheco R, Irusta S, Arbiol J, Ibarra MR, Santamaría J.Sustained release of doxorubicin from zeolite–magnetite nanocomposites pre-

pared by mechanical activation. Nanotechnology. 2006;17:4057.

13] Jiang M-Q, Jin X-Y, Lu X-Q, Chen Z-L. Adsorption of Pb (II), Cd (II), Ni (II) and Cu(II) onto natural kaolinite clay. Desalination 2010;252:33–9.

14] Wu P, Wu W, Li S, Xing N, Zhu N, Li P, et al. Removal of Cd2+ from aqueous solutionby adsorption using Fe-montmorillonite. J Hazard Mater 2009;169:824–30.

15] Aouad A, Mandalia T, Bergaya F. A novel method of Al-pillared montmorillonitepreparation for potential industrial up-scaling. Appl Clay Sci 2005;28:175–82.

16] Abollino O, Giacomino A, Malandrino M, Mentasti E. Interaction of metal ions with

montmorillonite and vermiculite. Appl Clay Sci 2008;38:227–36.17] Belkhadem F, Maldonado A, Siebenhaar B, Clacens J-M, Perez Zurita M,

Bengueddach A, et al. Microcalorimetric measurement of the acid propertiesof pillared clays prepared by competitive cation exchange. Appl Clay Sci

2008;39:28–37.18] Tan I, Ahmad A, Hameed B. Optimization of preparation conditions for activated

carbons from coconut husk using response surface methodology. Chem Eng J

2008;137:462–70.




19] Geyikçi F, Kılıç E, Çoruh S, Elevli S. Modelling of lead adsorption from industrialsludge leachate on red mud by using RSM and ANN. Chem Eng J 2012;183:53–9.

20] Hu J, Chen G, Lo I. Removal and recovery of Cr (VI) from wastewater by maghemitenanoparticles. Water Res 2005;39:4528–36.

21] Abdollahi Y, Zakaria A, Abdullah AH, Masoumi HRF, Jahangirian H, Shameli K, et al.Semi-empirical study of ortho-cresol photo degradation in manganese-doped zinc

oxide nanoparticles suspensions. Chem Cent J 2012;6:88.22] Fard Masoumi HR, Basri M, Kassim A, Kuang Abdullah D, Abdollahi Y, Abd Gani SS,

et al. Statistical optimization of process parameters for lipase-catalyzed synthesis

of triethanolamine-based esterquats using response surface methodology in 2-liter bioreactor. Sci World J 2013;2013:1–9.

23] Chaudhary N, Balomajumder C. Optimization study of adsorption parametersfor removal of phenol on aluminum impregnated fly ash using response surface

methodology. J Taiwan Inst Chem Eng 2013.24] Heidari A, Younesi H, Mehraban Z. Removal of Ni (II), Cd (II), and Pb (II) from a

ternary aqueous solution by amino functionalized mesoporous and nano meso-

porous silica. Chem Eng J 2009;153:70–9.25] Sohrabi MR, Amiri S, Masoumi HRF, Moghri M. Optimization of Direct Yellow 12

dye removal by nanoscale zero-valent iron using response surface methodology.J Ind Eng Chem 2013.

26] Huang P, Fuerstenau DW. The effect of the adsorption of lead and cadmium ionson the interfacial behavior of quartz and talc. Colloids Surf A 2000;177:147–56.

27] Zhu Y, Hu J, Wang J. Competitive adsorption of Pb (II), Cu (II) and Zn (II) onto

xanthate-modified magnetic chitosan. J Hazard Mater 2012;221:155–61.28] Gohari M, Hosseini SN, Sharifnia S, Khatami M. Enhancement of metal ion ad-

sorption capacity of Saccharomyces cerevisiae’s cells by using disruption method.J Taiwan Inst Chem Eng 2013;44:637–45.

29] Zhang F, Lan J, Zhao Z, Yang Y, Tan R, Song W. Removal of heavy metal ionsfrom aqueous solution using Fe3O4–SiO2–poly(1,2-diaminobenzene) core-shell

sub-micron particles. J Colloid Interface Sci 2012;387:205–12.

30] Chaudhary N, Balomajumder C. Optimization study of adsorption parametersfor removal of phenol on aluminum impregnated fly ash using response surface

methodology. J Taiwan Inst Chem Eng 2014;45:852–9.31] Sprynskyy M, Kowalkowski T, Tutu H, Cukrowska EM, Buszewski B. Adsorption

performance of talc for uranium removal from aqueous solution. Chem Eng J2011;171:1185–93.

32] Bhattacharyya KG, Gupta SS. Adsorption of a few heavy metals on natural and

modified kaolinite and montmorillonite: a review. Adv Colloid Interface Sci2008;140:114–31.

33] Sen Gupta S, Bhattacharyya KG. Immobilization of Pb (II), Cd (II) and Ni (II) ions onkaolinite and montmorillonite surfaces from aqueous medium. J Environ Manag

2008;87:46–58.34] Larraza I, López-Gónzalez M, Corrales T, Marcelo G. Hybrid materials: magnetite–

polyethylenimine–montmorillonite, as magnetic adsorbents for Cr (VI) water

treatment. J Colloid Interface Sci 2012;385:24–33.35] Daraei P, Madaeni SS, Ghaemi N, Salehi E, Khadivi MA, Moradian R, et al. Novel

polyethersulfone nanocomposite membrane prepared by PANI/Fe3O4 nanopar-ticles with enhanced performance for Cu (II) removal from water. J Membr Sci

2012;415:250–9.36] Zhu J, Cozzolino V, Pigna M, Huang Q, Caporale AG, Violante A. Sorption of Cu,

Pb and Cr on Na-montmorillonite: competition and effect of major elements.Chemosphere 2011;84:484–9.

37] Yuan P, Fan M, Yang D, He H, Liu D, Yuan A, et al. Montmorillonite-supported

magnetite nanoparticles for the removal of hexavalent chromium [Cr (VI)] fromaqueous solutions. J Hazard Mater 2009;166:821–9.

38] Xin X, Wei Q, Yang J, Yan L, Feng R, Chen G, et al. Highly efficient removal of heavymetal ions by amine-functionalized mesoporous Fe3O4 nanoparticles. Chem Eng

J 2012;184:132–40.



i.org/10.1016/j.jtice.2014.10.025

http://refhub.elsevier.com/S1876-1070(14)00326-5/bib0001






































































































































































































Rapid and high capacity adsorption of heavy metals by Fe3O4/montmorillonite nanocomposite using response surface methodology: Preparation,characterization,optimization, equilibrium

Documents